Re: [Emc-users] OT but built on linuxcnc machines:) affordable color sensor for project (24v model)

[Rafael Skodlar]

Hi Andrew,

On 5/11/22 16:29, andrew beck wrote:

hey guys

i am sure there is lots of experience here

i am looking for a color sensor to sense when a red color chip goes past on
a conveyor

normally the chips are just green and black.  they can be colored in with a
red marker if they are faulty and then we want this sensor to detect them.


Color intensity is important to differentiate between them. Most cameras 
can handle that these days.




currently i am getting prices back at 500 bucks nzd each and i need 4 of
them which seems a bit pricey lol

considering at the other end of the price point i can get a arduino option
for 8 dollars but it is not so plug and play. and is only 5v and needs a
chip too i think.

i want something that can be connected with a PLC to have a simple trigger
on or off .

you guys got any ideas?

regards

Andrew


Tensorflow came to mind right away. It's in the machine learning (ML) 
class. Amazing software if you ask me.


https://www.tensorflow.org/install

Obviously you'll need to do some research to connect things together. 
There is a tensorflow application for Android which I used on my PDA 
(phone is one app on it) to detect objects. It was fun to direct the 
camera towards a TV and detect moving objects on it.


I would start with RaspberryPi, add camera (even a USB kind), and 
install tensorflow. That way you could start experimenting with ML to 
detect different chips, count them, and kick off the bad ones from the 
conveyor belt. A RaspberryPi with a transistor activating an 
electromagnet could do the job. Estimated cost? $60-80 I think.


Be careful with handling low voltage RaspberryPi and more powerful side 
to handle the mechanics.


See examples of what can be done
https://www.tensorflow.org/lite/examples

Came to mind, it would be good to have a machine to sort different kinds 
of screws from a "collection box". US screws mixed with metric ones is a 
total nightmare for me.


Best wishes,

Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] LinuxCNC simulator

[Rafael Skodlar]

Hi Thaddeus,

On 3/15/22 13:52, Thaddeus Waldner wrote:

Hi,

I would like to set up a linuxcnc vm box for simulation/training in my 
classroom. What is the easiest way to do that and not have to mess with special 
kernels?

I installed LinuxCNC in my Virtualbox a number of times. RT kernel 
doesn't have a problem running that way.


I assume you have computers in the classroom. What's on them? I hope not 
trash OS. One way you could go around it with bootup from USB stick that 
would run functional LinuxCNC.


You could install Virtualbox on computers, and have students pull VM 
images with LinuxCNC from the storage on LAN.



I have access to a local exsi server so it would be a small matter to run an 
.iso. I also have a Debian bullseye vm running on it now, which I use for 
projects like this.

Thanks


I also installed LCNC as a VM in Ubuntu server that's running KVM 
kernel. When you have one VM you can just clone it but need to take care 
of hostnames, user login IDs, and networking for students.
Another option for a classroom setup would be to bootup VMs from the 
network.
ESXi is a challenge in many ways so my preference would be Ubuntu server 
with KVM.
RaspberryPi is yet another option. Students could take them home for 
their homework.


Not the easiest ways, just ideas.

Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Pi4 with LinuxCNC and MESA 7i92H

[Rafael Skodlar]

Robin,

On 3/6/22 11:19, Robin Szemeti via Emc-users wrote:

Have you succesfully got the USB3 adapter to work with the Pi?  If so, what
make/model is it, I will happily purchase one if it is known to work. I
wasted an afternoon trying to get one I borrowed from one of my kids to
work on the Pi4 and gave up, so a known working version would be a great
help.



use command
lsusb

to see if the kernel is recognizing that eth device you plugged in USB.

I just plugged in years old USB/eth adapter and get
Bus 001 Device 013: ID 0bda:8153 Realtek Semiconductor Corp. RTL8153 
Gigabit Ethernet Adapter


Realtek chip for ethernet is very common in many adapters.

Also, try to run
dmesg | less

and search for eth. That's the message from the kernel during bootup. 
System files under /var/log are your friends also.
I don't have a running RPi in front of me at this point but it would 
most likely be the same.


Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] who has used thin client pcs

[Rafael Skodlar]

On 2/26/22 09:11, Chris Albertson wrote:


Perhaps I'm just really unlucky when it comes to PC hardware stored in
damp conditions.



Could it be because you power the computers off.  If they run 24x7 then
they stay warm, moisture and corrosion then never happen. I'm
trying to understand "Wake on LAN" so I can shut down a PC but remotely
turn it back on.   In any case, if there is just a bt of heat inside the
chassis then the metal inside the PC will not condense water.


Chris Albertson
Redondo Beach, California



From Kubuntu distribution:

wakeonlan
https://github.com/jpoliv/wakeonlan

Description-en: Sends 'magic packets' to wake-on-LAN enabled ethernet 
adapters.
With this package you can remotely wake up and power on machines which 
have motherboards or network cards that support 'Wake-on-Lan' packets.


The tool allows you to wake up a single machine, or a group of machines.

You need the MAC addresses of machines to construct the WOL packets, 
but, in contrast to 'etherwake', you do not need root privileges to use 
the program itself as UDP packets are used.


=
Package: etherwake
Description-en: tool to send magic Wake-on-LAN packets
WOL (Wake-on-LAN) is a standard that allows you to turn on a computer
over an Ethernet connection. Computers with WOL-enabled network
interface cards can be woken from sleep mode, or powered up from
standby via a BIOS feature.
.
One feature that separates etherwake from other implementations is that
it also supports WOL passwords.

Read the man pages...

Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] who has used thin client pcs

[Rafael Skodlar]

On 2/26/22 08:47, Les Newell wrote:
I'm a big fan of contact cleaner/lube such as deoxit for contact issues 
and I have revived computers with it. However most of the failures I 


No need for that in PCs. More harm than good. Clean contacts with alcohol.

experienced were motherboard death, with complete failure to boot. Even 
conformal coating the motherboards did not help.


even worse.



The last desktop I had in a machine was in the Bridgeport I sold about 6 
months ago. Just before I put it up for sale I ran it though it's paces 
and it worked fine. After I sold it, on the day before it was due to be 
picked up I thought I'd give it another test just to be safe. Zip. 
Nothing. Dead as a dodo. No worries I thought - I had a few older PCs 
for parts, all stored in the barn. Not one of them worked! I eventually 
found a motherboard tucked away in my office which did work. I have lots 
of thin clients but this mill was running Mach3 with two parallel ports 
so it had to be a desktop.


Perhaps I'm just really unlucky when it comes to PC hardware stored in 
damp conditions.


Les


When you store computers in any relatively unclean (dusty) place put 
them in plastic bags to keep the dust and moisture out. Use antistatic 
plastic bags for computer components such as disk drives, PCBs.


Just like anything else, electronics components age. Capacitors are the 
most likely to cause problems after long time no use.


Use a vacuum cleaner and antistatic brush to get the dust out of PCs.
I revived a PC with removing fans, heatsinks and disk drives to get 
access to other parts with a brush and narrow vacuum cleaner nozzle just 
recently.


Put everything back together and use new thermal paste on heatsinks. The 
CPU and memory are much cooler now.


Les, barns are for horses, cows, or pigs not PC computers ;-)

Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] What Would You Suggest?

[Rafael Skodlar]

Hi Ken,

On 2/3/22 16:54, Kenneth Lerman wrote:

I'm considering converting a surface grinder to CNC. To start, I'll
probably just convert the longitudinal and transverse axes.

I'll go with steppers for this -- I'm thinking NEMA-42 motors.

My current Bridgeport clone uses servos and Jon Elson's hardware on a
little Intel Atom Box. I'm thinking of using a Rpi for this. It will need a
minimal display/control panel when completed, but initially will need a
display with touchscreen or mouse and possibly a keyboard. In the long run,
some buttons. and perhaps an mpg might be useful.





I'd like to use a raw Rpi without adding special hardware directly. That
probably means using a USB or ethernet interface to control the steppers.
I'm thinking of using Mesa hardware.

Can someone suggest the most cost effective way to do this? (Although I


Control steppers with Arduino and appropriate size drivers 
https://dronebotworkshop.com/stepper-motors-with-arduino/


and Rpi for GUI part and connections over the network for file sharing, 
backups and GUI as needed.


While these two are NOT industrial grade, they should be able to handle 
your project.



have to admit, that after buying the timing belts and pulleys, the
steppers, power supply, stepper drivers, ..., it's too late to be really
cost effective.). And the surface grinder only cost me $300.

Thanks,
Ken


Recommended reading:
https://openbuilds.com/
https://www.openplcproject.com/runtime/raspberry-pi/
https://www.openhardware.io/

Modern PC motherboards and freaking parallel port are a total waste for 
CNC (retrofit) in 202x! Using multicore CPUs with GBs of RAM to spin 
stepper motors with obsolete G-code is nuts!


Good luck,

Rafael Skodlar
Exclusive Linux user and Open Source Software supporter since Feb 1994


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Question about Raspberry Pi 4 reliability

[Rafael Skodlar]

On 6/13/21 1:49 AM, John Dammeyer wrote:

From: Andy Pugh [mailto:bodge...@gmail.com]

On 13 Jun 2021, at 00:28, Leonardo Marsaglia  wrote:

  I do have a Rpi3

...

I don't know if that's a good candidate to test LCNC.


I wouldn�t expect the 3B to work with Mesa Ethernet as the pre-4 Pis have the 
Ethernet in the USB bus.


I've also read that the Pi4 8GB is not ideal for LinuxCNC.  Apparently it has 
issues with that much memory on the Pi.  It works best with 4GB apparently.

John


Huh? That makes no sense. If nothing else, you can use it for fake swap 
if nothing else.


RPi is NOT good for any industrial use. This thing is for kids to learn 
Linux and open source programming techniques. RaspberryPi PCB design is 
just bad for industrial use: holes for mounting, using glue for CPU 
heatsink, connectors positions and their orientation, sandwich style 
header connector for DIO, silicon waste (for sound and video IO), etc.


To execute G-code using a large footprint computer makes no sense 
anymore. This open source based solution is gaining popularity for 
simple CNC:

https://openbuildspartstore.com/interface/  and other related HW.

Just the fact that Openbuilds is not using LinuxCNC is saying something. 
Dead end: no innovations, limitation in electronics/hardware selection.



--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] GUI Feature Request

[Rafael Skodlar]

Hi Todd,

On 1/5/21 6:33 AM, Todd Zuercher wrote:

I've noticed what appears to be a lot of new development of new GUIs for 
Linuxcnc lately.  Here is an idea that I would love to see implemented in a new 
GUI, but I do not have the programming skills to implement it.  For our 
production machines it would be nice for the supervisors to be able to set up a 
que of g-code files for shifts to run.  The que would be a list of g-code 
files, and their location paths (possibly a network location), the number of 
repeats to run, and possibly a material to be used description.  Then the 
machine operators would simply unload/load the machine, and press start, once 
the required number of runs is completed the gui would automatically load the 
next g-code file in the que and prompt the user what material to load for the 
next run.  How hard might something like this be to implement?


Todd Zuercher
P. Graham Dunn Inc.
630 Henry Street
Dalton, Ohio 44618
Phone:  (330)828-2105ext. 2031


At least part of your proposal seems to be a logistics problem that 
systems administrators have to deal with from time to time.

Think about it as a library of books or paper docs.

Good design should include hierarchy, document (file) naming scheme, 
aging, logging, and index.


${HOME}/jobs//
${HOME}/jobs//.jobs
${HOME}/jobs//2021-01-02/_partx-01.gcode
${HOME}/jobs//2021-01-02/_party-02.gcode

${HOME}/jobs//.jobs
${HOME}/jobs//_party-01.gcode
${HOME}/jobs//_party-02.gcode
${HOME}/jobs//_party-03.gcode
${HOME}/jobs//_partx-01.gcode

${HOME}/jobs/allJobs.list  would contain file names and short 
description about them in one line. Could be implemented in json format.


In general, file names alone should give you a hint where they come from 
or belong to.


Large part of this could be done simply with a set of bash scripts. 
Advanced design should be done in python IMO. Text based design with 
ncurses would work fine in terminal mode. Midnight commander for 
managing files is an excellent example of what I mean.
If the number of files grows into tens or hundreds I would use sqlite to 
keep track of it all.


One thing should be forbidden, spaces in the file names. File names 
should not start with uppercase characters because that requires two 
fingers to type which is not always convenient.


--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] I wonder if I can fix it?

[Rafael Skodlar]

On 1/5/21 9:25 AM, andy pugh wrote:

I bought an FHA-17C actuator from eBay. It was cheaper than normal,
for parts or not working.
It clearly had the cables cut off, so I decided to gamble.

It turns out that things were worse than I hoped.

https://photos.app.goo.gl/p3SL2Bwv2b9yd5XJ9

I am seriously considering trying to make a new flexspline. I mean,
how hard can it be?



Based on what I know about harmonic drives, extremely hard.
Each mechanism has a serial number on both tooth gears to be perfectly 
matched. You cannot mix and match. That explains high cost.


I happen to get my hand on few gears like that. See
https://www.harmonicdrive.net/technology

I know that people play with 3D printed plastic ones for toys etc. but 
that cannot replace metal.


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Raspberry Pi LinuxCNC 2.8 question.

[Rafael Skodlar]

On 12/12/20 1:41 PM, John Dammeyer wrote:

From: Nicklas SB Karlsson [mailto:nk@nksb.online]

To make a link to a file "ln -s filename filename", read manual page
"man ln", I always create the link in the wrong direction first time.


Oh gawd, that would mean using the command line.  How antiquated is that when 
the File manager drop down menu and dialogs do it for you?



John


Example for file name linking would be easier to understand this way
 ln -s original.file myFileName

Note that -s means symbolic link that works across the partitions.

Better yet just use manual pages for linking command
man ln

NAME
   ln - make links between files

SYNOPSIS
   ln [OPTION]... [-T] TARGET LINK_NAME   (1st form)
   ln [OPTION]... TARGET  (2nd form)
   ln [OPTION]... TARGET... DIRECTORY (3rd form)
   ln [OPTION]... -t DIRECTORY TARGET...  (4th form)

There is an excellent file management utility mc (Midnight Commander) as 
Gene has pointed out. It allows you to execute, copy, or look into any 
kind of file in text or hex mode which your windog crap can't do with 
default utilities! A GUI version of mc is krusader (in KDE) with the 
same layout.


Command line rules! Anybody can cut/paste a set of commands to tell 
others what to do for a certain effect. Impossible with GUI. What's also 
very important is the history of previously executed commands. You can 
repeat them or find out where you made a mistake.


Command line has one other advantage, it tells you names of commands as 
soon as you start typing first few letters if you tap Tabs key, example:

l Tab  <--- would respond with
Display all 167 possibilities? (y or n)
if you enter "y" then all commands will show up.

All that Desktop crap is for people with IQ bellow 100 which is what the 
internet has come to. The latest trend is to create same sub-directories 
in users home directory as in Windows and call the folders. What's folded???


That's total nuts. When the first letter of a directory or a file name 
is capitalized you need two fingers to handle it. How is that helpful to 
people with one hand or people with poor vision?


LinuxCNC should work without GUI in the first place. A decent text based 
menu should be enough to run most jobs. Original or old CNC machines 
have no GUI in the first place! ncurses were invented way back to create 
nice menus on text terminals. Clicking on pixel icons does not add value!


Enough of Computer literacy 001.


--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] question for Chris and Thaddeus

[Rafael Skodlar]

On 10/18/20 10:07 AM, John Dammeyer wrote:

From: Jon Elson [mailto:el...@pico-systems.com]
On 10/18/2020 03:57 AM, Chris Albertson wrote:

The above is not the way it works. No one would design
anything like that. But let's say you did. What would
happen? The motor would vibrate and slowly rotate and have
close to zero torque

The Geckodrives G-320 series, for one, does this.  A number
of other PWM servo drives from the well-known manufacturers
such as Servo Dynamics, Copley, AMC and others do the same.
The advantage is there's no dead zone around the null
point.  The disadvantage is a lot of power dissipation in
the transistors and motor.  Many of these drives require a
series inductor in the motor wires to control the
triangle-wave current.

Yes, the motor has zero torque at the null, that is by
design.  As soon as the following error is non-zero, the PWM
duty cycle shifts from 50-50 to an asymmetric ratio, and the
torque increases.  That's the whole point of a servo system.

Jon


I thought all servo motors did it that way.  I can see they could also remove 
the PWM from the windings completely for 0 power but if there's even the 
slightest bit of load on the motor the closed loop would cause a correction to 
hold it in place resulting in essentially the same behavior?

Also, the servo systems measure the current through the windings.  If the motor 
isn't turning then the back EMF from the current changing in the winding is 
pretty small and the motor reaches full current pretty quickly.  Then the 
circuits generally start chopping the waveform based on either a hard clock 
signal or just based on the LRC of the windings.

If the motor is rated at 3A for example, and the applied voltage is 100V the 
current will build up to 3A pretty quickly.  Hence the need for that extra 
inductance to just slow it down a tad so the measuring circuit can see the 
current and the circuit can remove the drive voltage.

Now the current has to decay in the windings and if the say CW drive signal is 
still asserted it will then re-apply the 100V holding the current at 3A.  This 
will continue throughout the CW time of the normal PWM drive rate (say 15kHz to 
25 kHz).

Then if the motor is stopped, when the PWM signal to the drivers has reached 
the 50% point it has a small dead band time for the devices to switch off, and 
then enables the CCW drivers.  Now once again, the 100V (in effect -100V) is 
applied by the other half of the H-Bridge to the motor and the same chopping 
action occurs.

Although the HP_UHU driver doesn't do this there is nothing to prevent a 
processor controlling all this to change the current detection from 3A down to 
0.5A if it hasn't see a request for motion for a period of time.

Or not even switch to a different comparator.  Just change to a different 50% 
duty cycle.  For example, CW for 5%, OFF for 45%, CCW for 5%, OFF for 45%.  
Rinse and repeat.

It's still a 20kHz PWM which is a period of 50uS.  So CW for 2.5uS, OFF for 
22.5uS, CCW for 2.5uS, OFF for 22.5uS.  The current limiting feature still 
comes into effect if the current increases past the example 3A but the overall 
average current for the entire 50uS period has dropped to 10%.  Now the motor 
doesn't heat up as much but is still locked.

I don't know if the HP_UHU with Henrik's module does that.  I'll ask him.

John Dammeyer


These issues were resolved long time ago. Philips office computers from 
1960s, and 1970s used DC motors and clutches for moving folded paper for 
printing and hard paper with magnetic stripe edges for data RW. There 
was no current in the motors when at standstill. Relays either turned it 
on or off.


I remember replacing worn out clutches as part of regular maintenance. 
Imagine cleaning dust from the clutches etc. All electronics was made 
with discrete components, no ICs. Logically, if I remember correctly, 
there was a simple position encoder with incandescent bulb that needed 
to be replaced on schedule.


It's much easier to design modern replacement for similar mechanical 
movements or situations. As already stated in this thread, it has been 
done so why reinvent it? CNC software should absolutely not need to 
bother with this kind of details IMO. Product design overkill kills it.


By the time you reinvent the wheel, you could make money to pay for the 
components bought from fine advertisers in Digital Machinist magazine or 
such.


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] CAD for parts to make on a lathe

[Rafael Skodlar]

Hi Ron,

Here is how I searched for what you are trying to do if I understand you 
right.


CLI is the best way to manage software in my experience. I rarely use 
GUI utility even in Kubuntu workstation.


As a root in my Kubuntu I have the following aliases:
alias acse='apt-cache search '
alias acsh='apt-cache show '
alias agar='apt-get autoremove'
alias agd='apt-get dist-upgrade'
alias agg='apt-get upgrade '
alias agi='apt-get install '
alias agr='apt-get remove'
alias agu='apt-get update '
alias asv='apt-show-versions -p'

On 9/15/20 9:37 PM, R C wrote:

Hello,


I have been using freecad for designing parts, and then milling them on 
a sherline mill, getting the hang of that a little bit.


I tried to get used to FreeCAD too but takes more time than I can 
dedicate at this point.





I have a lathe too, that works with CNC linux, but noticed heard, that 
you can't really  make parts, or g-codes, with it for a lathe.


So for program or related library search I do it this way:
* command
acse lathe

That returns more or less related lines with a description like this:

dxf2gcode - prepares drawings of parts for automatic machine tools
php-tcpdf - PHP class for generating PDF files on-the-fly
ruby-ami - Ruby client library for the Asterisk Management Interface

next command will tell me more about some related package:
acsh dxf2gcode

Homepage: https://sourceforge.net/p/dxf2gcode/wiki/Home/
Description-en: prepares drawings of parts for automatic machine tools
 This program reads 2D mechanical drawings of parts to be fabricated
 and produces G-code tool movement instructions for running on automatic
 machine tools (CNC machines) such as milling machines and lathes.
 .
 This is a graphical CAM (Computer Aided Manufacturing) program.
 It accepts input in DXF, PDF, or Postscript format.  It supports 
milling, drilling, and turning operations, as well as work-holding tabs.


I did not include lines with checksums etc. to make it clearer here.

What wold be a good choice for designing, simple, parts for a lathe, 
that will create g-code for it?


"It accepts input in DXF, PDF, or Postscript format.  It supports 
milling, drilling, and turning operations, as well as work-holding tabs."


That's impressive assuming it work as described.

-
To install the program I would use the command:
agi dxf2gcode

After that you can execute it from CLI or GUI menu.

Of course there is option to install as a unprivileged user:

sudo apt install dxf2gcode
but I'm trying to save keyboard keys ;-)



thanks,


Ron


I have not tried to install and try this program yet but might do it in 
one of my VMs later.


Searching for CAD I received one related app among other things:
solvespace - Parametric 2d/3d CAD

If the output is too long you may narrow it down with
acse cad | grep CAD

here's a description for solvespace:

Homepage: http://solvespace.com
Description-en: Parametric 2d/3d CAD
 SolveSpace is a parametric 2d/3d CAD program. Applications include:
 .
  * modeling 3d parts — draw with extrudes, revolves, and Boolean
(union / difference) operations;
  * modeling 2d parts — draw the part as a single section, and export DXF,
PDF, SVG; use 3d assembly to verify fit;
  * 3d-printed parts — export the STL or other triangle mesh expected by
most 3d printers;
  * preparing CAM data — export 2d vector art for a waterjet machine or
laser cutter; or generate STEP or STL, for import into third-party
CAM software for machining;
  * mechanism design — use the constraint solver to simulate planar or
spatial linkages, with pin, ball, or slide joints;
  * plane and solid geometry — replace hand-solved trigonometry and
spreadsheets with a live dimensioned drawing.

The best test would be to see how the generated (awful) G-code would 
behave in LinuxCNC of course.


This also reminded me about a book I bought in 1990s: The CNC Workbook 
(Addison Wesley 65600) that came with a floppy for DOS program. I 
misplaced the floppy but remember there was a program to simulate CNC 
lathe. Very good book with labs to learn CNC basics.


Good luck,
--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] LinuxCNC legacy or future

[Rafael Skodlar]

As soon as I saw Chris Albertson reply to Greg Bernard "Interesting GUI" 
I knew what will happen because I posted somewhat similar message about 
a year ago. Numerous attacks followed right away. The only surprise was 
a size of the response this time so I decided to split from that thread.


Attacks from some individuals that are stuck in yesterday is very 
disappointing. Others, including myself, that understand open source 
software and related development are in support of what Chris is saying.


IMO it's extremely bad to attack those who come up with "it would be 
nice to have this or that" in existing product. Saying, "it's open 
source, go fork and write your own code" is plain STUPID! We are not all 
programmers! [1]


[1] Before you (individual) attack me, let me tell you that I too worked 
on computer systems in the early 1980s. Oscilloscope, soldering iron, 
spare parts were my tools to get PDP-*, HP-*, S-100, Multibus, back in 
functional order. Those were all industrial systems, some very critical 
like X-ray machines, power plants, electric grids, etc.


Working on electronics side of computer systems kept me away from 
software side so I never became a programmer. To learn at home I had to 
smuggle my first ZX-81 and Amstrad (CP/M) behind the Iron Curtain.


As soon as the PCs became affordable and I made some money here in the 
US I bought XT clone and started learning whatever was possible decade 
before the Internet. One manager told me that my experiments with 
introducing Linux to National Semiconductor in 1994 have no future. 
Solaris was the king for engineers.


As technologies changed, my career veered into systems administration 
supporting software developers mostly on Unix and Linux systems. Since 
then I've seen my share of bad starts in coding projects. They are clear 
indication of what code or style the first few programmers knew.
Lack of OS experience is the most dangerous one. "Running around with 
Windows laptop" and using "porta"potty to connect to Linux is one 
example I've seen too many times and it shows in final products all the 
time.

---

I follow this LinuxCNC mailing list for years mostly as a systems 
administrator. I would like to see and support it in industrial 
environments. Unfortunately the demand is not there. When I talk to 
potential users it becomes very clear that what Chris, Greg, and some 
others were saying is true.


I paid for and supported open source software including buying Linux on 
floppies and CDs. I sent donations to software developers and would not 
mind doing so for LinuxCNC project (hardware and software) if it helped 
me make money. If I had a chance to install it in an industrial 
environment I would suggest the company contributed to
SW development. One of my managers paid hundreds for GNU CD package in 
the mid 90s.


What LinuCNC is doing in its core was done on relatively simple systems 
without GUI long time ago. I believe that it's time for architectural 
change. Split GUI from RT section, and move away from the dependence on 
terrible PC architecture to industrial SBCs make sense to me. Imagine, 
we still have DOS functions in BIOS! And that is "emulated" in virtual 
machines in the data centers these days!


Some SBCs are very expensive, others are cheap but poorly designed from 
the electro-mechanical point of view. For example, RaspberryPi, BB, and 
clones are good from programming side but not ready for industrial use. 
Try to use scope on a middle board sandwiched between the RPi and 
another PCB. What's most important is compatible interfaces developed by 
different vendors.


Digital Corporation and HP encouraged that with open bus architectures 
that handled multiple boards with well designed RT OS in PDP-* buses 
over 40 years ago. Unibus, HP-IB, etc. I don't see that in LinuxCNC 
world. It's great to see some people making $$$ from it but there is no 
big industry or competition behind it.


One comment in that long thread mentioned whyfy as a totally unsuitable 
way to communicate on machine shop floor. Time to read about net 
technologies and related industry trends. Learn about 5G and how it 
makes it possible to rearange CNC robots on the manufacturing floor 
without rewiring the whole shop! Granted, we hobbyists and small machine 
shops cannot afford this from the beginning but all things trickle down 
eventually.


Another comment was negative towards python. Too bad. Python library is 
huge which saves you from writing a lot of code. Python is used 
extensively in Blender, ROS, graphical presentation of data, simulation, 
systems administration, etc. Perl is dead. Most scripts I wrote in it 
are gone with the companies.


I would rather see a resurrection of Multibus like architecture with 
modern CPUs and peripherals for CNC use than running "special test" to 
find out which freaking motherboard with parallel port on PCI card is 
suitable for Linux RT kernel and LinuxCNC. Intel designed Multibus for 
RT 

Re: [Emc-users] Best VM to run Linuxcnc in

[Rafael Skodlar]

Hi Marius,

On 9/1/20 9:29 AM, Marius wrote:

Hi Todd

I just need to run a simuation to show some GUI's and fucntionality. 
Thanks I will try VitrualBox.




My test with 2 different version of LinuxCNC in Virtualbox works fine. 
They share space with other 16 VMs for different tests like ROS, Ubuntu 
firewalls, kubernetes, BSD, CentOS, etc.


Virtualbox is inside my Kubuntu workstation. I say go for it.

It would be more impressive if you had a low cost industrial SBC with 
functional LinuxCNC and use it for simple demo with LEDs, switches, and 
a stepper motor or two.



On 2020/09/01 18:18, Todd Zuercher wrote:
I don't think it really matters much.  Just remember that a VM is only 
capable of running Linuxcnc in simulation mode, because VMs are 
incapable of realtime.

I have only used Virtualbox myself and it works fine for simulations.

Todd Zuercher
P. Graham Dunn Inc.
630 Henry Street
Dalton, Ohio 44618
Phone:  (330)828-2105ext. 2031

.
--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] How to get data to HAL from USB, serial or socket?

[Rafael Skodlar]

On 2020-06-05 10:16, Chris Albertson wrote:

On Fri, Jun 5, 2020 at 9:58 AM N  wrote:


USB should be able to handle real time but think it is complex which might
be a reason to avoid it and there are no support in Linuxcnc.


Mixing USB into RT kernel space is not something I would recommend due 
to it's purpose and position in computer architecture.




Andy posted a clear example that shows a simple way to move data from USB
to LinuxCNC.
  https://emergent.unpythonic.net/01198594294
Read the Python code that you find by following the above link.
Technically this is easy.Now to think about what a user should see on
the hand controller and how he should interact with it.


I currently don't have hardware to test this possibility. What came to 
mind is software tool I use for troubleshooting message passing or 
transferring bytes across local or remote sockets:

netcat sometimes linked as nc
Backup example:

- server
cat backup.iso | nc -l 
- client
nc serv_host  > backup.iso

or
- server
dd if=/dev/hdb7 | gzip -9 | nc -l 

- client
nc serv_host  | pv -b > hdb7partition.img.gz
pv - monitor the progress of data through a pipe

Socket is a file in Unix so you could put it into command above like this:
dd -if= | nc -l 
and connect to it from other systems.



I think e-stop is a desired feature but it will never be as reliable as one
that is wired to the machine that does not depend on computer software.


Right, that should be wired to physical switch/relay to drop power to 
most if not all CNC electronics in case of emergency.




USB does have a short maximum cable length but it should work.   The
example linked above would work for WiFi or BlueTooth too.  Once you see
how it works the method of communication is unimportant, you are simply
setting and reading HAL pins from userspace.



5G technology under development is going to be one of the best ways to 
solve this issue. However, it's wireless implementation will be 
sensitive to interference.



There are however a lot of cheap development boards available which may
make it a very cheap and good choice. There are standard protocol for
keyboard and mouse but do not know about this kind of input you want.

Switching noise might be a problem if connected to power electronics but
not to hand held device.

There is Modbus available in Linuxcnc. It might be a little bit slow
nowadays but expect it more than good enough for buttons. Think it is also
simple and no problem to build a device using one of these development
boards available but otherwise a Modbus Remote IO Module might be a good
choice.


Emergency stop, any plan for this?



--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] cloning hard drive to new SSD

[Rafael Skodlar]

On 2020-04-23 07:56, R C wrote:

well, you can do it with dd, but the details are tricky at times.


You are not just "cloning" a disk, like you used to copy a disk. I drive 
has multiple things, like an MBR, and


partitions and such. Most of the time to just clone a disk, copy it if 
you will, works, but what one should really


do is copy partitions, figure out how big they need to be. And then 
there's things like layout etc etc.



What clonezilla etc do, is use the same stuff, pretty much dd, except 
they have some "logic" in their software on how


why mess with installing that when core utilities and a bit of bash can 
do the same.




to copy those partitions, MBR/fat and all that. Some drives don't care 
about alignment for example, and work not that


optimal, others might not work like that.

A "fail safe" way to do it is use sync or so.  There are different ways 
to do that though.  If you do not want to "hammer"


the drive that is failing, you still create an img/iso, mount that and 
then create a disk by partitioning/formating it, and after


that you use the mounted image with sync or so to move the files to your 
new disk.



dd works really well for copying devices, especially if they are the 
same or very similar, or for copying a device to a file/iso/img


or from an "image file" to a CD/DVD or SSD memory. Mostly used for 
RapsberryPi and such.


dd is just a tool, very versatile and powerful, but as with all tools, 
you need to know the details on how to use it and for what.


Bravo. I'm horrified reading recommendations to use dd for cloning files 
on storage devices. dd copies fragmented files as is so you are messing 
new drive for performance issues from the get go.


The best use for dd is in computer forensics and virtualization 
environments for the same purpose or "deep troubleshooting" to find out 
why a VM has issues.


I mentioned rsync to be one of the best and most effective utilities for 
cloning locally or remotely in my experience. One option "--dry-run" 
allows you to see what will happen without making a big mistake if you 
are not careful.


There is another way to copy/clone files I learned in my Sun OS/Solaris 
days.


Drive 1, /dev/sdb mounted /tmp/disk1
Drive 2, /dev/sdc mounted /tmp/disk2

(cd /tmp/disk1; tar cfp - *) | (cd /tmp/disk2; tar xvf -)

* assumes all directories but you can just name a few for the process.

tar utility was always on all Unix systems, rsync was not. There is one 
more utility worth mentioning: cpio
You can change file ownership during file copy/clone process which is 
some times necessary.


I miss days when utilities mt and rmt were needed for files 
manipulation! It's magic to watch tape reels spin one way or the other. 
No silly G-code needed to spin reels or have the drive suck tape into 
vacuum chambers  ;-)


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] cloning hard drive to new SSD

[Rafael Skodlar]

Andrew,

On 2020-04-22 19:22, andrew beck wrote:

Hey guys.

Just a quick question here

I recently heard some funny clanking noises in my old 2nd hand hard drive
on my VMC and thought I better change it out and get a SSD in there.

I have a bunch of stuff loaded onto the hardrive for probe basic gui and
other stuff and would like to clone the drive and keep everything.

I can manage a windows cloning I am just not sure if the process will work
on a linux system.  I am using a crucial brand SSD and can download the
drive cloning software (it is rebadged acronis cloning software)

anyway some help would be appreciated.


regards

Andrew


Linux comes with all software to manage system in any way needed.

There are different scenarios you can use to do what you need.

You can start with booting up from a CD or it's image on USB stick if 
the main drive is not booting up. Otherwise just add second drive and 
bootup. Find how is second drive recognized:

dmesg | less  <--- in x-terminal
look for lines SCSI, ATA and such to see what the second drive is 
recognized as.


To partition the disk I use fdisk command; see 'man fdisk' for details. 
Make sure you do not do it on original drive!!!


I happen to have two drives and mount the second one like this:
/dev/sdb1 1.4T  877G  441G  67% /backup
/dev/sdb2 672G  208G  451G  32% /virtual

Mount old and new drives and sync files across.

For example:
/dev/sda is old drive,
/dev/sdb is the new one.

Let's assume there is only one partition for the files on old drive 
/dev/sda1


You would make one partition on the second also. However, you need to 
add a swap partition about 2 to 5 times RAM size to make "Linux happy". 
Swap can be found this way:


swapon -s
Filename   TypeSizeUsedPriority
/dev/sda3  partition   8388604 0   -2

mkdir /mnt/sda1
mkdir /mnt/sdb1

mount /dev/sda1 /mnt/sda1
mount /dev/sdb1 /mnt/sdb1

In case there is more than one data partition you might want to do the 
same on the second drive. In any case, I would use a separate partition 
for /home so that work files are in one place separate from the OS itself.


When the partitions are mounted you may use rsync to sync files to the 
new drive partition. Example:

rsync -av /mnt/sda1/ /mnt/sdb1
rsync -av /mnt/sda2/ /mnt/sdb2<--- for second partition.
and so on.
Pay attention to '/' (slash) at the end of source directory to ensure 
correct way of file sync.


rsync is great because you can transfer files across the network to 
other systems. Read man pages for details

man rsync

If you want to know how long it takes for task to complete use
time rsync -av /mnt/sda1/ /mnt/sdb1

Note, the above commands need to be executed as user root. Alternative 
is to prepend sudo to the above commands but I prefer becoming root this 
way:

sudo su -
and enter password you use for login. That's assuming root does not have 
special password. I always have one x-terminal tab dedicated to root for 
sysadmin work. Note that the prompt will change to # at the end.


One more thing, if the new drive has a partition on it, very likely, 
then your OS might automount it during login. You need to unmount it 
before you use fdisk to delete and create new Linux partitions on it.

Check with
df -h

Good luck,

--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] linuxcnc on modern hardware

[Rafael Skodlar]

On 2020-04-19 17:46, Bari wrote:
The bottleneck with CNC machines is not Linuxcnc, it's the mechanical 
hardware and materials. All modern CNC machines may easily be controlled 
by LinuxCNC on older hardware. What we really need is new tech and 
materials that are cost effective to use to make end products and be 
machine faster and easier.




Why the hell you keep promoting this OLD CRAP??? That's because you 
never read or watch about anything worthwhile!


"All modern CNC machines may easily be controlled by LinuxCNC on older 
hardware"  What a load of crap is that? With that kind of thinking no 
wonder US never adopted metric system and have to use Soviet era 
technology to send American astronauts to space from Russia.


An armchair project manager like yourself could easily spend some time 
learning about physics, mechanical engineering and materials science to 
provide us with your "insight" to accomplish this.




Here you go again genius/scientist/machinist and spokesperson for 
LinuxCNC and CNC industry in general! No arguments just personal 
attacks! So typical for your kind. How many patents are there under your 
name?


Just the fact that you are top posting and hide behind incomplete name 
tell it all!


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] linuxcnc on modern hardware

[Rafael Skodlar]

I see this tread circling around the same issue as the one after I 
posted my comments on this subject matter when I was exploring possible 
options for new CNC machine.


On 2020-04-19 10:12, René Hopf via Emc-users wrote:

Hi,
For reasons I dont understand, many people prefer to use extremely legacy
hardware to run linuxcnc.
the hardware list in the wiki is extremely outdated, and there is no real
recommendation on what to buy currently.


Tell me about it. Threads in the past few months will tell you how much 
beating I was taking for bringing up related annoying issue with need 
for testing motherboards for suitability with RT kernel, etc.



just for fun I tried to run it on modern hardware, and it worked really
well.
I used a Pentium Gold G5400 on a rog strix h370 mainboard(dual ethernet),
booting from nvme.
64 bit, and efi boot, you cant legacy boot from nvme.
I tried the 4.19 and 5.4.19 rt preempt kernel from the debian repo.
both get extremely good latency of 1500-1 ns, depending on bios
settings and kernel args.
but even the defaults are fine.
I tried on a few other non-legacy PCs, and could not find a single one that
doesnt work well with rt preempt.


That's great. Trouble is not everybody can afford or have means to test 
different platforms these days when you need to buy most of this stuff 
from the Internet.


HW architecture needs to be taken into consideration in any case. What 
good is it to run LinuxCNC on multi-CPU and multi-core boards when only 
a fraction of it is actually used for CNC functionality?


Mix of GUI and RT on the same SBC is just not a good idea. Yet, some 
will argue about it again and again.


Those who want to use old junk just use old junk. I'm sure that this 
kind of messages or threads turn potential manufacturers, small or large 
(?), away from LinuxCNC.




imho rt preempt is the future, as it is maintained, and you can get modern
kernels with it.

just trying to kill the rumor that linuxcnc doesnt work on new PCs.

Rene


Thanks,

--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Gecko Failure

[Rafael Skodlar]

John,

On 2020-04-05 21:24, John Dammeyer wrote:

Anyone ever run into this sort of thing with a G213V driver?
http://www.autoartisans.com/mill/G213V-Failure.jpg


This is typical for poor connection in environment with mechanical 
vibrations as in CNC. Gecko driver itself has nothing to do with it I 
believe. In my hardware support years I checked female parts of pins to 
make sure they are narrower than the male part thickness every time 
there were "power related issues".


You can use a scribe or a larger needle to push halfs of female section 
towards the center. In some cases I pressed that part out of the plastic 
first then used pliers to squeeze the metal part then pushed it back 
into plastic part of connector.


Power connectors for hard drives and floppy drives needed that treatment 
in old PCs using much less current than your CNC motor. Same in power 
supplies on mainframe systems.


Without seeing that kind of connector from close it's hard to say if 
this is doable or needed in your case. The burned out connector needs to 
be replaced first and male part needs cleaning. In some cases you need 
to re-solder it because it gets lose during overheating.


You can try to use different kind of connectors with screws to tighten 
the connection. That depends on space between the pins on Gecko side.




It was running the Knee with a 1200 oz-in motor and 60VDC power supply.

John


I hope it's helpful.

--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Open source CNC architecture

[Rafael Skodlar]

On 2020-04-02 03:14, Przemek Klosowski wrote:

On Wed, Feb 19, 2020 at 1:51 AM Rafael Skodlar  wrote:



Think about LinuxCNC and it's packages. Using Slackware method you could
try to use different version of LCNC to see if it's good for your CNC
setup. If the new version fails (breaks your tool?) you could run
"cnc-admin script" to roll back, i.e. relink the app to previous version
and start it.

/opt/linuxcnc-v2.6/bin  <- binaries or scripts
/opt/linuxcnc-v2.6/etc  <- config files
...





/opt/linuxcnc-v2.7/bin
/opt/linuxcnc-v2.7/etc
...


What you suggest is not that simple. For starters, there are cross-package


It's simpler than have a full blown workstation to do the work of what a 
much smaller distribution could do. Now you have to manage much larger 
number of packages.


ls /usr/lib/ | grep python
python2.7
python3
python3.6
python3.7
python3.8

You just tell which one you want to use and all is fine. Same can be 
done with other software.


I believe that LinuxCNC would benefit with using Yocto because you can 
build software for different platforms at the same time. Used widely in 
automobile industry, avionics, and robotics as well.



dependencies, so in general you will have to carry various system libraries
required by each version. Then, modern programs use various forms of IPC,


We already have different versions of libs. Ever looked under */libs*?

Anaconda is the best utility to manage different versions of python in 
user space. Yocto takes it to another level.



so they need version specific versions of those IPC endpoints (things like
DBus, etc). IN the end, you can do it, but you replicate huge parts of the
OS. RedHat/Fedora tried to do Modularity, which is something like what you
propose except it turned out that you could have multi-version availability
but not multiversion installability (it was easy to switch versions, but
you could install only one). They put a lot of effort into making
dependencies work automatically, but in the end it turns out that lifecycle
management (patching/updating) is hard in a multi-version world: multiple
versions of multiple programs lead to combinatorial explosion of
dependencies and unresolvable conflicts when one program depends on ver. X
of something and another one on ver. Y. Currently Modularity is in retreat.



Software written so tight that it works only with a certain library but 
not on a newer one is just bad software. I hear this from other 
engineers developing for large software houses here in Silicon Valley.



The technology to do that exists---containers like Docker or Podman. The
downside is that your system is now a mess of versions, and you need to


What mess??? It's a flexible design that solves dependencies issues. 
When setup right it works automagically.



worry about patching and updating them. Containers provide a partial
solution to the Modularity problems---you can isolate such conflicts to
separate containers, but you still need to worry about lifecycle management.

If you are serious about those issues, read up on containers and
modularity---don't invent your own solutions, as a lot of people tried to
do it right and it's worth to learn from their experiences.


Using your logic nobody would ever invent or improve anything. So what 
if they failed? Somebody with make a difference then others will follow 
as usual.


When you spend as much time using and managing virtualization on servers 
as long as I have, install and setup containers etc. then ... I got 
tired of this kind of responses when you don't read on the subject 
matter before telling me I don't know anything.


The fact is that LinuxCNC is a mess in a sense because too many insist 
on basic real time functionality and top it off with full GUI instead of 
advancing from that original architecture. It helps if you study what 
large commercial manufacturers of CNC machines make. I provided links to 
some already.


How about telling others to go read about going metric and stop using 
STUPID system that is based on hundreds years dead British king butt or 
finger size?


G-code is archaic also. It's like programing PCs in assembly language in 
1980s. Some bad things just never die.


--
Rafael
Linux specialist since 1994


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Open source CNC architecture

[Rafael Skodlar]

On 2020-04-02 07:13, R C wrote:

my 2 cnts;


I work in/with HPC, and run into that stuff all the time, and it is 
unavoidable.


what stuff? It would be much easier to read old thread if related lines 
were in related place.




Since HPCs run diskless, and boot in/from a network, we simply build a 
complete new image, (and keep


the older ones around). We never even update an image, we simply build a 
new one from scratch, since


And stop services when Saltstack, fabric, or mush could be used to 
update software without much downtime ... Nonstop rebuilds are not a 
solution to everything.




an update on an existing system never works and it is easier to rebuild 
a repo (at least in RHEL it is).


Libraries etc, specific to applications either get relocated, or are 
merged with the OS ones on a virtual file system.



Of course that is pretty much undo-able, impractical, unaffordable to do 
at home, so what I do: I use different drives with


That's possible since 1990s when we could buy first removable drives for 
PCs. Some were IDE, other SCSI based. Same idea as in IBM, DEC, HP, and 
other mainframe computers.




separate installs (I use these now very inexpensive CRU data trays to 
swap drives, and SSDs are really inexpensive now)



And indeed, let's not even get started on "rolling back" within an image.


containers; that's one of these things that don't seem to work 
consistently yet. I know people (at work) that are working




Ever heard of Google? See more bellow

with it, developing in it, but I have not seen it work reliably/stable 
yet. It will definitely go there, but as of yet, at least


??? It's being used in so many places that would spin your head.
https://www.sdxcentral.com/articles/news/t-mobile-to-slash-30m-in-cloud-costs-with-kubernetes/2020/04/



at scale, it is not working. (there are lots of issues that come down to 
latency/timing and rdma issues and we don't


even use real time kernels etc. most of what I do is based on RHEL and 
application specific RHEL 'flavors')



as I said, just my 2 cts,


Ron

I study technologies while you watch sports...

--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] connector ignoramus

[Rafael Skodlar]

Hi Dave,
On 2020-03-18 11:05, dave engvall wrote:

Hi,

Or maybe it is web search dumb. Usually I do better but:

I have the osh park board for Rpi4b to 7i90 ordered. However I think I 


Which OSH board? Interested to know for my own use.

can plug a 40 pin (2 x 20) idc directly into the pi and come off the 
other side of the board with a 26 pin (2x13) directly into the 7i90. No 
cable to fool with. Just carefully making mounting that matches the 
physical requirements. Using a right angle connector to get to the 7i90 
would make things neater. IOW   40 pin to osh-park board to 26 pin (R/A) 
or straight to 7i90.


Quality of flat cables properties is not always the same. When IDE and 
SCSI disk drives were top of high tech, there were two types of better 
flat cables. One kind had a shield on one side, the other had twisted 
pair wires next to each other. I keep old flat cables around for 
critical CNC or other embedded use. I cut off unneeded sides in some 
cases and add new connector.


You have to be sure which side you put the connector on. You may 
accidentally "rewire" pins. Guess how I learned that?


My small mill now has glass scales (0.0002)  on X and Y so making simple 
stuff is easy, especially out of aluminum. I do almost nothing in Al so 
I am continually amazed at how easy it mills. Especially with Zr 
coating. Surplus Boeing from a long time ago; before they closed the 
surplus store.


Any dust related problem?



Part numbers from Mouser, Digikey, Allied would be most helpful.

TIA

I suspect I'm house bound ... with a 100' leash to reach the shop for 
several weeks but UPS/BBT (big brown truck) still works. Stats say one 
is fairly safe unless you have a condition that makes you more vulnerable.


I carefully clean any package delivered to my house. You never know 
what's inside or who touched it intentionally or by accident. This 
reminds me of the following from the 80s.


One company I was working for was a DEC representative in part of EU. 
Many spare parts came wrapped in bubble packaging. That was a novelty 
east of Iron curtain.
Some coworkers kept popping bubbles. That stopped only after I said that 
bubbles might contain special gas sent by "capitalists" to trigger 
people to overthrow the commie paradise.


What if ...? Never mind, all knowing genius will attack me anyway.



This state has been really hit hard; the nursing home in Kirkland didn't 
help. Even worse is that a person can pass it on before they have symptoms.


https://informationisbeautiful.net/visualizations/covid-19-coronavirus-infographic-datapack/?fbclid=IwAR0CAPMO2oHOLZyB07vBhdbLvAjUcQ6RiifnOZJDt6sdMrSV1MWL-2Y0bv8 



Hang tough, be careful.

Dave


Thanks for civilized email exchange and helpful link.

--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Open source CNC architecture

[Rafael Skodlar]

On 2020-03-17 01:30, Robert Murphy wrote:

Raf,

Get a 7i76e and you are done.

Don't scratch your head and touch sensitive components.


That's just one possibility which depends on how much hair and size you 
have. Just walking around certain kind of floors you generate static 
electricity that can kill modern electronics very fast. I have yet to 
see technicians in IT using anti-static wrist strap.I keep two in my 
toolbox.




What you have said is technically and correct and in all the books,it's
even what the teach in tertiary educations when you're working towards
your certs.  it's not exactly what I've seen in the field. That's was
day in and day out repair EFTPOS terminals, mainboards of pokies and
other sundry products by a certain large supplier of EFTPOS equipment.


When you use a not so common (?) acronym it's nice to spell it out.


Even with the latest exercise equipment with all the bells & whistles
the biggest issues are failure of output drivers due to lack of
mechanical maintenance by owners.

Not that I'm advocating working with electronic equipment in a vinyl one
piece jump suit with balloons attached.

I'm not too sure of the certification status of the RPi for industrial
use. Where as there is a variant of the BBB that is.


It depends on what temperature range components you use and how you 
build the boards.


https://thepihut.com/products/italtronic-din-rail-raspberry-pi-model-b-plus-case

https://revolution.kunbus.com/revolution-pi-series/  pay attention to 
the diagram on the sides. 12-24V DC. And QR code to find additional 
information easily. It looks to be very good product IMO, all based on 
open source. [1]


[1] And I was told earlier to get lost after I commented on LinuxCNC 
architecture issues.


DIN is a German standard I'm aware of since I mixed neutral power line 
with hot 220V. I built a simple one transistor receiver and used a 
neutral wire as an antenna. That idea (?) came from something I was 
reading in the 70s if I remember correctly. One day I connected my 
detector to "antenna" on the wrong side ;-(


Anybody used oscilloscope to troubleshoot switching power supplies? Not 
modern battery powered scope mind you. That was fun.


Since then I touch unknown circuits with one hand but only if I have 
good shoes on. We have protective gloves now but that was not available 
in commie paradise.


Building radio detector was one step in my way to learn electronics 
trade. Problem is that stupid plug standards in continental EU allow you 
to plug single phase power cord two ways.


American standard for power plugs is way better but I don't recommend to 
use N for an antenna unless it's an emergency and the whole nation is in 
kernel panic mode. Oh wait, we are.



--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Open source CNC architecture

[Rafael Skodlar]

On 2020-02-19 00:11, Robert Murphy wrote:


On 19/2/20 5:47 pm, Rafael Skodlar wrote:

. snip

examples of embedded system customization:
https://wiki.st.com/stm32mpu/wiki/OpenEmbedded
http://wiki.openmoko.org/wiki/OpenEmbedded

To naysayers; there's gold in those hills:
https://www.crunchbase.com/hub/hardware-companies#section-recent-activities 



One more comment on application distribution. It's based on my
experience from few years ago where "cloud service" was designed on
Centos distribution. Programs, libraries, and configuration files were
packed as RPMs and installed as such. I proposed much simpler and more
flexible packaging, tar files as in Slackware. The advantage would be
in architecture where more than one version of service could be
installed in parallel without interfering with another one.

Active service would be a link 'from the top directory'. Switching
from one version to another one would be simple:
- service  stop,
- relink top directory,
- service  start.

What we were doing was order pizza, stop services, install RPMs on top
of old ones, start services. That process took about 3 hours on all
servers in the evenings when the network traffic was low across the USA.

In some instances QA requested we roll back. Argghhh! Obviously, same
install process that lasted another 3 hours or so ... My proposed
architecture would be done in less than 15 minutes in my estimation.

Think about LinuxCNC and it's packages. Using Slackware method you
could try to use different version of LCNC to see if it's good for
your CNC setup. If the new version fails (breaks your tool?) you could
run "cnc-admin script" to roll back, i.e. relink the app to previous
version and start it.

/opt/linuxcnc-v2.6/bin  <- binaries or scripts
/opt/linuxcnc-v2.6/etc  <- config files
/opt/linuxcnc-v2.6/lib  <- libraries
/opt/linuxcnc-v2.6/man  <- manual pages
/opt/linuxcnc-v2.8/log  <- log files

/opt/linuxcnc-v2.7/bin
/opt/linuxcnc-v2.7/etc
/opt/linuxcnc-v2.7/lib
/opt/linuxcnc-v2.7/man
/opt/linuxcnc-v2.7/log

/opt/linuxcnc-v2.8/bin
/opt/linuxcnc-v2.8/etc
/opt/linuxcnc-v2.8/lib
/opt/linuxcnc-v2.8/man
/opt/linuxcnc-v2.8/log

Latest release would be:
/opt/linuxcnc -> link to /opt/linuxcnc-v2.8.
Your path to LinuxCNC binaries or scripts would always be
/opt/linuxcnc/bin etc.



Dunno why you are calling that the "Slackware Method" and I've been
running a 24/7 Slackware server for years. Never in all my years I have
seen that method used in Slackware or mentioned in the docs. Actually I

Really?
http://docs.slackware.com/slackware:package_management
" Slackware packages can be found with any of the following extensions:

tbz - Slackware package archive compressed using bzip2
tlz - Slackware package archive compressed using lzip
tgz - Slackware package archive compressed using gzip
txz - Slackware package archive compressed using xz

... Slackware does not automatically track dependencies and install 
dependencies when you install a file"


What's simpler than that???


can't recall any official Slackware packages installing into /opt and
linking as you say. Have you used Slackware ?


Slackware is the first one I used in 1994. Before Ygdrasil, Redhat, 
Mandrake, and Debian.


Don't take everything so literally. I was just giving an example for 
files locations and a method to switch versions without major changes. 
See above.


Ubuntu uses /srv a lot. Put LinuxCNC wherever you want, /cnc, 
/usr/local/cnc for what I care. What I wanted to point out is the 
simplicity in managing multiple versions of the same application or 
service on the same system if tar was used rather than other packaging 
system. Dependencies will always be an issue as it was pointed out in 
other threads a lot recently.




Puppy Linux and Salix use a "Plug in" file system for apps. But usually
a squashfs file system that can be loaded on boot.

If you read the Linuxcnc docs you'd be aware of a Run In Place install.
Which is the recommended method before a full upgrade.

try this one:
emc2-dev/src/emc/motion/teleop-notes points to
http://jmkasunich.com/pics/emc2-motion-dataflow.pdf show me that file.



I'm beginning to think you don't know too much about Linuxcnc and are
trying a push it something that suits your needs or business model.


Keep thinking. I can't read. I'm glad to see you know _everything_ and 
how things should be. I observe trends in the industry and point them 
out to PC parallel port zealots that suggest using motherboard with 
multicore GHz CPUs, and EPP PCI adapter to push 8 bits at a time. So far 
I haven't pushed any business model.




Why not do your own crowd funding, fork Linuxcnc and pay for someone to
break it up and rearrange it to suit your needs.


That's because I don't have your permission to fork from PPC or as much 
money as you do.


Why did I decide to resurrect this thread? My first response did not go 
through a few wee

[Emc-users] EMC email in /dev/null land

[Rafael Skodlar]

This is a test for LinuxCNC broadcasting system.

For over a week I was wondering if my replies to the thread "Open source 
CNC architecture" ended in /dev/null. Because there were no other 
messages from the list it crossed my mind that _somebody in power_ 
removed me from the mailing list because they don't like my position on 
that subject matter. Saved emails indicate that some replies to my 
messages were rather hostile.


Is it possible that some email is still stuck in the queue?

My solution to this kind of issue is a proposal to setup a cronjob that 
sends out a short note about mailing list statistics once a week. Just a 
few lines or a link to related web page. That way we would know that the 
mailing list and our subscription, email servers, filters, etc. are all 
still functional.


Now that the messages are flowing in again I'm assuming that you guys 
are simply too busy with your CNC machines to bother with the mailing 
list. This is good sign.


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Open source CNC architecture

[Rafael Skodlar]

s on all 
servers in the evenings when the network traffic was low across the USA.


In some instances QA requested we roll back. Argghhh! Obviously, same 
install process that lasted another 3 hours or so ... My proposed 
architecture would be done in less than 15 minutes in my estimation.


Think about LinuxCNC and it's packages. Using Slackware method you could 
try to use different version of LCNC to see if it's good for your CNC 
setup. If the new version fails (breaks your tool?) you could run 
"cnc-admin script" to roll back, i.e. relink the app to previous version 
and start it.


/opt/linuxcnc-v2.6/bin  <- binaries or scripts
/opt/linuxcnc-v2.6/etc  <- config files
/opt/linuxcnc-v2.6/lib  <- libraries
/opt/linuxcnc-v2.6/man  <- manual pages
/opt/linuxcnc-v2.8/log  <- log files

/opt/linuxcnc-v2.7/bin
/opt/linuxcnc-v2.7/etc
/opt/linuxcnc-v2.7/lib
/opt/linuxcnc-v2.7/man
/opt/linuxcnc-v2.7/log

/opt/linuxcnc-v2.8/bin
/opt/linuxcnc-v2.8/etc
/opt/linuxcnc-v2.8/lib
/opt/linuxcnc-v2.8/man
/opt/linuxcnc-v2.8/log

Latest release would be:
/opt/linuxcnc -> link to /opt/linuxcnc-v2.8.
Your path to LinuxCNC binaries or scripts would always be
/opt/linuxcnc/bin etc.

--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] RPI4 is pretty close to a decent machine

[Rafael Skodlar]

On 2020-02-18 15:23, Chris Albertson wrote:

This is a problem with many people using Linux to do something complex like
build a CNC system or using ROS to build a robot.  They are doing something
complex with Linux before they try to do simple things with Linux.


Among other things, ROS is wrongfully named Robot Operating System, when 
it's a service. Must be a tradition in naming things the wrong way. 
Cookies for example are baked in bakeries not cookeries. Therefore they 
should be called bakies!



The best approach is to fist use Linux for everyday tasks, like reading and
writing emails, watching youtube videos.  Simply use it for all your normal
tasks.Get rid of your Windows system.


It takes some courage and experimentation to use Linux for your work. 
Granted, too many silly software companies don't create their programs 
for Linux platform since 1990s when they wrote them on Unix platform in 
the first place.


This month is 26 years since I started using Linux on my home and work 
computers exclusively. Over the years I tested all kinds of 
distributions and settled on Ubuntu based Supermicro servers in my 
garage and Kubuntu for my workstation and a laptop.


Some silly recruiters ask me to send my resume in "word format". I 
respond with either simple ASCII file or PDF exported from LibreOffice 
Writer. LibreOffice set of programs is good enough for office work. 
There are other utilities and programs that make my life exciting. I use 
Kate for writing bash and Python scripts, Jupyter for testing python 
scripts, etc.


There are other great programs for taking notes, editing or viewing 
pictures, creating 3D pictures with Blender, Thunderbird for handling 
email, listening or editing music and videos, etc.


I rarely turn on large TV because I can watch TV programs on HDhomerun 
that connects over LAN to SiliconDust TV box in my garage. That's 
running in parallel to other programs for my work.


One of my favorite utilities is Virtualbox which allows me to test 
different Linux distributions (Centos for example) or other OS like BSD 
that I need for professional work. I installed LinuxCNC as a virtual 
machine to see how it looks like. RT kernel is not important in this case.


Virtualbox is available for Windows also. You can create Linux VMs for 
testing. The other way around is possible also. Yes, start using Linux 
for most if not all of your work.



It is a little like buying a harbor freight mini lathe and then your first
project is a working gas airplane engine.



Oh my, I won't be able to build a working gas engine. Little Greta told 
the World that gas engines are forbidden now anyway.


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Open source CNC architecture

[Rafael Skodlar]

On 2020-02-15 01:05, Bari wrote:

On 2/15/20 1:42 AM, David Berndt wrote:

How did we get to the point where we decided that the goal is a 
"relatively simple embedded system"? I for one am not looking to trade 
off the current gui and it's features for what you describe.


It seems like a lot of this thread seems to steer itself in the 
machinekit direction in terms of apparent goals/ideals, so I guess I'd 
ask, why not start there?



We haven't. Someone just thinks new or different is better. Someone that 


And you? You think that doing things the same old way is better? Steam 
engine driving machines with pulley transmission in your work shop?


When I was little, we had to pull the weeds by bare hands on a small 
farm. Our family would prefer doing it this way:

https://agerris.com/
https://farmwise.io/
https://www.deepfield-robotics.com/en/

Only people who think differently in a positive way make a positive 
difference. Have you heard about little automobile company Tesla that 
happens to be a bike ride from my home?


has not really worked much with CNC machines but does have an idea that 
they think would make things better. It would be newer without 


In my case, true. However I see what others with lots of $$$ in the 
industry are doing. It would be beneficial for small shops or hobbyists 
to use similar things on a smaller scale. No need for running test on 
the motherboard to ensure RT kernel will work as expected.


That's how PC industry started in the 1970s. People started copying SW 
and HW ideas from mainframes. I'm not inventing anything here. Just 
pointing to possibility to simplify LinuxCNC and expand it's use.


explaining why it would be better, only telling us that it will be 
better. So they think newer is better just because it is newer. I hope I 
have cleared this up.  :)




Explanation or hints were given by others and myself. For some it's 
obvious, for others it's not. History is full of naysayers to everything 
new.



Newer is better. Look into my eyes and repeat "newer is better".


Looking into your anonymous eyes:
Metric system is 200 years old. Some of us know it's better since my 
grandma told me so long ago. Yet whole North American Continent is stuck 
on archaic system that is based on dead British king finger, foot, and 
butt size. So "newer is better".


"Make America Metric" I say.

Need explanation?

--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Open source CNC architecture

[Rafael Skodlar]

On 2020-02-11 01:04, Chris Albertson wrote:

   I said people *want* to use CNC like a laser printer. Most setups are
not that good.  It is a goal and if designing a new system.  It is good to
set the bar high and try to do what can't be done today.What I really
meant was that with a printer, all the critical timing happens in the
printer.  There are no servo-loops on the PC and you don't need a real-time
OS to print to paper. I think people want CNC to work this way.



That's how some CNC machines work. I came across a small woodwork 
business owner with very nicely garage that was converted into workshop. 
Win PC in the corner for designing parts in CAD, large Axiom CNC machine 
with a pendant to control it. Tiny LCD is all that's needed to select 
the job, i.e. file from a USB stick.


The owner did not know what's inside the CNC machine itself and he 
doesn't care. That's what you say Chris I think and I agree with.


I don't know if there's an option for connecting that CNC machine to 
LAN. I would not use wireless connections for such as it's security 
issue due to hacking possibility in the neighborhood. Another 
possibility is noise on WhyFy frequencies from appliances, bad power 
lines, etc.


In any case, that CNC machine does not have or need a PC computer with 
modern GUI interface connected directly to run it. That's why I started 
this discussion. X-windows is waste of resources, it's another thing 
that needs to be maintained and updated in some instances. Too many 
things to go bad in what's supposed to be a relatively simple embedded 
system.


The tiniest user interface would be possible using extended ASCII 
characters as in old DOS. We used to play with that in old email 
signatures. My fun with ASCII art in the 1990s:


. ..   .  . . . . o o o o O o
  ___   ___   _  O
 |  Rafael Skodlar| |   LINUX   |   ]OO|_n_n__][.
 | ra...@mydomain.com |=|  Support  |=||=[]_|__|)<
  oo~~oo~   ~~oo~~~oo~~  oooo  'oo -| oo\_
+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+

Add colors, lines, and block characters and ... you see the picture 
that's taking extremely little memory by today standards.
Simple ASCII DRO + G-code scroll window and 4x4 keypad would be enough 
for most work. No need for keyboard, mouse and X-windows on large monitor.



On Mon, Feb 10, 2020 at 3:06 PM Bari  wrote:


A laser printer is a good example of how people really want a CNC mill to
work.  You design you document on the computer then press "print" and the
printer creates it.   After the last of the data is moved to the printer
you can turn the PC off it you like.



Why it's not that simple:


https://www.machinedesign.com/3d-printing-cad/article/21122653/top-11-myths-of-cnc-machining


the article states:
Myth #4: G/M Code is a Thing of the Past

That's true too. It amazes me that the industry did not go away from 
primitive code by today standards. G-code was only modified or updated 
by some CNC machines manufacturers as far as I know but most of G-code 
is still the same. Compare that to computer/software advancements since 
1980s. Perl, php, python, Go, html, etc.


Using G-code is like writing computer programs in assembly language! It 
time to upgrade it to something like HP-GL with addition for Z and other 
axis obviously. Such a language would make it much easier for human(e) 
use. 4 to 6 letter long abbreviations for tool manipulation would still 
make code terse enough to fit on smaller LCD displays and we could 
remember the commands for small jobs after a while.


For start, HP-GL commands would need to be modified to accommodate 
relative or absolute CNC tool movement.


Magazine Digital Machinist has some very cool CNC related articles but 
you need to wait long for the next quarterly issue to follow them. None 
of the advertisers mention LCNC ;-(


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Linuxcnc 2.9.0 hangs

[Rafael Skodlar]

On 2020-02-02 17:00, Jon Elson wrote:

On 02/02/2020 05:04 PM, Thomas D. Dean wrote:

On 2020-02-01 20:46, Jon Elson wrote:

On 02/01/2020 08:35 PM, Thomas D. Dean wrote:
I have been using linuxcnc 2.9.0~Pre0 from sources, run in place.  
Until today, it has worked, mostly.  Before today, about 2x per week 
it would freeze and I had to cycle power to get it working.  Now, 
linuxcnc freezes.


Does it ONLY freeze when LinuxCNC is running?  But, just keep working 
if you DON'T start LinuxCNC?
In that case, the most likely problem is the BASE_THREAD has too 
short a period.  Try increasing

the BASE_THREAD period in the .ini file and see if that fixes it.



I changed the BASE_THREAD period by a factor of 10x then 100x, with 
the same result.


The OS seems to run fine.  I did:
> cd RTAI/linuxcnc
> git pull
> cd src
> make clean
> make
> sudo make setuid

All this went OK.  Elapsed time 20 min, or so.

OK, so ONLY your GUI is freezing?  One other thing is if you load a very 
long and complex toolpath, then the

3D preview window can lock up for a long time.

Jon
Perfect example of poor design where the OS with RT kernel is configured 
with all the software needed for GUI and multimedia!


It took me 20 minutes to install linuxcnc-2.7-wheezy.iso in virtual 
machine. It's not the same as on bare metal of course but most things 
should work under such circumstances.
Linux linuxcnc 3.4-9-rtai-686-pae #1 SMP PREEMPT Debian 3.4.55-4linuxcnc 
i686 GNU/Linux


rs@linuxcnc:~$ lsmod | wc -l
83
one a set of modules is sound related
snd_intel8x0
snd_ac97_codec,snd_intel8x0,snd_timer,snd_pcm,snd_seq,snd_seq_device
Default installation has sound modules active on CNC?

LinuxCNC program started fine. I was able to power it on.
When I started milling test axis.ngc it failed due to RT error. Not a 
surprise for a VM. I was watching a TV program on my workstation 
connected to Silicon Dust box over the network at the same time as VM 
was running. As soon as I turned the TV app VLC off I was able to start 
virtual CNC milling. Amazing how many things can be done on one system.


I poked around CNC kern.log and other logs. It's easy to see that there 
is a lot of useful information for troubleshooting.


--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Open source CNC architecture

[Rafael Skodlar]

p of fine men. This list is not necessarily complete. If anyone
would like to add to the list then I encourage the addition.

I appreciate each and every one on this list.



Strongly agree.


I like the latest discussions and hope a way forward is found to continue
the development and proliferation of LinuxCNC.

Regards
Stuart


That's my desire to see also. We do whatever we can in our capacities. 
Some of us support open source with providing feedback, free or low cost 
support or buy early related products for our own use. CDs and DVDs with 
free Linux distributions, laptop with Linux on it in my case. I kept 
most of them as evidence to claims on my resume ;-)


If there was an easy way to help somebody remotely on the Linux(CNC) 
side I would gladly help. In person I can only help in Silicon Valley 
but here we don't have many machine shops anymore. I doubt we can count 
those with LinuxCNC with fingers on one hand.


Best wishes,

--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Open source CNC architecture

[Rafael Skodlar]

On 2020-01-27 05:41, Eric Keller wrote:

On Sun, Jan 26, 2020 at 11:52 PM Rafael Skodlar  wrote:


LinuxCNC is not only about RT kernel as some threads seem to spin around
to no end. It seems that Chris understood my original post the best.
What he pointed out is use of modern technologies that makes it possible
to create advanced systems we could only dream about years ago.

It's clear you want something other than linuxcnc and also seem to be

unable to contribute anything other than ideas.  Ideas are great, but in
isolation they are not so great.  Reminds me of Elizabeth Holmes saying she
invented a time machine when she was 9.  What she did was draw a picture of
a "time machine."  Not quite the same thing.


People spend time to come up with ideas. Some ideas are free, other are 
paid for by taxpayers or people who find them usable. We have patent 
system to protect ideas that can be put in practice. Don't isolate 
ideas, exploit them and give due credit to their inventors.


I'm not inventing here, I just want to use advanced technologies ...


Refactoring linuxcnc is a big job and the resulting product would take many
years to work as well as linuxcnc.  If it were a simple matter, someone
would have done it.  That's partially where machinekit came from, but it
seems to have fizzled a little.



That's clear to me. Every software development company in my career went 
through phases of initial idea, first lines of code, debugging, advanced 
coding techniques, trends in the industry or technologies, etc. Input 
from end users is always important at any stage.


It's not my intention to throw away what's good and working. Some parts 
might need recoding to be more flexible and work with new functionality.
Replace primitive parts with new formats which can be verified easily 
with libraries that were created for manipulation or verification of such.


For example, replace silly ini files with yaml or json in configuration 
or even in G-code files. That way it's easier to detect anomalies or 
mistakes.



At its core, linuxcnc is a machine controller.  It is very flexible, but


LinuxCNC is software for the machine controller.


what it does best is control one machine.  It's a very useful piece of
software.  As long as there are computers with i/o that works and the
mechanisms that make linuxcnc possible exist in the operating system, it's
modern enough.


We know that don't we? How about this; first guy invented a hammer using 
a stone and a wood stick. First open source tool, no royalties. 
Thousands of years later another guy invented metal hammer and I bet 
there was an argument that it's too hard to dig the dirt and extract 
metal when you could simply use stone for the same result.


We either stick with PC architecture for years to come or we open 
magazines like Electronic Design, Machine Design and such and see what 
technologies are advertised or in demand. Taking costs into 
consideration we select technologies and design systems with them. 
Software is the glue.


Flexible LinuxCNC could cover machines for use by artists, hobbyists, 
experimenters, students, and manufacturers. Start small with Raspberry 
Pi like SBC, move to larger and more reliable products based on ETX, 
PC/104, m.2, SSD, etc. when your needs grow.
Among others https://acces.io seem to have wide variety of building 
blocks for modern or legacy needs.


The smallest CNC system could run without GUI for the most part. You 
could manage it with simple keypad and few lines of LCD display or a 
terminal in text mode. Oh, that's what's done in laser and 3D printers 
for years.


For more advanced use you attach GUI over the network or other fast 
protocol/cabling such as HDMI or USB. Linux CUPS comes to mind. You can 
check the printer status from more than one CUPS server.


I just found out this excellent option for advanced machines use; CNC 
metrics https://www.machinemetrics.com/production-monitoring

That site has a lot to learn from. At least for me.



  It's not like a big change like this would attract a number of developers
committed to a multi-year uncompensated work with minimal payoff.
Eric Keller
Boalsburg, Pennsylvania


Well, LinuxCNC was developed after all. Crowd sourcing could solve 
motivation/costs for SW development.


--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Open source CNC architecture

[Rafael Skodlar]

Never expected that saddle on a cow comment would trigger so many 
patties ;-)


LinuxCNC is not only about RT kernel as some threads seem to spin around 
to no end. It seems that Chris understood my original post the best. 
What he pointed out is use of modern technologies that makes it possible 
to create advanced systems we could only dream about years ago.


Those of you that have a problem with that please do some research on 
industry trends. Then stick heads together and see how much of that 
could be built for DIY crowd that fall in the following categories:

- hobby,
- light industrial use,
- advanced industrial use

GNU Linux and other open source software made it through those stages in 
early years. In my opinion CNC systems need to be looked at as robotic 
devices instead of exotic things that cannot be improved beyond what 
LinuxCNC and alike (?) are doing at this time.


This page describes pretty much what Chris was saying I believe:
https://www.mmsonline.com/articles/modern-cnc-control-systems-for-high-speed-machining

"The CNC consists of the following main components:

Operator panel based on industry-compatible PC technology for the 
man/machine interface (MMI)

CNC control unit (NCU)
Programmable controller (PLC)
Drive modules for machine tool axes and spindles (feed drives VSA, 
and main spindle drive HSA)

Motors (AC motors and linear drives)
Supply and energy recovery unit (S/E unit)

As each drive module, the NCU, the PLC and the operator's panel each 
contain a processor, a modern CNC can also be seen as a multiprocessor 
system."


That makes it clear that it's CRAZY to use PC motherboard for running 
all functions of a CNC system. I see some on this list keep mentioning 
PPort; please get off the dead horse or jump the ship!


If I could, I would tell this to Ken Olson, DEC CEO, in the early 90's 
when DEC started to deteriorate in the 90s. To fill the gap, there was 
at least one manufacturer making PDP compatible boards that you could 
plug into PC-AT if I remember correctly. They might be still around.


Architectures like PDP-8, PDP-11, and HP-1000 that I supported long ago 
were designed for industrial use: power plants, electric power 
distribution, steel mills, etc. Why is that important to bring up here? 
Computer architecture. PC sucks! It's based on the 80's 8086 and beyond 
that. It was not designed for industrial use in the first place. At 
best, PCs were used as dumb terminals to mainframe size computers.


PDP CPU, instruction set, memory, interrupts architecture, and Unibus is 
where the magic was. Interfaces/controllers were able to interrupt main 
program or OS based on their priority. Critical interfaces in the Unibus 
backplane had higher priority; disk drive controllers or DIO over 
printer or terminal, etc.


What about LinuxCNC? You want to poll parallel port in a loop and watch 
it in fake scope? Count motor steps or it's encoder? Really? Or have 
smart motors or sensors tell exactly the tool position? CNC components 
are getting cheaper so we can afford more and more of them to assemble 
CNC machines or upgrade the existing ones and have LinuxCNC handle all 
of it.


CNC systems also have sections that have higher priority than others. 
Comparing this to supercomputers is just silly. I don't care what any 
individual uses for their work as long as it does not scare the horses.


I admire and understand that those who converted huge CNC machines 5+ 
years ago want to keep them running 'as is' as long as possible. Some 
machines might be used as fully functional museum artifacts and that's 
fine too. More power to them.


My interest in robotics made me come across interesting but expensive 
solutions in that field. Robots loading and unloading material and parts 
are interacting with CNC machines. How would LinuxCNC work in that 
environment? Watch GUI? Not easily. Give me a break. Leave user GUI off 
the main system!


https://www.universal-robots.com/how-tos-and-faqs/how-to/ur-how-tos/remote-control-via-tcpip-16496/
https://www.universal-robots.com/how-tos-and-faqs/how-to/ur-how-tos/real-time-data-exchange-rtde-guide-9/
https://blog.universal-robots.com/how-to-get-robots-to-talk-to-each-other
MODBUS anybody?

No need for GUI all the time:
https://www.zacobria.com/universal_robots_zacobria_remote_control.html
Best (???) CNC would be able to run from the pendant like peripheral, 
X-terminal or some such as John Dammeyer and Chris mentioned in their 
extensive studies on this subject matter.


Thanks to all for extensive research and comments elsewhere. I learned 
allot.


I hope this message does not turn into "nuclear thread explosion" like 
the other one did. I keep some valuable messages for future reference. 
Sorry but top posting, untrimmed messages, and "signatures" that are 20 
times longer than the answer don't fit that category.


Think it this way; trim threads to make them easy to read on pocket size 
devices so we 

Re: [Emc-users] X forwarding with linux-cnc

[Rafael Skodlar]

On 2020-01-21 13:08, Todd Zuercher wrote:

Correction on my cable lengths, I was thinking 30ft which is about 10m not 3 
(transposed the 1 and the 3 in my head.)

Todd Zuercher


To make it through my DIY projects I had to install a conversion app to 
handle fractions, F, and such on my PDA. Sad state of the Union.


... snip

-Original Message-
From: Todd Zuercher
Sent: Tuesday, January 21, 2020 4:04 PM
To: Enhanced Machine Controller (EMC) 
Subject: RE: [Emc-users] X forwarding with linux-cnc

Cable and wiring needs will vary from application to application, but where 
I've had to replace machine cabling I have usually used Igus' Chainflex cables. 
 But I'm generally working with large cnc routers with long cable runs (3m or 
so) and a lot of long moving cable chains where good long-term flexibility is a 
must.

PS if you can buy your encoder cables pre-made it is usually worth the extra 
cost.  Hand crimping (or worse soldering) those tiny high wire count cable ends 
is a pain in the...)

Todd Zuercher

...

I appreciate your pointer to what I was looking for. The cost seems to 
be reasonable. Some encoders I've seen come with connectors that are not 
too hard to connect to cables.


I don't mind soldering though. I've done it thousands of times when 50 
or 100 pairs phone cables needed to be connected in cable racks for the 
phone company for my intern job. Here in the US it's not done that way. 
They must be afraid of burning their fingers ;-)


SMT components are now a challenge for me. My eyes have seen too many 
small things and need magnification for small parts/print ;-)


Thanks to you guys there is hope for me to build something.

--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] X forwarding with linux-cnc

[Rafael Skodlar]

On 2020-01-21 11:31, bari wrote:

On 1/20/20 11:55 PM, Rafael Skodlar wrote:

That's fine. You can experiment with software/HW any way you want. I
was pointing out what makes sense and what not. You can try running
X-windows over PPP connection through 56k modem over the phone line if
you want.

I would rather "experiment" with headless CNC machine located in EU
from  my GUI front-end on Linux workstation in CA. Modern bandwidth
would allow for video from a camera mounted above CNC machine.

- headless LinuxCNC core with USB for keyboard and simple text display
needed for basic OS installation, configuration, or troubleshooting
- GUI fronted for machine operator; platform independent (i386, Arm)
- API for special additions and future development


How will the CNC machine operator at the remote location with the GUI
(many miles/kilometers away from the machine) handle loading the
material and unloading the parts from the machine? How will they handle
issues with chatter, worn or broken tools, etc etc?


That's a matter of $$$. It's possible to handle such tasks already. In 
case of LinuxCNC is not easy to provide remote support. Lets say LCNC 
architecture with headless controller allows a VPN connection to API to 
see or manage the machine remotely or collaborate with operator that's 
not familiar with Linux. Some surgical robots already make this 
possible. I bet that's way more critical RT system than a CNC machine 
which ends up with broken router bit when "oops" happens.



Is this for some sort of fully automated manufacturing plant in the future?


Well, future is here and I get very excited just reading about it:
https://www.universal-robots.com/
CNC machines are robots in a sense.

However, we should not forget how we made it to where we are. I do 
support and participate in efforts to restore old computing equipment in 
computer museum for future generations to admire. It's a lot of fun when 
you meet people who programmed those machines when or before some of us 
were in diapers ;-)




What exactly are the problems that need solutions or that you want to solve?


I would like to see low cost solutions based on LCNC for hobby or small 
business users so that they can do their work more productively and 
perhaps grow to the point to be able to buy more advanced robotic systems.


It's my intention to put together a small CNC to be able to do more than 
what I can do with not very accurate Grizzly lathe/mill combo. I don't 
mind paying some $$ for components to put a simple CNC machine together. 
If those components were built by members of this list we would all 
benefit. Software and HW QA, feedback, and promotion elsewhere.


One possibility for motivating software developers would be to help them 
buy industrial grade embedded computer/electronics set that they could 
use for writing code and quick test. That could cover upgrade for long 
list of existing machines or help you build new machines. This has been 
done before by HW manufacturers that needed drivers for their products.


I see Mesa covers a range of components that could be used that way but 
it's showing it's age and it's only a single source.


Other times such efforts end up as products, open software and hardware; 
example: https://ardupilot.org/


Generic PC motherboards with multi-core CPUs, GBs of RAM, for CNC use 
make no sense anymore.


Aluminum extrusions make it possible to make all kinds of machines that 
don't exist yet. It's perfect material for people with small work space.


I spent a lot of hours researching this subject on the Internet or 
talking to people at different trade shows but they all seem to sell 
full expensive solutions with proprietary software. Nobody showed any 
interest in LCNC.


I found interesting snap cnc-designer which needs some work but it's the 
only CNC related piece of software in Linux containers space that I'm 
aware of.


Following this mailing list for very long I have yet to see a discussion 
about flexible cables for CNC machines; specific types, wire sizes, 
brand names, etc.


Thanks for not kicking me around ;-)

--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] X forwarding with linux-cnc

[Rafael Skodlar]

Hi Ron,

On 2020-01-20 12:10, R C wrote:

Hello Rafael,


First;  As I said, I am setting up a test bed, not even a CNC, just to 
work out how to make some combination of linux-cnc, stepper drivers and 
stepper motors work.


Second; When I get it figured out, and I am sure I will, because this 
list/forum is great in providing solutions, suggestions, pointers and 
positive directions in general etc.  I


  probably won't use X forwarding,  but then again, I might, 
who knows.


That's fine. You can experiment with software/HW any way you want. I was 
pointing out what makes sense and what not. You can try running 
X-windows over PPP connection through 56k modem over the phone line if 
you want.


I would rather "experiment" with headless CNC machine located in EU from 
 my GUI front-end on Linux workstation in CA. Modern bandwidth would 
allow for video from a camera mounted above CNC machine.


- headless LinuxCNC core with USB for keyboard and simple text display 
needed for basic OS installation, configuration, or troubleshooting

- GUI fronted for machine operator; platform independent (i386, Arm)
- API for special additions and future development

third; I have no Idea what a caw is, however on my travels I once heard 
a breed of parrots being referred to as a caw, and indeed, it would not 
make much sense to put a saddle on a bird


     like that. Unless you meant cow, in that case, and because of 


caw was a trap ... cow it is. You passed the test.

losing track of the topic, I would have carved out some time to spend on 
reading/writing English from those decades


  of research if I were you.


thanks for English lesson. Plenty of opportunities for your police work 
on this mailing list. Perhaps you can cleanup "Pandemonium of rabbits" 
and untrimmed emails that are impractical to read on tiny mobile devices 
for years now.




fourth: Your email sounds like a Chevy vs Ford,  Windows vs Mac vs Linux 
rant, no one probably cares.


Failed test. Better comparison: embedded systems vs. CNC on PC; DOS vs. 
RTOS; metric vs. feet/inches;




The fact you, and few others, don't have a solid understanding of linux, 


"The fact"? based on what? My installations of Linux containers, 
clusters, Saltstack, Openstack, ESX,...?


Compiled Linux kernel hundreds of times, installed OS on thousands of 
servers or VMs in the data centers over VPN from 5 to 1000 miles away. 
Add BSD, Solaris, HP-UX, and other Unix to that list. I only used 
X-windows on workstations installs. When you exceed that number I'll 
sign up for you to give me more lessons.


Fact: millions keep dragging feet and inches over 200 years after much 
better system was invented for science and daily use by people born with 
10 fingers to count.


is not an argument. The fact it is open source isn't either. HPC, super 
computers, worth hundreds of millions, run


some version/distro of linux. One of the reasons being that the ways it 
can be configured, tweaked, modified are enormous, better then one 
single OS vendor would ever be able to provide.



"better then"? Thanks for English lesson again.



Also, I am not sure, but most on this list/forum, are probably not even 
professional machinists, there are probably a lot of "CNC enthusiasts" 
around, some know a bunch about




that's precisely why I was hinting at need for growing up from "parallel 
port" days. Perhaps discuss how SATA, PCIe, M.2 technologies and related 
protocols could be taken advantage of. May I remind you that this is 
"Python3 year".


CNC machining, some are more software/programming inclined, some are 
more OS or electronics inclined. That's probably how linux-CNC came of 
the ground, which is pretty cool.


Exactly. We do agree on something. I believe that we reached the point 
where significant improvements can be or should be done to carry 
LinuxCNC forward.





If you look at it, linux-CNC is pretty slick. Granted I had some trouble 
installing it, using it, it is not based on the OS of my choice (but I 
don't mind/care, and


I never had problems with installing it in VM to test software 
functionality. Granted, I'm not using it for CNC at this point. With 
attacks on my comments make it even less desirable now.




no one else, except you, probably really does.). The reason for my CNC 
"troubles" , I am sure, largely have to do with my own shortcomings, not 
exactly knowing


a lot about linux-CNC, or CNC machining in general. It is an open source 
project, which means that it is available to anyone. It might not be the 
perfect solution, BUT, what is?


I have nothing against people trying this or that. Experiments lead to 
new discoveries but not all.


My active participation in Linux install-fests held at Cisco in north 
San Jose is history. Anybody was welcome to see how our personal 
computers with CRT monitors work. We would help visitors setup their 
systems. I remember a gentleman who asked me to come to his home to help 

Re: [Emc-users] X forwarding with linux-cnc

[Rafael Skodlar]

On 2020-01-19 16:50, R C wrote:
well,  I just want to test a setup, it by no means is going to be a 
permanent setup


So X-forwarding would work, and is easy


Ron


Running X-windows on CNC machines makes as much sense as mounting saddle 
on a caw.


Based on my research LinuxCNC is undesirable in production environments 
or as an option in new CNC products. I spent countless hours to find out 
if any CNC manufacturer is recommending or including LinuxCNC with their 
products. Most small to medium size CNC machines in built or kit forms, 
come with Mach or some other thing. When the subject matter comes up I 
don't recommend their product because it's only available on crippled OS.


After decades of proven good records, there is a lot of misunderstanding 
about using Linux in small business environments. I hate to write it, 
but LinuxCNC is not ready for software option with new or DIY CNC 
machines. When I tried to get small business owners or others at trade 
shows interested in LCNC I get questions that are impossible to answer. 
What kind of computer and other electronics HW are needed, who's 
supporting it, how much does it cost, etc.


I can't tell people to come ask questions on this mailing list. 
Discussions more often than not degenerate from the lists main purpose.


Suggestions to find a used PC or a motherboard with parallel port are 
just silly. That's fine for hackers with more spare time than $$$ in 
their pockets but not for serious business owners. During my visit to EU 
I could not find a used PC to demonstrate LinuxCNC around.


LinuxCNC is like Apache and such in the 1990s. No serious commercial use 
and support. I haven't seen any job listing LCNC as one of the 
requirements advertised anywhere.


If things were different, open source community would embrace LCNC 
instead of putting their effort into GRBL for example. My choice would 
be a headless Linux based CNC controller with suitable drivers for 
different size machines. GUI would be running on separate system 
connected over ethernet, USB, or even wireless in some cases. Numerous 
robots work that way.



--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] 2 installs, one network just works, nothing makes the 2nd work.

[Rafael Skodlar]

On 5/2/19 9:30 AM, bari wrote:

On 5/1/19 11:03 PM, Gene Heskett wrote:

What used to be so hard to build and expensive in analog is often
replaced by software.

So it seems, but does it work as well?

As always the answer is "it depends".

How good are your coding skills? How well do you apply the math? How
well is that ADC/DAC really performing? Code doesn't drift much over
time the way electronic materials do in the analog world.

I recall having a discussion with Bob Pease about "audiophiles" and
distortion you can measure and distortion only they can hear.


This reminds me of my early years at National Semiconductor where I 
supported PDP-11, peripherals, Sun Sparc systems, and at one time fixed 
Bobs PC. His office was something I would take a picture of but that was 
not easy in early 90's. Search for the messiest office in Silicon Valley 
and you should find it ;-)


At one point Bob took his PC up to the roof of one building and dropped 
it to the ground. I agree with his view and attitude toward PC 
architecture. We are emulating PCs as virtual machines these days; BIOS, 
ATA, SATA, SCSI, etc. How bad is that? No serious progress since Sun 
(Microsystems) went down.


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Motorman Robot with absolute Yaskawa

[Rafael Skodlar]

On 4/9/19 6:09 AM, Sam Sokolik wrote:

STMBL amps are really quite cool

Me playing with one as a spindle drive..

https://www.youtube.com/watch?v=7LAtx8o48jI



Impressive Sam. Added a brownie point to your video. It would be nice to 
see the block diagram of your setup. What impressed me is the fact that 
you are using a Dell laptop with LinuxCNC. This is exactly what I had in 
mind when I started a discussion about problems with looking for a 
motherboard that will run RT kernel many weeks ago. Precise timing 
should be offloaded to microcontrollers then any PC could run LinuxCNC 
frontend.


I have no problem with somebody making $$$ when they make a line of 
products that can handle different work environments from hobby to large 
production sizes. As long as some of it is open source and agreed on 
communication standards or protocols then competition takes care of the 
rest of it; marketing, options development, advertising, support.


Looking for more details about your setup ;-)

--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Posting order

[Rafael Skodlar]

On 4/8/19 3:32 AM, Erik Christiansen wrote:

On 08.04.19 05:04, Gene Heskett wrote:

On Monday 08 April 2019 02:36:49 Erik Christiansen wrote:

...

And that requires html encoding, also high discouraged on most lists. And
usually skipped by this reader.


Nooo-, that'd be true if it were done on a sent message, but it's
only a display option I use on received messages. I'm as averse as you
to html, read only the plain text portion of a multipart post, and any
html-only post is auto-converted to plain text - then the unquoted text
is given another colour in the xterm, by mutt¹. Plain text does not
irremediably confine us to a monochrome display. (OK, it may be a bit
"fawncy" for some tastes, but makes for quicker discernment of what's
the reply, and a second saved can be spent on lunch.)

Erik

¹ It's just a regex detecting /^> /, then flipping those lines to the
   chosen colour. No html at all, even on the receive side.


Thunderbird does this so nicely.

Since top quoting folks do not trim emails, messages start to grow like 
rabbits. Ratio of noise/message size was about 80% in most mails at 
work. Besides font descriptions there are pixels of gifys and other 
nonsense that take more space than messages in many cases. This started 
with windows PCs connected to the internet.


Since emails are kept for long time to search back for messages, they 
occupy space on disks, mechanical or SSDs. Each one uses a lot of energy 
eventually. I don't know how much of this stuff is "out there" but it's 
likely in megawatts of energy shuffling emails back and forth. 
Conclusion, top posting without trimming is contributing to global 
warming more than normal style.


So there,

--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Home-limit switch measuring

[Rafael Skodlar]

On 3/31/19 8:07 AM, TJoseph Powderly wrote:

hi gene
maybe redesign to drive past the switch
like you can brush a wall mounted light switch by swiping your hand over
the wall surface
rather than punching it :-)
( try mounting so the direction of switch activation is 90 degress to the
joint motion )
tomp



Good idea. Some computers from the 70's and 80's had a lot of mechanical 
parts. I remember seeing a double switch solutions to detect position of 
some kind of a sliding mechanism. Thinking about it, that kind of 
solution was possibly used on head assembly with voice coil in disk drives.


Switch assembly had tiny wheels on spring loaded lever that was pushing 
gently on microswitch. First switch would detect adjustable position 
close to the limit, second switch was the absolute limit. When the 
electronics detects first switch it slows down the motor and the 
mechanism continues to move to the absolute limit switch then stops.


A: No.
Q: Should I include quotations after my reply?

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?

--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] SPI comms for linuxcnc (was Re: Rock64 pre-orders on Banggood.)

[Rafael Skodlar]

We are making progress...

On 3/27/19 11:16 AM, Chris Albertson wrote:

What makes SPI nice is not it's speed.  On the Pi it can go up to 250 MHz,
so Ethernet beats it for speed and Ethernet cables can be 100 meters long.
   SPI wins because it is fast enough for most things and is very simple,
just connect the wires.  But it only works over a short distance.   But it
is SIMPLE, conceptually no  unlike TTL level serial.


Let's see what one of my former employers, an IC manufacturer has to say 
about SPI:

https://www.maximintegrated.com/en/app-notes/index.mvp/id/6208
"The maximum distance of an I2C bus depends on the capacitive loading. 
In typical applications, the length is limited to a few meters in 
standard mode. This is because a system has to be built to accommodate a 
maximum bus capacitance of 400pF to meet rise time requirements listed 
in the I2C bus specification (Rev. 6 – 4 April 2014)."


Still want to use parallel port? ;-)



   Ethernet needs some support hardware, a small transformer at least.


Not necessarily. There is plenty of boxes connected with fiber cable. 
Seen that done that.




Unlike simply using GPIO pins, SPI does not need much CPU time to send data.

But if writing software you should not have to decide of lock in a certain
communications method.   I like whatht ehauthors is ros-serial did, that
w=said "Use any communications method that has these four fuctions: open,
close, read, write.  THat pretty much means anything from a network socket
to an RS232 cable.



All protocols have some kind of handshake. If designed properly then 
dedicated circuits to handle interrupts take care of it. That beats 
anything that I've seen years ago in one over 40 years old CNC machine 
where transistors and other discrete components were in such odd shape I 
wasn't sure what they were in most cases.


You could extend SPI connection with optical drivers and fiber optic 
cables also. 1km distance would not be impossible. Granted, this would 
be theoretical more than a practical solution not sensitive to electric 
noise or atmospheric electric discharge that killed a lot of RS-232 
circuits which I ended up replacing with 60mA loops in my career.


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Rock64 pre-orders on Banggood.

[Rafael Skodlar]

On 3/25/19 8:39 PM, Chris Albertson wrote:

Yes, you have to patch Kernels.  I think this is the #1 weakness of
Linux CNC.   It would be good to eliminate the need for a real-time Kernal.


Patching is not a weakness IMO. This is done all the time. FPGAs for 
example would get loaded with "binary blobs" during bootup to do their 
things as programmed. Another way to handle this is with kernel modules. 
Load modules for whatever a particular SBC needs.



How?   Move the real-time parts to hardware such as the uP, PRU or Mesa
FPGA.Clearly, it can be done because people are controlling machines
with only a uP, PRU or FPGA.


Brilliant IMO! In this scenario, microcontrollers accept commands from 
LinuxCNC for example and handle their tasks in loops or whatever program 
flow until the task is finished then send interrupt back to main 
processor board saying "I'm done".


Subcontrollers could be advanced in some cases, a RaspberryPi or some 
such running it's own Linux for specialized tasks: drive stepper motors, 
read sensors, etc.


Easier said than done I know, but better than do that all in RT kernel 
to fiddle with parallel port and other silly obsolete things.




The interface was designed around a parallel printer port when outbound
processors had poor performance andhad to be on custom PCBs.Today we
have 32-bit STM32 uP on a board.The real-time code can move off the Linux
PC.


I like this idea. Kernel params can be tweaked so that certain processes 
are 'niced' above the others. I remember tweaking kernel parameters long 
ago for mouse, keyboard, etc. response time. Workstations one way, 
servers that were not using keyboards and mouse most of the time 
another. In this case, LinuxCNC would get highest priority over anything 
else in the system for example.


--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Autonomous machines

[Rafael Skodlar]

On 3/25/19 3:51 PM, andy pugh wrote:

On Mon, 25 Mar 2019 at 22:18, Rafael Skodlar  wrote:


PCs were NEVER intended to be used for CNC yet some of you out there use
it for that purpose.


That is the _entire_ raison d'etre of LinuxCNC.



Do you still use fax or pager by any chance ;-)  I kept my pager to show 
it at work to the latest diaper generation that hasn't seen it or know 
what it was used for.



The point of the EMC project was to do machine control on
off-the-shelf PC hardware rather than on specialist hardware.


COTS industry is not consumer industry. I know how it started but it 
does not need to stay that way. If GNU Linux followed only your 
definition of where it should be running then we would not have it on 
ARM architecture, RaspberryPi, etc.


COTS:
https://www.abaco.com/open-architecture-cots
https://en.wikipedia.org/wiki/Commercial_off-the-shelf

Pay attention to Obsolescence on that page. CNC machines, as far as I 
know, are not being replaced as often as mobile phones or TV sets so 
it's critical to keep that in mind when selecting computing hardware to 
run it.


This is how our tax money defines it:
http://adsabs.harvard.edu/abs/2014ESASP.725E..44P
https://www.faa.gov/aircraft/air_cert/design_approvals/air_software/media/03-77_COTS_RTOS.pdf



Maybe that no longer makes sense, if so then LinuxCNC no longer makes sense.



That makes no sense either. It needs to evolve into application that's 
able to run on different platforms. As I mentioned in my original 
message, GNU Linux by itself does. LinuxCNC depends on good folks that 
created it and still maintain it. Younger generation might jump in and 
take it to the new levels that are significantly different from PC 
architecture from 1980's.


I too evolved from computer hardware and peripheral support person to 
systems administration, and DevOps support in virtual environments. Open 
hardware initiatives are also appealing IMO.


--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] ConFusion 360

[Rafael Skodlar]

On 3/25/19 4:01 AM, Les Newell wrote:


Is this a mailing list about LinuxCNC or windows crap?! 


Someone asked why there isn't a Linux version of Fusion and I explained 
some of the technical reasons why.


I run "stock trading app" written in Java on Linux just fine. It's 
obvious the app is  mostly focused on Windows. It's very poorly 
designed based on what I see in html files, config, and logs because 
windows is their development platform but it still runs on Linux. I 
see people complaining about windows crashes on it's chat channel many 
times. Lost trade is a losing trade.


Java has a lot of overhead and is not a practical language for processor 
intensive work such as 3D CAD. There are very good reasons why very few 
mainstream PC applications are written in Java. As far as your stock 
trading app being buggy, that is down to the software developers, not 
the OS.


I did not say write CNC program in Java. It's portable language that 
happens to work on different platforms just like python.




and Mac have consistent libraries. I can take a binary that was 
written for Win2K and run it on Win10. Conversely I can build a 
program in Win10 that will run on Win2K (within certain limits). 
Linux keeps changing and binary compatibility is lousy. If I build an 
application on a current 


Extreme bull!


No. Fact. If you don't believe me download a copy of SheetCam TNG from 
my website. The copies on the site were built a few months ago on Win10. 


See, you did not read what I wrote or tried to say. "... within certain 
limits ..." ahhh?


I said that I do not have a single machine, including VMs, at home 
running crap. My position is based on facts, i.e. experience over years 
as a solo sysadmin where I had to deal with 4 different versions of 
Unix, MAN, and DOS/Windows PCs or listening to others complaining about 
complete lockups after "antivirus updates" and such. Downtime = lost $$$.


They will run just fine on Win2K. Yes, I do test compatibility with 
operating systems this old. Download the very old version of SheetCam 
Standard I keep to support old users (built on XP if I remember 
correctly) and it will still run on Win10.


OK, so if you are a software developer, I respect your trade. I assume 
you are making money. Great. If you ported that program to Linux then 
perhaps you could sell to different crowd and brag about it also.


I'm not in "metalwork business" but do work on hobby projects that 
require me to spin up lathe or turn on a mill. Nothing professional mind 
you. A CNC would help me speedup some projects I have in mind and that's 
why LinuxCNC is appealing to me.


version of Linux it won't run on a distro that is even a few years 
older. An app built on Ubuntu probably won't work properly on Fedora. 
There are tricks to mitigate these issues but they all involve lots 
of work and don't provide a reliable solution.


Unbelievable what speculations I have to read these days. 


Again not speculation. I've invested hundreds of hours working on this 


That's how it sounds. I used numerous programs that were originally 
written on one Linux distribution that was different from my favorite 
ubuntu. Takes some effort and it usually works after a bit of effort.



and testing different solutions. Can you show me any proof that I am wrong?

Les


Many discussions on mailing lists end up nowhere because it's fruitless 
to argue who's right and who's not. I just don't see a point about 
pushing software written for one platform only if it's on unrelated 
mailing list.


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Autonomous machines

[Rafael Skodlar]

On 3/24/19 11:30 PM, Chris Albertson wrote:

This little board is being discussed non-stop in many other forums.   I
intend to buy one but not for Linux CNC

This is way overpriced for what you get if you want to run Linux CNC on if.


This statement is precisely what I have problems with. That's why 
clueless people and MS emphasize all the time, how can free software 
make you any money. I have no problem paying $300 for a computer that 
will run a CNC machine for years to come.


The way you put it can be interpreted that $30 would be too much for you 
to pay.



   You are paying for the 120 CUDA core GPU which I don't think can be used
for graphics.   The GPU is intended for computation, either for vision
processing or machine learning


Computation is computation, like math is a math. Some chips are more 
suitable for certain functions than others. If a board is designed to be 
used in robot systems including motor control, read all kinds of sensors 
including visual spectrum camera, then it's suitable for CNC also. Some 
CNC systems work with an arm like mechanism used for welding for 
example, where capability to detect things similar to human vision would 
be a bonus. Co-robot, co-CNC machines,... meaning they work with humans 
without a danger to injure them.



If what you are looking for is a faster BBB, then the Pine64 is that.  Or


I described my gripes about poor mechanical design on these boards long 
time ago. Cables all over the boards or their sides, inappropriate 
connectors in respect to current, vibration, thermal properties is just 
bad design.



if you are OK with spending $100 on the Jetson then why not spend $100 on
an Intel Celeron based board?


You mean PC motherboard, unsuitable computer design for CNC from the get go?


It's a good product but this is not its intended use.


PCs were NEVER intended to be used for CNC yet some of you out there use 
it for that purpose. I don't blame you. In my research for CNC hardware 
I'm still not clear which motherboard to use for CNC because you need to 
run a special test to ensure proper behavior with RT kernel which cannot 
be done before you buy it.


Jetson-nano board designed for use in robotics and other real time 
systems is sure more suitable than RaspberryPi, Baegle board and such 
that were designed for toys/education, not industrial use.


The board is not even out yet but definitely worth looking into it IMO.


Don't worry about the single source, drivers or any of that.  This is a
standard ARM board that runs Ubuntu Linux and will run just about anything
that can run on Ubuntu Linux but the part that makes this unique is the low
price of the Nvidia GPU.



You are missing the point, overall PCB design with edge connector that 
together with a decent backplane could be used the same way as PDP 
computers from Digital were used in industrial environments. Boards with 
edge connector was in the bottom plugged into Unibus backplane, most 
cables from interfaces were in the back with few on the top, all 
enclosed in well ventilated box.



--
Rafael Skodlar
Some people have vision, some do not, and some simply follow what a self 
imposed elite tells them to do.



___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Autonomous machines

[Rafael Skodlar]

This seems to be a good platform designed for embedded systems:
https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-nano/

Trouble is in the bus not being a popular standard supported by numerous 
manufacturers with interfaces for additional GPDIOs, drivers for stepper 
and other motors, etc. Single manufacturer is also a concern in the long 
run but who knows what the future holds. Price is acceptable for serious 
CNC application IMO.


We know Linux is running on it. Add RT kernel with LinuxCNC and you have 
a CNC platform. CNC is autonomous machine after all.


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] ConFusion 360

[Rafael Skodlar]

Sorry Les,
but ...

On 3/23/19 1:37 PM, Les Newell wrote:

Hi Andy,

(Which is based on BSD, so a Linux port would be easy, if they chose 
to do it)


Porting closed  source code to Linux is very difficult to do. Windows 


Is this a mailing list about LinuxCNC or windows crap?! I run "stock 
trading app" written in Java on Linux just fine. It's obvious the app is 
 mostly focused on Windows. It's very poorly designed based on what I 
see in html files, config, and logs because windows is their development 
platform but it still runs on Linux. I see people complaining about 
windows crashes on it's chat channel many times. Lost trade is a losing 
trade.


and Mac have consistent libraries. I can take a binary that was written 
for Win2K and run it on Win10. Conversely I can build a program in Win10 
that will run on Win2K (within certain limits). Linux keeps changing and 
binary compatibility is lousy. If I build an application on a current 


Extreme bull!

version of Linux it won't run on a distro that is even a few years 
older. An app built on Ubuntu probably won't work properly on Fedora. 
There are tricks to mitigate these issues but they all involve lots of 
work and don't provide a reliable solution.


Les


Unbelievable what speculations I have to read these days. No wonder the 
latest diaper generation is so confused about anything. I see them 
coming out of school and what they think they know.


There's so little progress in computer science and anything else since 
1980's PC. The most notable difference is we virtualize that hardware in 
software and keep throwing more Intel like CPU cores and RAM at it these 
days. I see it every day at work. Just look at an icon on typical app 
for saving your work.


--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] ConFusion 360

[Rafael Skodlar]

On 3/22/19 2:14 PM, Jeff Johnson wrote:

Anyone on here have opinions on Fusion 360 Cad/Cam by Autodesk?



Opinions is one thing, experience is another. After few mentions about 
Fusion on this mailing list in the past I briefly looked at it and that 
was it. Anything from here is my _opinion_ based on a lot of experience.


Any engineer worth his keep ran CAD CAM programs on minicomputers of 
some kind; Digital, HP, Silicon Graphics, Sun Microsystems. All that 
kept healthy competition going until it became bastardized with ports to 
Windows".
Anybody or a company that is so narrow minded to produce their software 
product for one crappy OS is STUPID! They are under false impression 
that "free software" is only free in terms of $$$ so there's no money in 
it. Wrong! We proved this time and time again, free software is free as 
freedom which is not necessarily absolutely free and run if there is any 
grain of decency in for profit user.


Last week of February this year I "celebrated my 25 years of Linux" with 
creating a wiki page on internal site at work. In there I described my 
experience and inserted numerous pictures of Linux distributions I 
bought in the past. First was a set of 49 floppy disks from Linux 
Systems Labs which besides at home, I also installed it at National 
semiconductor as the first Linux server. That was running at least two 
more years I was told after I left the company. Why stay when the 
manager told me there is no future in Linux. She sent me to "Windows 95 
reeducation camp" in San Francisco that year.


My Linux experience helped me find interesting jobs because I kept 
supporting and studying it with buying 6CD set from ImageMagic, Progeny 
Linux, Redhat, Mandrake, etc. and FreeBSD.


I believe that if I use "free software" of some kind commercially I have 
a moral obligation to contribute in some way and I encouraged that at 
different work places. One of my managers approved purchasing GNU 
package for Windows in late 90s.


I would call "greedy bastard" anybody that would run a "(machine shop) 
of some kind making huge profits with free software without providing 
some kind of support to people that developed it. Send them $$$ or 
invite them to your shop for consulting; travel expenses included. That 
would give you bragging rights to advertise "We actively support 
" on your main web page.


A number of us, members of Silicon Valley Linux Users Group, volunteered 
for years installing Linux on personal computers for those who brought 
them to Linux Install-fest organized at Cisco, North San Jose. This was 
a good will to software developers mainly so that they would be 
encouraged to develop software for/on Linux platform.


There is a number of good examples of software projects that started as 
open source and evolved into commercially supported version with 
proprietary enhancements only available for $$$. I have no problem with 
that as long as the data and config file formats remain compatible with 
free version.


This kind of "business model" allows people to learn basics of a 
software product for free then switch to commercial version (because 
they know it) for profit.


My conclusion after reading long thread that grew out from "Fusion 360" 
is ConFusion. Now I wonder about their Eagle acquisition also. "2 
schematic sheets, 2 signal layers, and an 80cm2 board area." PCB with 
80cm2 only? Really?

Sorry, but Autodesk "embrace and extinguish" business model is not for me.

--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] OT: Where's them fires. [Was: Re Conversational mode.

[Rafael Skodlar]

ow to do it.


This OT is way too long. I wish we spend time discussing about new 
minicomputer architecture for CNC use instead of failed political dreams 
and techno scams.


--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] No clean solder flux, how to buy in Sweden

[Rafael Skodlar]

On 2/24/19 1:14 PM, Chris Albertson wrote:

Lead solder is not yet banned in the US.  But it is banned in almost all
  the world.   The US is not the biggest user of solder and few companies
want to make a US-only product so we get what is available on the world
market.But you can still find lead solder. But I've 100% moved on.
The best stuff is silver based now.  Use that.


Not always. Old equipment needs to be worked on with older kind of 
solder. If nothing else it is to preserve the artifacts in museum(s). 
Silver is harder to work with because of the temperature difference so 
PCB would get damaged before the solder joint would be complete.



The reason they have banned lead is that EVERY electronic device eventually
ends up in a land fill.  The lead then leaches into the water table.

Not exactly true!!!


Sounds silly that tiny phone might contaminate an entire aquifer but Apple
sells something like five million phones every month and now that the
market is started this means 5 million phones go into landfills every


that is unlikely. People do recycle stuff these days. Manufacturers are 
taking back old equipment for recovering precious and other metals that 
can be used in manufacturing again.


We have scrap metal and electronics collection in place here in Silicon 
Valley. Every few months recycling specialists pickup anything of that 
sort left in our driveways with attached note. They don't dump it in 
landfills.



month.   Or another way to think of this is that EVERY pound of lead solder
manufactured eventually goes to a land fill.  Lead is mostly not to
harmful to older adults but has a serious effect on babies and children and
the effects of exposure can't be removed, the effects are permanent.

They passed a law years ago in China that ALL phones must use USB chargers
with a standardized USB-A connector.   Notice that the entire world is now


Any such laws come from CA in US first as far as I know. China keeps 
polluting their sea and that can be traced across Pacific to US West 
coast. Watch nature programs that show pollution affecting animal world 
on the way.



following China's charger regulations.   The effect is millions of tons
fewer used chargers in landfills.   Now that chargers are standardized
people keep them.  China got that one right.


China is the last one to be given credit for cleaning up the environment 
based on what I see on Asian TV programs alone. Japan, S. Korea care 
about the environment way more than red PRC.



About freight for stuff on eBay.  Most of what I buy there is with free
shipping.The free shipping is mostly from items shipped from China as
postage is very low there.


That postage is subsidized by USPS based on some old obscure law that 
was put in place to help undeveloped countries I heard.


--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] DIY CNC builder dilemma, open request for comments

[Rafael Skodlar]

On 2/23/19 7:29 PM, Dave Cole wrote:

On 2/20/2019 8:42 AM, Ken Strauss wrote:



-Original Message-
From: John Dammeyer [mailto:jo...@autoartisans.com]
Sent: Wednesday, February 20, 2019 12:07 AM
To: 'Enhanced Machine Controller (EMC)'
Subject: Re: [Emc-users] DIY CNC builder dilemma, open request for

comments

My mill is equivalent to a Grizzly G3606.  From the now defunct House of
Tools.  They handle the broken parts and assemble and make sure 
everything

is there before I ever saw it.  No real complaints but it's not a HAAS.

Of

course the price was nowhere near that either.

I know for sure that I'm going to have to either rebuild (redesign) 
the X
axis nut or just install ball screws (preferred).  The PC, the CNC 
control

software are such a small part of the whole thing.

The BeagleBone design is published and can be used in other layouts.
Nothing stops someone from producing a more S100 if you will (but 
smaller

cards) system using it as a backbone processor.  Although it's video
processing is pretty pitiful.  The Pi is not public.

But again, 1000 hours  (about 6 months work).  $100 to $150 per 
hour.  Say
with all the design work including metalwork etc. that you can create 
that
magic Linux CNC based box for $150,000.   The customer base will 
probably

only want to pay $500 at the most for it since you can duplicate what it
does with used PCs and some hardware.  The motors, couplers, power

supplies

etc. remain constant regardless of the install.   So if you want to make
back your investment and earn a living then $150,000 R / $500 per unit

is

300 units.

And even with $500 per unit the end user still has to modify his machine
which is where all the work and money is.   If the need was there it 
would

already be filled.    IMHO.

John

Selling 300 units at $500 each would only recover the $150K R if the
supplied hardware were free with no advertising/support required for the
sales. I suspect that needing to sell 3000 units is closer to reality.





The PC is not going away anytime soon.    I do a lot of industrial 


Some form of PC will stay around for time to come. However, industrial 
computer is a different beast for number of reasons. Tablets and 
Chromebooks are changing this the most in user landscape.


control work and more and more PCs are being installed in factories to 
allow access to the MES system, quality control/part tracking system 
(IBS QMS,etc) and provide employee log in/log out.   Act as cell 
controllers for machine cells.  They are everywhere.  Most of them are 
small fanless units with one or more Ethernet  ports.


We need specialized SBCs for industrial and CNC systems obviously. It 
would be nice to have a handful of well known brands to select from for 
DIY projects. They could be functionally the same but use different CPU. 
I would participate in related Kickstart assuming the design was good 
and supported for at least 5 years ;-)


The factories are using off the shelf desktop screens since they are 
dirt cheap.   Standard keyboards etc.  They break, they replace them. 


Dirt cheap are the key words here. Non-technical bean counters select 
the technology too often because they don't know better. I've seen it 
too many times in industrial and data center environments and small 
startups. Computer technology keeps revolving around simplistic PC 
architecture even in virtual systems these days. I see users requesting 
virtual systems with 4 coreCPUs, 8-16GB RAM, 300-500GB of virtual disk 
space just to run simple compiler jobs and such on a daily basis.


I've even seen standard desktop PCs used in steel mills out in the 
shop.  Its popular to mount the PC behind the screen.


The only reason why PC installation may slow in factories is increased 
use of robots which eliminates the employees who would interact with the 
PCs.


CNC machines are robots too. They use different language to run and 
interact with humans in some steps. CNC machines are a class of co-robots.




Remember when the parallel port was going away 10+ years ago, 
wellyou can still buy NEW PC motherboards with parallel ports.


Yes, and some printers come with Pport to plug in to those obsolete 
ports also. It's like bulky DC power connector designed for use by 
stupid smokers in cars decades ago.


Thanks to all for participating in DIY CNC builder dilemma. I learn so 
much from such discussions. Too bad we can't get together and solve CNC 
automation problems as a company selling an ultimate CNC product.


--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] head scratcher

[Rafael Skodlar]

 keyboard, video connection, Linux in USB memory or CD 
(if you have such a drive) is enough to start. Unplug all other 
interfaces to eliminate their possible interference in troubleshooting 
steps.


BIOS generates beeps or diodes indicate hardware errors in some cases. 
Check MOBO manual.


Good luck,

--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] head scratcher

[Rafael Skodlar]

Hi Gene,

On 2/20/19 10:50 AM, Gene Heskett wrote:

Thought maybe I'd see if my memory (in the machine, I know /mine/ is
shot)
was getting finicky. And the uptimes on this D525MW board haven't been
that great recently, but I reboot and choose memtest86 from the grub
menu, and it advises the mem is wrong, and reverts to the grub menu in
about 5 seconds.



One way to troubleshoot is to swap DIMs or whatever type of memory you 
have and you should see if the memory error moves. Older motherboards 
might have electrolytic cap problems. One time I bought a bag of them 
and replace all on one motherboard.


Series of caps made in China was as fake as CNN news. They stole part of 
manufacturing process from Japan.



Run synaptic, see there is a slightly newer version available under a
slightly different name, but installed, it obviously needs someone 100%
familiar with the map available just to run it.

The msg at the top of the screen is wrong address, 0x99100, 0x8F000

And it reverts to the grub menu in about 5 seconds.


I would remove HDD and test booting from USB memory stick. Eliminate or 
narrow down components needed to see anthing on the screen to find the 
problem.




Anybody have a clue whats going on?
  
Thanks all;


Cheers, Gene Heskett



Good luck,

--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] DIY CNC builder dilemma, open request for comments

[Rafael Skodlar]

On 2/19/19 5:28 PM, John Dammeyer wrote:

I did find the actual supplier web site.
http://newker-cnc.com/index.php?l=3

Having some trouble downloading the English manual but these are the units
sold through Alibaba.


There you go, classic problem with made in china.


I'm not in any way saying these are better than LinuxCNC on a PC or
MachineKit on a Beaglebone.  They do claim to be using ARMs.  The point is


I am not against modern CPU architecture. It's the sandwich design that 
makes no sense to me. And lack of simple connection with easy 
interchangeable interfaces for DIO and stepper motors drivers. Capes, 
hats, and underware are not professional solutions. It's toys on top of 
toys for school projects. Might as well put that in Lego plastics.


If Beaglebone came with a decent PCB design so that companies could make 
professional interfaces for it I would be all over that architecture. I 
don't see it anywhere. Trouble is that modern designers don't bother to 
see what people created with much more limited resources in the 70's and 
80's. S-100 bus would be better than a "sandwich with header connectors".



for the money they are asking it's just another example of a complete
package, ready to go, that probably does some level of machining well enough
to make parts.  And fundamentally it's all about making parts as cheaply as
possible.


You get complete packages here in the US also. The cost is not adjusted 
to DIY and experimenters. Kickstart with LinuxCNC based on open 
architecture would be good start ;-)




Unless I win a lottery I can't see buying one just to play with.  And I
already have my $25 PCs I bought to run LinuxCNC or MACH3.  But by the time


Lottery would make it possible to buy good US made product with manual 
you can read and mobile phone to make a call to person that understands 
what your problem is.



all that stuff is assembled and mounted the costs really do exceed the price
of the far east units.  Just like anyone building a Knee Mill in North
America wouldn't be able to compete with the same size units made in the far


Industrial revolution started mostly in USA so there is no reason it 
could not be done again. Just depends on how hungry people are.



east.  That's why the Grizzly, Tormach etc mills are all from the far east.

John


My Grizzly machine was delivered upside down, broken parts inside, 
bearing for lead screw seized first hour of use resulting in broken cast 
iron gears. When you look at steel and finish quality you see that's 
worth less than what you paid for.


--
Rafael


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Dedicated Processor hardware for LinuxCNC

[Rafael Skodlar]

On 2/18/19 5:00 PM, Greg Bentzinger via Emc-users wrote:

( Greg )
Every kid would love to ask for a pony, if they thought they could get it they 
would ask for a unicorn. Reality however usually proves this to be impractical 
to impossible.
Everything is changing today at a hyper accelerated pace. Back in the day Motorola produced the 68000 CPU, it was used in all sorts of equipment from 

^^ not true for everything.

Apple PC's to Okuma machine controls. I would guess that that CPU was 
still being used to manufacture products 20 years after debut.

Today PC cpu and chip sets are what a 3-5 year life cycle before going OOP. 
Same with many ARM CPU. Look at the Arduino and how many generations have come 
down the road. Most new version have moved to 3.3V and are no longer compatible 
with 5V shields.
Even if someone did come up with a backplane card rack for most of the interface cards, 
that main CPU board would have to be updated nearly constantly because key components 
would become un-obtainable. Todays global manufacturing tries to run as close to true JIT 
as it can.(JIT="Just in Time") companies want as little space and capitol tied 
up in inventory. When I worked an Apple PC assembly line they were so JIT focused they 
only kept 4-6 HOURS worth of components in the facility. This meant that trucks were 
delivering Mem, HDD's ect. were being unloaded every few hours and being reloaded with 
boxed product ready for market.


Good to know. I love computer history.


Several Chinese machine tool builders have offered LinuxCNC as the control to 
reduce the overhead of building the machine. I believe it would be used more if 
tool builders were sure they could not be held liable for a system they sold 
that had been modified by the buyer.
GRBL has come along way - and it was a project to fit a stripped version of the 
early EMC/LinuxCNC into cheap Arduino hardware. the current v1.1 IIRC had to 
strip down the boot loader and some other items to still be able to squeeze the 
optimized assembly code into the Atmega328p chip. What did GRBL have to give up 
to fit in an Arduino? Tool table capability {G43}, Tool radius comp {G41-G42}, 
Minimal look ahead buffer, No program storage - its all drip feed via serial, 
no program editing, X-Y-Z only no additional axis, no spindle feedback 
(tapping), requires second device to stream G-code and operate the control. Now 
I like GRBL and it has the honor of being the founding code which virtually all 
extruder type 3D printers is based. I hope someday there is a port for a rotary 
axis, I would love to use it to engrave on cylinders using X-A-Z.
For now though, the movement towards SSerial and interface from control PC via 
Ethernet allows all sorts of flexibility.
As for HP-GL, I had to work with HPGL for tool paths for several years and it was a complete disaster. Mathematically a "line" has no physical width, not exactly so in HPGL so you will have gaps, broken chains, lines intersecting not at there endpoints and all arcs are output as splines. While at the scale for a HP pen plotter this worked out, for CNC use it was a dismal failure of epic man hours wasted trying to get each file into a usable state. I hope HPGL 


HPGL was not designed for CNC work if we assume that plotter is not a 
CNC machine. However, the reason I brought it up was in it's simplicity 
of writing raw code for it be that in Basic in HP-85 for example, or 
whatever else. It's easier to remember PA (Plot Absolute), PR (Plot 
Relative), PT (Pen Thickness), SC (Scale), etc. than G-code.

http://paulbourke.net/dataformats/hpgl
For 3D somebody would need to come up with language extensions.

is banished from the face of the earth long before G-code begins to 
fade. I agree that G-code is far from perfect, but there is no other 
method out there that even comes close. Early AutoCAD had the same sort 
of dysfunction issues by the use of the "polyline" construct.

Lastly - May I ask the status of LCNC v2.8? Is there any potential release date 
on the horizon?
Thanks ( /Greg )


G-code is like Latin used by doctors 50 years ago. How is that working out?
Just because your implementation did not work for you HP-GL is not bad. 
Repeatability and accuracy on HP plotters was very good even for pens in 
my experience. With a bit of adjustment for a CNC environment HP-GL 
would be at least as good ad G-code in 2 dimensional space. Code is just 
a code. One is easier to read than the other but it has nothing to do 
with machine accuracy.


It was only used by engineers not a crowd that only knows how to insert 
paper into inkjet printing machine. I've seen some very impressive IC 
layout designs plotted with HP-GL on HP plotters that I maintained or 
repaired in the past.


--
Rafael
Exclusive Linux user since Feb 1994


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] DIY CNC builder dilemma, open request for comments

[Rafael Skodlar]

r entry level CNC. 
That alone should make it clear that it's silly to search for suitable 
PC motherboards that can run RT linux kernel when Arduino can handle 
much of it. Why did GRBL designers go that route? My guess is that 
LinuxCNC was too complicated from their point of view.


My choice for CNC control would be a system with the following major 
components:
* backplane with power supply and connectors for accepting CNC computer 
board and interfaces, all using edge connectors.


* headless computer board running LinuxCNC (non-gui) with RT kernel and 
USB port to get the job related commands from the front end GUI part of 
LinuxCNC, or independent APIs, and execute them in real time. Small CNC 
machines could work with Odroid, BeagleBoard, or RaspberryPi like 
system, bigger ones on more powerful would use 64bit CPU systems.


* two types of servo boards would be plugged into backplane directly. 
First kind would include motor drivers for use in smaller CNC systems. 
Another kind would provide signals for more powerful external drivers 
with large heatsinks like Gecko etc.


* DIO interfaces for connections to different sensors such as magnetic 
or optical encoders, limit switches, keypads, etc. Manufacturers could 
make different kinds of IO interfaces that would be able to interrupt 
main CNC computer in real time.


* front end computer for user interaction with CNC system and outside 
world. A PC with (ubuntu) LinuxCNC gui or a "tablet" with Android for 
example. It would be connected to the CNC system over USB3 to control 
jobs; not motors, relays, or other components in detail. Hint: ROS for 
managing robots. Bluetooth could also be an option. This would allow for 
one person to walk around and manage a set of CNC systems quickly.


Sorry it's so long.

Best wishes to all in 2019,

--
Rafael Skodlar


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] If there's one thing I hate, all the noise, noise, noise, noise

[Rafael Skodlar]

Tom,

Hard to tell from the picture but I would make sure that ground is 
properly connected in star configuration, i.e. there is only one 
ground point [1] in the cabinet and the CNC machine. I suspect you get 
some ground currents somewhere which possibly cause enough noise to mess 
up your VFD.

[1] ground could be a copper bar with enough holes to attach all ground 
wires from all cables, CNC chassis, and power filter to it. Connect 
ground wire to the door itself and whatever is mounted on it as well.

It might be worthwhile to add ferrite toroids on critical (analog) lines 
including parallel cable; see 
https://en.wikipedia.org/wiki/Toroidal_inductors_and_transformers

On 05/11/2015 06:00 PM, Tom Easterday wrote:
 Early on we made a decision to put our VFD into the same cabinet as the rest 
 of the electronics on a lathe retrofit - I originally had two separate 
 cabinets, one for power with vfd and one for electronics.  This was to save 
 space and bring the size of the cabinet on the machine down to a more 
 reasonable size so it might fit through doors, etc.  In retrospect, perhaps 
 that was a bad idea.  But here we are trying to address VFD induced noise 
 problems.   The cabinet is shown in a picture here:
 https://www.flickr.com/photos/37438950@N00/17219261571/in/album-72157651167328249/
  
 https://www.flickr.com/photos/37438950@N00/17219261571/in/album-72157651167328249/
   There is actually more wiring completed now then what is shown in the 
 picture but you get the idea.

 After having read about folks often using Rasmi input power filters to solve 
 noise issues I purchased one from ebay and installed it today.  It didn’t 
 help, and may have actually made the problem worse.  I installed it very 
 close to the VFD input power terminals as recommended.

 When I run the spindle motor on it’s lowest RPM I hear a high pitched whine 
 (at the motor) and strange things begin to happen in Axis.  Windows pop up, 
 perhaps a homing window, perhaps a touch off, perhaps Axis switches to MDI 
 mode, sometimes it turns the machine off, sometimes it turns the machine off 
 but the spindle keeps moving!   Sometimes the VFD shuts off and displays oL 
 1” on the screen.

 So now I want to understand how this noise is getting into the PC. I first 
 thought it was because the keyboard, mouse, and video cables ran past the VFD 
 in the cabinet and noise was being induced on the keyboard cable.  So in 
 trying to isolate where the issue was I disconnected those cables and ran 
 them far away from the VFD.  No help..  I then wondered if it was coming in 
 the AC power to the PC, so I rerouted the PC power to a completely different 
 outlet outside of the cabinet.  No help.  I then rerouted the network and 
 video cables to get those away from the VFD, no help again.   Even with the 
 door open (as you see in the picture) I have noise.  The only thing 
 connecting the PC to the rest of the system is the parallel cable which is 
 about 12” long that connects to the Mesa 7i85s card (and again, that is at 
 the other end of my cabinet from the VFD).  Today I borrowed a friend’s 
 0-1Ghz spectrum analyzer to see if I can find the source/frequency of the 
 noise and/or where it migh
t be getting to the PC.  I will start playing with that tomorrow.

 I am wondering if anyone has any ideas of where i can look, or what I can do?
 -Tom

-- 
Rafael

--
One dashboard for servers and applications across Physical-Virtual-Cloud 
Widest out-of-the-box monitoring support with 50+ applications
Performance metrics, stats and reports that give you Actionable Insights
Deep dive visibility with transaction tracing using APM Insight.
http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Network Broadcast?

[Rafael Skodlar]

On 02/05/2015 10:18 AM, andy pugh wrote:
 On 5 February 2015 at 18:10, Kirk Wallace kwall...@wallacecompany.com wrote:

 Maybe?
 http://www.computerhope.com/issues/ch000783.htm

 it seems to be easy Mac-to-Mac, PC-to-PC or Linux to Linux..

 Less so in a heterogenous network.

That's not so. There is plenty of ways to communicate between the 
systems on LAN or beyond including with less desirable OSes.

Python has been ported to all OSes worth mentioning. It should not be 
too hard to come up with a script to pass messages and present them one 
way or the other.

One way to push a message is presented here:
http://www.html5hacks.com/blog/2013/04/21/push-notifications-to-the-browser-with-server-sent-events/

You can always use netcat or nc to pass messages and pipe them into 
scripts etc.

If you have wireless router on your LAN you can use your PDA (they call 
them smart phones for some strange reason these days) to alert you on 
desired events.

I would not use email as it is indeterministic event.

-- 
Rafael

--
Dive into the World of Parallel Programming. The Go Parallel Website,
sponsored by Intel and developed in partnership with Slashdot Media, is your
hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials and more. Take a
look and join the conversation now. http://goparallel.sourceforge.net/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Printing on our version of wheezy

[Rafael Skodlar]

On 01/09/2015 08:54 AM, John Thornton wrote:
 I tried to print to a network printer from sneezy and never got that to
 work. Something about tried 3 times and gave up... odd that it could not
 see the network printer either when I tried to configure it.

 JT


Printer daemon needs to listen on port 631 and firewall has to allow 
connections on that port.

ss -lnt
command returns the following line when cups is listening:

LISTEN 0   128   *:631   *:*

see more bellow

 On 1/9/2015 9:41 AM, Gene Heskett wrote:
 Greetings;

 I've been screwing around, trying to make my printers work, and while I
 can wake them up, no paper comes out.

 The cups error_log:
 HL3170CDW: error while loading shared libraries: libpoppler.so.5: cannot
 open shared object file: No such file or directory

 Seems to be the show stopper.

 I've no clue where to get this dependency because the previous ubu
 install did not use it. Googling around, it isn't the latest version
 either.

 That error above was generated AFTER I had downloaded the most recent
 debs for that printer from support.brother.com, installed them with dpkg,
 and reconfigured it in a root session of firefox, restarted cups, and
 attempted to print the cups test page.

 So, since there does not appear to be a libpopplar.so.5 existing on
 either the wheezy drive, nor in the old ubu 10.04.4 LTS drive, where can
 I get it?

 Thanks guys.

 Cheers, Gene Heskett

Gene, have you tried apt-get search libpopplar? It returns a number of 
related package names on Kubuntu 14.04:
libpoppler-cpp-dev - PDF rendering library -- development files (CPP 
interface)
libpoppler-cpp0 - PDF rendering library (CPP shared library)
libpoppler-dev - PDF rendering library -- development files
...
among them. However, I have none of that installed on my system with 
working CUPS. I doubt you need it.

apt-get search cups
returns a number of packages, some of which might be needed in addition 
to what you already have to get your printer going.

Have you tried CUPS setup with web browser http://localhost:631 to 
manage the printer?

Try to use simple postscript driver if your printer model is not listed. 
Worked for me many times.


-- 
Rafael

--
Dive into the World of Parallel Programming! The Go Parallel Website,
sponsored by Intel and developed in partnership with Slashdot Media, is your
hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials and more. Take a
look and join the conversation now. http://goparallel.sourceforge.net
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] new motherboard doa, twice

[Rafael Skodlar]

following this for a while ... but

On 12/15/2014 02:43 PM, Matt Tucci wrote:
 3 beeps, no ram installed, won't be any video because it still hasn't
 booted to bios. Incompatible ram seems like it could be your issue, but 2

that is not correct. PCs do not boot into BIOS. BIOS is the code that 
runs self tests and initiate bootup from the selected peripheral. Some 
advanced BIOS boot mini version of Linux but that's not common 
unfortunately. Blame Microsoft.

It's the BIOS code that makes beeps when self test detects an error. 
Beeps are somewhat standard. http://www.biosflash.com/e/bios-beeps.htm

 bad motherboards out of the box? I have never had one out of the box bad in
 25 years. Anytime something was screwed up, it was something I did. Also
 Keep in mind that some motherboards want a certain wattage Power Supply. My

That statement is inaccurate. All motherboards require certain wattage 
whatever that be. It all depends on the chipset, type of memory and CPU. 
What's also important is peripherals such as disk drives, internal tape 
drives, video cards, etc.

 Daughter built a computer last year, i gave her a P.S. (450 watts), I had
 and it wouldn't boot, it was 50 watts less than the MB recommended, went to
 microcenter got a 700 watt and it has been fine. Good Luck.

That's suspicious to me. My system with 4 core CPU, 12 GB of memory, and 
Nvidia video card is happy with 450W PSU. Unless that PC has more stuff 
in it, 450W should be sufficient.

I work with 1U or 2U servers here and there and they do not use that 
much power, over 450W, with 2 multicore CPUs.

http://extreme.outervision.com/PSUEngine


 On Sun, Dec 14, 2014 at 3:11 PM, kqt4a...@gmail.com wrote:

 On Sun, 14 Dec 2014, jrmitchellj . wrote:

 In the past, some of the motherboards I have bought came with the CMOS
 jumper in the reset position.  That holds the system settings in a
 cleared
 and won't allow the system to start.  You may want to check for  that.

 Ray

 --J. Ray Mitchell Jr.
 jrmitche...@gmail.com
 (818)324-7573

Mixing old memory with new motherboard is most likely the issue here. 
Make sure you have memory type that manual recommends and your PC will 
help you cut material soon.

-- 
Rafael

--
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration  more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=164703151iu=/4140/ostg.clktrk
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Sign in issues

[Rafael Skodlar]

Allen,

Still struggling with your login?

On 11/21/2014 06:56 PM, Allen wrote:
 I hope this finds someone that can help.

 I have my user name and password written down, the system doesn’t like
 them anymore!

you must have written it incorrectly. Make sure your Caps Lock is not on 
or numerical part of keyboard is not in edit mode.

 I have tried to get a new password and asked for my user name as suggested.
 I get no emails with either user name or password.

Where is that suggestion coming from? This is not on some server or in 
the cloud where applications make it possible to recover from login 
problems.


 What now?

 Allen

Bootup from CD or it's USB equivalent and select a rescue mode. That 
should make it possible to change the password. I don't remember where 
the root partition will be mounted by default if at all. Regardless, you 
need to open a terminal and run a command 'df' which would tell you if 
/dev/sda1 is mounted or not.

If it's mounted, go to the mount point, /tmp/target for example, and 
edit file /tmp/target/etc/shadow. But let's assume you used Ubuntu CD to 
boot from and that the hard drive was not mounted. That happened in my 
test with LinuxCNC in virtual environment, Virtualbox.

Open a terminal. You are going to be logged in as user ubuntu. You 
either need to prepend all suggested commands with sudo or become user 
root, my preference.

Switch to user 'root' with command 'sudo su -'. Run the following 
commands one by one:
fsck /dev/sda1
mkdir /tmp/sda1
mount /dev/sda1 /tmp/sda1

Now you should be able to run
ls /tmp/sda1
That is disk drive partition 1 root or top point. You should see 
directory named etc among others in there.

cd /tmp/sda1/etc
cp shadow shadow.bkp-- creates a backup copy.

ls shadow*   -- command shows two shadow files in that directory. Most 
likely 3 as one is shadow~ also a backup created by the system at some 
point.

Now edit file named shadow and remove the password, that is the part 
between the first and second ':' following your user name. In your case 
it might look like:

allen:$6$K5NgZYUK$3s2qEljrPGeX4LLLyuVVDGjA104:15942:0:9:7:::

   ^ --- remove  ^

Make sure you don't remove anything else. Save the file and reboot from 
the hard drive. You should be able to login without a password 
afterwards. Then change your password. Make sure you test login before 
you logout to prevent lockout again.

For editor you can use either vi or nano.

There are other ways to do it but that's the easiest IMO. The same would 
work for most if not all Linux distributions.

-- 
Rafael

--
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration  more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=157005751iu=/4140/ostg.clktrk
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] mach3

[Rafael Skodlar]

I have no horse in his race but found this thread amusing until now.

On 07/05/2014 05:23 AM, Stuart Stevenson wrote:
 Not sure what Aram is looking for but why is there so many catty remarks
 and so much sarcasm. I can think of more than one reason to explore a
 different system.

When you are exploring a different system then ask questions about it on 
related forums.

 If you just want to put someone down why don't you pick a target that will

You need to go read email exchange between some members of this list and 
target individual where he was extremely rude IMO [1] not very long 
ago. Only then you'll understand the reaction to his latest silly 
question about a product that has nothing to do with LinuxCNC.

[1] That kind of experience would keep me quiet for a year before asking 
for help again. I understand there is a language barrier but that 
exchange was not about problem with understanding English language. It 
was about ignorance after they tried to help him and people got tired of 
it. This reminds me of mailing lists rule no. #1: don't offend people 
that (try to) help for free.

And there is this: I've done a lot of linux support in early days 
mailing lists and news groups when there were only a handful of sites 
with anything on Linux and there was only one printed collection of 
messages from the Linux news group in first Linux Bible. It was 
understandable that software developers needed help to get systems up 
and start writing or porting applications to Linux but they lacked 
hardware or system administrative skills that some of us were able to 
provide.

These days people are too _damn_ lazy to search on one of numerous 
search engines before they send a request for help on a mailing list. 
That's why we see so many of the same (newbie) questions over and over 
again.

Searching on the Internet requires some skills too but this mailing list 
is not appropriate place for learning it IMO.

 fight back. All I see from Aram is questions about what problems he is
 having and then attempted positive comments to some of the negative
 response he gets. People should either not engage if they are not inclined
 to help or help.
 Just sayin
 Stuart

Your new target,

-- 
Rafael

--
Open source business process management suite built on Java and Eclipse
Turn processes into business applications with Bonita BPM Community Edition
Quickly connect people, data, and systems into organized workflows
Winner of BOSSIE, CODIE, OW2 and Gartner awards
http://p.sf.net/sfu/Bonitasoft
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] New toy (4'X8' Router) for conversion

[Rafael Skodlar]

On 06/10/2014 12:30 PM, Dave Cole wrote:
 Is buying a machine like that really worthwhile?


Socialist design never impressed me; delivered in bad crate even less.

My experience with Chinese made lathe was so bad I would never buy from 
them again. When you see it from close and start using it you see how 
sloppy is everything.

Trouble is they use American names for their products to make you 
believe it's US made.

 After you remove the dents, repaint the machine to cover the damage,
 replace the controls etc.

 Why not just build a machine from scratch?


Exactly. Why waste time on fixing what should come in working order in 
the first place. You can get very good CNC kits made in USA. I was 
impressed by http://www.cncrouterparts.com at Makerfaire in San Mateo 
last month.

Their kit is simple to assemble, strong to carry heavy loads, you can 
buy whatever you want and add your own parts as needed or desired. 
Something worth considering.

 If the crating was bad its quite possible that the frame could be
 tweaked also.

 I've heard of stories like this a few times.

 Dave

-- 
Rafael

--
HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
Find What Matters Most in Your Big Data with HPCC Systems
Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
Leverages Graph Analysis for Fast Processing  Easy Data Exploration
http://p.sf.net/sfu/hpccsystems
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Need help compiling the simulator on Ubuntu 14.4

[Rafael Skodlar]

Erik,

On 05/18/2014 12:42 AM, erik wrote:
 Hello everyone,

 Following these instructions
 http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Build_A_Simulator_Manually
 worked for my last setup with Ubuntu 12.4 (32Bit) quite well.

 Now I have a fresh install of Ubuntu 14.4 (64Bit) running and have
 trouble with the dependency libgnomeprintui2.2-dev. apt-get install
 does not find it (same for :i386). All other missing dependencies were
 installed without any problems.

 What do I have to do next?

 Cheers
 Erik


try this:

apt-cache search libgnomeprintui2

libgnomeprintui2.2-0 - GNOME print architecture User Interface - runtime 
files
libgnomeprintui2.2-common - GNOME print architecture User Interface - 
common files
libgnomeprintui2.2-dev - GNOME print architecture User Interface - devel 
files
libgnomeprintui2.2-doc - GNOME print architecture User Interface - doc files

cat /etc/issue
Ubuntu 12.04.4 LTS \n \l

--
Rafael

--
Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free.
http://p.sf.net/sfu/SauceLabs
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Hardware compatibility lists ?

[Rafael Skodlar]

On 12/07/2013 07:07 PM, Jon Elson wrote:
 On 12/07/2013 07:44 PM, Gene Heskett wrote:
 Generally speaking, both pata and sata dvd-rw drives have
 been Just Works(TM) for 3 or more years. If it doesn't,
 I've found its a good chance the drive is funkity, and a
 fresh $25 drive from Wallies plugs in and works. Not at
 all the problem child they were 10 years back up the log.
 OK, great!  I used to scan the compat. lists before buying
 anything the least bit special, but CD burners and sound
 cards were especially difficult to find ones that were
 Linux compatible.  Glad I don't have to fret over that
 stuff anymore.

 Thanks,

 Jon

I haven't had problems for over 10 years now but why bother with boot 
from CD-ROM or DVD anymore? Modern motherboards boot from USB device. 
You simply copy the same image to USB memory stick and boot from it.

Same goes for writing data or backups. USB memory sticks keep getting 
cheaper with bigger capacity, and you don't need special device to use 
them. 32GB go for less than $20 these days. Beats everything including 
BlueRay drives.

It's also easier to carry 2 or 3 USB sticks in the pocket than DVDs.

As far as compatibility goes, no problem with sound but Nvidia chip 
based video cards are the biggest pain to setup in my experience. Even 
Linus has strong opinion about them.

-- 
Rafael

--
Sponsored by Intel(R) XDK 
Develop, test and display web and hybrid apps with a single code base.
Download it for free now!
http://pubads.g.doubleclick.net/gampad/clk?id=111408631iu=/4140/ostg.clktrk
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] gcmc - v1.1.1

[Rafael Skodlar]

On 11/18/2013 02:45 AM, Bertho Stultiens wrote:
 On 11/18/2013 06:06 AM, Rafael Skodlar wrote:
 First, I'm not familiar with the program details. However, I wonder why
 unit of measure throughout the program matters at all?

 There is a simple reason why it matters. The output stage, where values
 are output to XYZ coordinates, will use a conversion depending on how
 you run gcmc (-i option).

 (Please note that the following interpretation will be consistent from
 version 1.3.0 (which is in the pipeline).)

 Without -i, gcmc will default all none units to mm.
 With -i, gcmc will default all none units to inch.

 If you use scripts solely for your own use, then there is probably no
 problem. However, when you start to share, then differences will
 eventually pop up. It is the eternal portability problem all coders have
 to deal with.

That's why I jumped on it ;-)


 Mixed hardcoded measurement units inside functions is unnecessary or
 even dangerous [1] I believe. Mixing metric, inches, etc. is like saying
 I'm going to create a program that will compile code written in python
 sprinkled with some functions in java, or perl in between into a running
 program.

 Not necessarily a bad thing. It depends on what you are trying to
 achieve. Units are a far cry from mixing languages. Units are a concept.


 If I understand this correctly, default units are metric. Why not
 convert everything to metric? Don't all CNC machines understand it?

 Gcmc will convert everything to millimeter on output by default (unless
 you specify -i). Using units at the output-level should in my opinion
 always be either/or, but the problem resides not in the output.

 The reason for units in gcmc is to support designs that are mixed units.
 The problem I regularly deal with is making a box (metric) for a PCB
 (imperial). Now I can specify the coordinates in the design's source
 units and have everything fit correctly. No longer any need to do manual
 conversions, which are prone to error.


 A unit is a unit and should not change throughout the gcmc program.
 Units represent distance relationship between places of work (drilling,
 welding, tapping, extruding material, etc.).

 Units do not change as such. You have distance units and angular units.

Exactly. Write once, resize with scale at the top. Everything else stays 
the same. That's my opinion.


 You normally only deal with units when adding them to constants. The
 point being, calculation will be distance or angular throughout the
 calculations.

 The problem is when you start mixing it with none. A value with no
 units associated can be interpreted in different ways, due to scaling.
 You need to /know/ what you are doing.

Don't assume things. I see programmers assume too many things and the 
rest of us have to deal with it later on.


 Gcmc has a fixed set of rules to deal with these cases. Like any
 programming language, the programmer must express things in a way that
 will reach the proper result.


 Well designed programs come with configuration file(s) which define
 default values. If you work with inches all the time then set that in
 the configuration file and all will be translated to metric in the
 final code. If you use something else, like the ear size of current
 crown prince of England, then that would translate to G-code
 accordingly. [2]

 Configuration files are a hell to work with. Which files make up a
 project? I'd rather have everything in one source and define a few
 variables at the top to make the setup flexible.


 It would also be easy to resize (scale) working units on the fly, i.e.
 change from the original to whatever you want. At the top of gcmc
 programs you would change/redefine default variables into whatever is
 needed in a particular program. Ex.
 units = inch
 scale_x = 1.75
 scale_y = 1.75
 scale_z = 1

 Scaling is an issue that can be seen on different levels. It is easy to
 scale any defined path in gcmc using vector math. However, there is no
 scaling at the output-level. The program defines what is output.

 Scaling at the output level raises a lot of issues, but you surely could
 implement an affine matrix to post-process all output.


 Knowing you dare dealing with inches, a special function would be called
 to swap metric G-code to US specific (G20 and G21, etc.). There are
 other G-codes exceptions, specific to some older machines, that would
 need code adaptation but everything else can stay the same for basic
 positioning functions I believe.

 Gcmc will not insert G20/G21 anywhere in the output other than in the
 prologue, which defines the initial setup. Everything else in the g-code
 will always use the same output unit format.

 You can, of course, insert them yourselves, but that may lead to
 unpredictable results. It is like changing a CPU from binary to decimal
 in the middle of a program.


 [1] If you are going to mix units (cutter.gcmc):

   move([-,-,-2.0mm]);
   tracepath(padpath + offset, -2.0mm

Re: [Emc-users] gcmc - v1.1.1

[Rafael Skodlar]

Hi Bertho,

On 11/17/2013 02:23 AM, Bertho Stultiens wrote:
 On 11/16/2013 11:37 PM, Tony Zampini wrote:
 The way I have been using GCMC is to not use any unit specifiers after
 numbers.
 This way, all I have to do is make sure I include G20 (program in inches) at
 the top, and everything is
 treated as inches, since GCMC will not do any conversions when no explicit
 units are
 specified (am I correct, Bertho?)

 In my example, I failed to mention that I used G20 at the beginning of the
 program.

 Yes, specifying no units at all will treat all measures as none and
 translate the values as is to the output.


First, I'm not familiar with the program details. However, I wonder why 
unit of measure throughout the program matters at all?

 However, you must take extreme caution because some built-in functions
 will add units due to implicit conversions. I do not think that it is
 wise not to specify units because future versions of gcmc may interpret
 things differently (see below).


Mixed hardcoded measurement units inside functions is unnecessary or 
even dangerous [1] I believe. Mixing metric, inches, etc. is like saying 
I'm going to create a program that will compile code written in python 
sprinkled with some functions in java, or perl in between into a running 
program.

If I understand this correctly, default units are metric. Why not 
convert everything to metric? Don't all CNC machines understand it?

A unit is a unit and should not change throughout the gcmc program. 
Units represent distance relationship between places of work (drilling, 
welding, tapping, extruding material, etc.).

Well designed programs come with configuration file(s) which define 
default values. If you work with inches all the time then set that in 
the configuration file and all will be translated to metric in the 
final code. If you use something else, like the ear size of current 
crown prince of England, then that would translate to G-code 
accordingly. [2]

It would also be easy to resize (scale) working units on the fly, i.e. 
change from the original to whatever you want. At the top of gcmc 
programs you would change/redefine default variables into whatever is 
needed in a particular program. Ex.
units = inch
scale_x = 1.75
scale_y = 1.75
scale_z = 1

Knowing you dare dealing with inches, a special function would be called 
to swap metric G-code to US specific (G20 and G21, etc.). There are 
other G-codes exceptions, specific to some older machines, that would 
need code adaptation but everything else can stay the same for basic 
positioning functions I believe.

[1] If you are going to mix units (cutter.gcmc):

 move([-,-,-2.0mm]);
 tracepath(padpath + offset, -2.0mm, DWELL);
 move([-,-,-1.0mm]);

 goto([50mil, 0mil] + offset);

I bet you will break the tools if not worse because of a typo or you'll 
miscalculate something somewhere. What happens if you enter 
move([-,-,-1.0m]) instead of move([-,-,-1.0mm]) Bam!

Opinions on syntax:
move([-,-,-1.0mm]) - better: move([0,0,-1.0]) or
move(0,0,-1.0mm) is it's used in other languages.

Line termination is also too perl-ish or java-ish IMO. LF tells you it's 
the end of command or comment, why add extra character to it? I hated ; 
in perl and javascripts.

Looking at Gcmc example on the website I see need to declare the 
variables. Python for example takes care of variables by itself.

[2] I just wish that US would stop using silly imperial inches! We are 
in 2013 with dual or four core computers in everybody's pocket so it's 
easy to convert to metric. There are free HP calculator emulators 
available for android for those who insist dragging feet into this century.

 To clarify my previous email, I use the literal statement to
 include the G20. I don't use the -i switch. So GCMC thinks its
 working in mm, and by not using any explicit unit specifiers, no
 conversions are done. Then when I run the gcode, all values are
 treated as inches.

 Adding literal to the code is like specifying assembly-code in a
 C-program. It is non-portable as hell. It works fine for your use in a
 specific case, but may have side-effects as things evolve.


Agree! Don't do it. Unit of measure should be the same in all functions 
in the same program to be safe.

 The -i option was introduced to enable change of metric/imperial
 defaults. That was not implemented at first, but the new version *will*
 implement complete and throughout change of default. That change also
 covers the implicit conversions in built-in functions and therefore will
 no longre require you to insert a literal change of units.


Command line options are ok, but it's hard to remember them all. Some 
are awkward or missing: --help is there, -h is not. When run without 
parameters, the program should either dump a help page or present a 
menu. It's not hard to provide a simple menu to handle input or output 
file locations etc. select in bash can do this very well.

Adding default 

Re: [Emc-users] Correct use of subroutines

[Rafael Skodlar]

On 05/12/2013 01:48 PM, Steve Blackmore wrote:
 On Sun, 12 May 2013 10:08:10 -0700, you wrote:


 code.
 On 05/12/2013 02:16 AM, Steve Blackmore wrote:

 These days unlimited code length is the norm and subs are frowned
 upon commercially in my experience. The operator can't easily alter
 the code, if required, on the fly.  Some of the subs I've seen here

 But they can make changes somewhere in te middle of thousands lines of code?

 Of course they can - they have to understand Gcode, but not some obtuse
 programming language or complex math. They often simply pause the
 machine, alter the code and restart.

Of course, if they spend the same amount of time to learn some advanced 
code but still simple code instead of G-code they could do much better.

Using your analogy we should pause faulty applications in the middle 
somewhere, change assembly code, or registers in the CPU, and memory 
locations, and proceed from there. I only see that being done on PDP-1 
and similar machines in computer museum.

Where did I mention need for complex math? Knowing basic geometry is 
needed even in non CNC machining. No?


 and elsewhere are so complex that they could only be written to show
 how clever the author is, not for ease of machine control. A typical
 op has no chance of understanding them.

 That might be unfair to the programmer.

 But often true.

 If the operator does not understand complex operation, that's precisely
 why a subroutine should be used IMO. Granted, I do not know much about G
 code and machining in general as my work mostly revolved around
 computers and connections to other stuff, sometimes CNC. However, code
 is code, it automates processes and minimizes redundancy if written
 properly. That goes for CNC machine operations or OS administrative tasks.

 Your key statement  I do not know much about Gcode and machining in
 general as my work mostly revolved around computers and connections to
 other stuff - You then go on to assume that applies to CNC operations.

Machines behave predictably based on what the code says. There is no 
magic there. Bad code, bad results in my real life experience. Who would 
want machines that do not behave predictably???

 Please do some reading up on Gcode - a good place to start is Peter
 Smid's CNC programming handbook. ISBN 978-083113347-4

I have a book on CNC programing and a floppy disk that came with it. I 
used to run programs from that book under OS I left behind 19 years ago. 
However, code is code, including ~50 years old G-code that never evolved 
much. I know, old industry doesn't die fast but some parts of it just 
should. A number of attempts have been made over the years but no 
solution has emerged.

 An excellent book that everybody who thinks they know, or wants to know
 about CNC should read.

Of course, the more you know the better you are. However, knowing G-code 
is irrelevant for this discussion. The fact is that G-code was outdated 
long time ago, proved by the fact that manufacturers started to use 
different implementations to accommodate new functionality.

Perhaps you should check what's going on around the world 
http://reprap.org/wiki/G-code
http://www.ghielectronics.com/community/forum/topic?id=11023
to understand the limitations that G-code is imposing on end users.


 Drawing an analogy from the sysadmin world, I can say that well
 defined, understood, and tested functions (subroutines) are the best way
 to develop, maintain, and use code efficiently and consistently in any
 environment.

 You also know that is the holy grail in most environments. Getting
 programmers to document their code is the bane of computer ops.

We have an agreement here.


 BTW - I've been there, done that and threw away the tee shirt. I was a
 Senior Ops Analyst for British Gas for some years in an ICL VME
 environment and wrote and controlled SCL, but gave up that boring world,
 moved up to designing and managing installations of Distributed
 Mainframe Systems, their HVAC environments and CHP plant development.
 Finally went back to my roots with Industrial Engineering and CAD/CAM
 solutions when British Gas was devolved.

 I hate to see systems administrators write more or less same
 functionality, sometimes (?) with hardcoded IP numbers, hostnames etc.
 into their scripts which make it extremely hard to read or maintain
 because some variable with no name like i or j is defined deep into
 the code, line 315 or whatever.

 The same could be said for Gcode as when it fails on line 315 the answer
 is on something defined in line 20

Of course, the issue is understand the code. It would be easier to 
read/enter one line like selecttool 3; position 11, 22; drill 5mm, 
than using whatever equivalent number of obscure G-code. At least for 
occasional CNC users and emerging personal 3D printing.


 When the issue is known, a well defined subroutine would account for
 that and drill/peck depending on depth and perhaps drill bit size.
 Program once, 

Re: [Emc-users] Correct use of subroutines

[Rafael Skodlar]

On 05/13/2013 03:56 PM, RogerN wrote:
 From: Steve Blackmore
 .Sent: Monday, May 13, 2013 4:48 PM
 To: emc-users@lists.sourceforge.net
 Subject: Re: [Emc-users] Correct use of subroutines

 On Mon, 13 May 2013 22:50:51 +1000, you wrote:


 Given that it is now mainly a machine-to-machine data standard, the
 fact that it is archaic, antiquated and poor as a programming language
 is largely irrelevant.

 True.

 True, and well expressed, but the LinuxCNC elves have added much needed
 flow-control, looping, and subroutine language elements to LinuxCNC
 gcode, so that it is nearly as good as assembler now. Named variables,
 too, have done a lot for its utility as a hand-wrought language.

 Your LinuxCNC elves are working for a tiny niche market only. Nobody
 except programmers seem to think it's necessary. It certainly isn't to
 produce a part and all the loops and subroutines in the world don't
 machine that part any better or quicker.

 Your never going to persuade thousands of companies or the big players
 like Fanuc or Siemens to drop traditional Gcode. It's been tried before
 and then, as now, changes were as popular as pox in a brothel.

Never? One software company kept telling us in the mid 90's that 
Netscape, java, and Linux are not worth mentioning. There are plenty of 
similar examples of disruptive technologies coming from little guys 
that all of a sudden become something everyone wants to have.


 Commercially they have too much time, skill and money invested in
 something that already works and if it ain't broke there is no need to
 fix it.

 Steve Blackmore
 --

 A program wouldn't have to give up it's ability to understand G code to
 understand additional instructions.  For example a canned routine for

Finally somebody that understands what I tried to say in my earlier 
emails which veered far away from it's subject line.

 drilling doesn't mean the control no longer understands milling

exactly.

 instructions.  My Anilam control only had 1000 instructions program memory
 but by using loops it could execute much more than 1000 instructions per
 press of the start button.  As an example, one part I made was cut out in a
 sheet of Delrin, the size of piece that fit in my vise allowed me to machine
 3 rows and 7 columns, 21 pieces.  By using nested loops I was able to
 machine all 21 parts using the code for 1 part repeated in 3 rows and 7
 columns.

 So continuing the same idea, if the program understood it, I could define
 one part, a turbine blade maybe, and repeat the part multiple times rotated
 to different positions.  When you have so much memory and hard drive space
 it hardly seems worth writing efficient code, but if you can correct one
 blade it could correct every blade, versus having to edit the same problem
 for every different blade.


Exactly again.

 I wish LinuxCNC understood G code plus something like C++ or Basic language.

 RogerN


HP developed a specialized CNC language HP-GL to drive NC (Numerically 
Controlled) machines, plotters. Because plotters use microcontrollers, 
they could handle higher level language with simple commands to select a 
pen, move it to starting position, put it down, and draw a line. 
http://en.wikipedia.org/wiki/HPGL

Similarly could be done in CNC machines. One problem remains, standard. 
G-code remains in it's place precisely because we don't have such a 
standard.

In computer world we have standards like XML, JASON, etc. which are 
interpreted in our browsers and other applications to interact with 
powerful server clusters in data centers. No such thing for CNC.

If there was such a higher language like 
http://en.wikipedia.org/wiki/APT_programming_language but more advanced 
and object oriented, and XML perhaps, it would be possible to prevent 
CNC machines to do something stupid. For example, you enter the size and 
position of the raw material and tool size. The machine could tell of 
there was a possibility for tool collision if you manually entered bad 
command. With controllers that only handle G-code there is no such 
verification, or is it?

Fact, systems that handle G-code are technically still on the same level 
as computers using paper tape for program input decades ago.

It's easy to recognize the resistance from those who are afraid of (CNC) 
changes. However, we do not need to be scared into thinking that we need 
to know the G-code in order to understand CNC in general: 
http://en.wikipedia.org/wiki/Numerical_control
Somebody did an excellent job on that page!

-- 
Rafael

--
AlienVault Unified Security Management (USM) platform delivers complete
security visibility with the essential security capabilities. Easily and
efficiently configure, manage, and operate all of your security controls
from a single console and one unified framework. Download a free trial.
http://p.sf.net/sfu/alienvault_d2d
___
Emc-users 

Re: [Emc-users] Correct use of subroutines

[Rafael Skodlar]

On 05/12/2013 02:16 AM, Steve Blackmore wrote:
 On Sat, 11 May 2013 21:40:55 -0700 (PDT), you wrote:

 Re: Correct use of subroutines

 My old school answer is that most subs should be written in (G91)
 Incremental mode.

 My more modern method can be either G90 or G91 depending on the
 application but I use variables for D, F, Z and loop to run Z level
 roughing and finishing often using several tools.

 RE: G84

 BTW it is the programmers job to lie to the machine - the machine
 should NEVER lie to the operator.

 The value you use for your F should be exactly what the machine
 feeds at. No IF's ands or buts... 100% F This is also a key safety
 point - The operator needs to be able to read the G-code and KNOW
 exactly what the machine will do when it tries to execute that
 code.

 Which is the reason that many commercial shops never use subs now.

 Now if you are using a using a specific tool which needs to run at
 a reduced feed then you as the programmer should alter the
 programed feed rate. In this case if you wanted to use a Osub to do
 your tapping and you wanted it to feed at 95% thats fine - its all
 man readable in the Osub - Fixed cycles don't display the inner
 workings so others would not know of your 95% mod.

 While I started out on a Bandit CNC way back and subs were critical
 with controls that could only handle like 99 command lines.

 These days unlimited code length is the norm and subs are frowned
 upon commercially in my experience. The operator can't easily alter
 the code, if required, on the fly.  Some of the subs I've seen here

But they can make changes somewhere in te middle of thousands lines of code?

 and elsewhere are so complex that they could only be written to show
 how clever the author is, not for ease of machine control. A typical
 op has no chance of understanding them.

That might be unfair to the programmer.

If the operator does not understand complex operation, that's precisely
why a subroutine should be used IMO. Granted, I do not know much about G
code and machining in general as my work mostly revolved around
computers and connections to other stuff, sometimes CNC. However, code 
is code, it automates processes and minimizes redundancy if written 
properly. That goes for CNC machine operations or OS administrative tasks.

Drawing an analogy from the sysadmin world, I can say that well
defined, understood, and tested functions (subroutines) are the best way
to develop, maintain, and use code efficiently and consistently in any
environment.

I hate to see systems administrators write more or less same
functionality, sometimes (?) with hardcoded IP numbers, hostnames etc.
into their scripts which make it extremely hard to read or maintain
because some variable with no name like i or j is defined deep into
the code, line 315 or whatever. One silliest example is seeing their 
email address embedded in scripts, multiple times in some cases. That 
only becomes known when all of a sudden there is a major
failure somewhere and nobody got any warnings prior to that because some 
user account was disabled.


 If you are hand coding they can save some typing and cut 'n' paste
 but in the CAD/CAM world they are pretty worthless.

I do not agree. The problem is (very) awkward UI in CNC consoles from
what I've seen on older systems. It's very archaic going all the way 
back to paper tape days. Can't say the same for newer ones.

 One of the big shops I work with don't even use canned cycles unless
 they really have to! They run a prize scheme where if an operator
 saves time or tooling cost on code they are rewarded. One operation
 they do is a deep drilling op in a nasty cast steel. They use a
 through coolant drill but it still clogs up with chips. It was soon
 realised that it was pointless withdrawing the drill until it had
 reached a certain depth to clear it. The deeper it gets, the more it
 needs clearing etc. The drilling operation is done with G1 and G0
 moves, easy to adjust to get the optimum without unnecessary machine
 moves. It's ended up as hybrid drill, peck, deep hole operation and
 saves several seconds. That saved time is more profit.

 Steve Blackmore

When the issue is known, a well defined subroutine would account for
that and drill/peck depending on depth and perhaps drill bit size.
Program once, reuse tested code countless times. I would give the
operator double brownie points for modifying, documenting the
subroutine, and send feedback to the programmer. Why? Because other 
operator on the line could use the same subroutine when
Joe is not there or caches in on his last check from that company.

-- 
Rafael Skodlar

--
Learn Graph Databases - Download FREE O'Reilly Book
Graph Databases is the definitive new guide to graph databases and 
their applications. This 200-page book is written by three acclaimed 
leaders in the field. The early access version is available now

Re: [Emc-users] Correct use of subroutines

[Rafael Skodlar]

On 05/12/2013 10:34 AM, andy pugh wrote:
 On 12 May 2013 18:08, Rafael Skodlar ra...@linwin.com wrote:

 If the operator does not understand complex operation, that's precisely
 why a subroutine should be used IMO. Granted, I do not know much about G
 code and machining in general as my work mostly revolved around
 computers and connections to other stuff, sometimes CNC. However, code
 is code, it automates processes and minimizes redundancy if written
 properly. That goes for CNC machine operations or OS administrative tasks.

 I think that you might be mistaking G-code for a programming language…

It is a programing language.

G-code is the common name for the most widely used numerical control 
(NC) programming language, ... see http://en.wikipedia.org/wiki/G-code

It was created to program machines that make parts over and over again. 
Computers were very limited in computing power and memory in those days 
so the code was very terse and limited. Some of the CNC machines did not 
have computers at all. There was some electronic logic to read the paper 
tape and drivers to flip relays, breaks, clutches, etc. Seen that.

G-code began as a limited type of language that lacked constructs such 
as loops, conditional operators, and programmer-declared variables with 
natural-word-including names (or the expressions in which to use them). 
It was thus unable to encode logic; it was essentially just a way to 
connect the dots where many of the dots' locations were figured out 
longhand by the programmer.

Unfortunately that did not improve much over the years. There is no 
longer reason to keep saying G tis or G that when a more powerful 
descriptive language could be derived for the same results.

Improvements made to G-code were small and show it's age, just look at M30.

Since about the mid-2000s, the era has finally arrived when the death 
of manual programming (that is, of writing lines of G-code without 
CAD/CAM assistance) sometimes seems to be approaching. However, it is 
currently only in some contexts that manual programming is obsolete.

CNC is stuck in it's own assembler age. If the industry switched to a 
modern intermediary language like python, you could say

from geometryset import circle
from geometryset import square

circle = circle(x, y, radius)
circlePoints = 5# number of holes on the circle
startingPoint = 0   # 0 degrees from X- axis
drillDepth = userInput
drill(circle, circlePoints, startingPoint, drillDepth)

squareX = 10
squareY = 15
squareSize = 55
squareAngle = 45
drill(square, squareX, squareY, squareSize, squareAngle)

to drill number of holes in the circle and in square. You could add user 
input for each variable to make it more flexible.

help(geometryset)
square, circle, star, 

help(drill) 

Much simpler to learn and remember than Gxx with it's limitations. It's 
all vector graphics at the end. Operators could learn it just as fast as 
they learn G-code. If you wrap this into functions, i.e. subroutines and 
stored them into library, you could run complex jobs from the command 
line very fast.

I'm sure it's been done in some form but G-code keeps being the focus in 
discussions. G-code shows it's limitations in 3D printing and is not 
something that masses will adopt for quick one of a kind products.


 It can be used as one, certainly, but then so can Postscript.


By that logic, we would program web sites code in assembler. But then 
millions use inches and feet even when they tweet ...


-- 
Rafael

--
Learn Graph Databases - Download FREE O'Reilly Book
Graph Databases is the definitive new guide to graph databases and 
their applications. This 200-page book is written by three acclaimed 
leaders in the field. The early access version is available now. 
Download your free book today! http://p.sf.net/sfu/neotech_d2d_may
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] off topic opinions

[Rafael Skodlar]

On 04/26/2013 06:11 AM, kqt4a...@gmail.com wrote:
 Is anyone using http://www.grizzly.com/products/Combo-Lathe-Mill/G9729 or 
 http://www.use-enco.com/CGI/INSRIT?PMAKA=328-1310PMPXNO=25221739PARTPG=INLMK3
 They appear to be the same machine except the enco is higher priced
 How is the quality and company support

 Richard


G0516 Combo Lathe w/ Milling Attachment (hobby) user here. Nice picture 
but way different from seeing it close. The whole thing is as sloppy as 
8 floppy disk. Mechanism for moving cutting tool is not smooth and 
noticeably different when used in one direction or the other.

My first delivery was in upside down and partially broken crate, with 
damaged lathe that I had to reject on the spot. They sent me another one 
(?) weeks later. That one broke down few weeks later. Sleeve bearing for 
the threaded rod seized which caused brittle gears to literally grind to 
stop. The stiff coupling between the threaded rod and connection rod 
with a gear has a small angle which of course puts huge load on the 
sleeve bearing to seize.

They sent me replacement gears and sleeve bearings but that not better. 
When I turned the gears manually I realized that the connection rod and 
the threaded rod are not in one line! I'm sure it would break if I ever 
use it again. The gears are made of cast iron not steel!

With exception of motors, there are no ball bearings in lathe so you 
start hearing squeaking noise after months of little or no use. Need to 
take the gears apart to lube the damn thing.

Remember, I'm a hobby, not a machinist user and it still broke down with 
so little use. What happens when you want to use it more seriously?

Another unfortunate thing is that it's not metric. They could not 
provide simple metric conversion kit either.

Conclusion: piece of crap from China, unsuitable for precision work, if 
you ask me. I've seen the same thing in blue color sold in Europe. Stay 
away from it. I would never buy it again. Should have saved $$ and buy a 
US (is there such a thing?) or German made Proxxon.

Combo also has a drawback, there is little space between the lathe 
headstock and milling area. They are too close to each other for many 
things in my experience.

-- 
Rafael

--
Try New Relic Now  We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app,  servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] a bit of New Year levity

[Rafael Skodlar]

On 01/04/2013 08:59 AM, Kent A. Reed wrote:
 Gentle persons:

 Too bad this was an April Fool's joke (I know, me it took a while to
 notice it, but I've been kind of busy). I could really use a case of
 Cut Away!, not to mention their MetalDhruker.

 http://www.onlinemetals.com/cutaway.cfm

 Regards,
 Kent


If it's good for metal rods it should be good for fingers too. Ordered 
it right away.

-- 
Rafael

--
Master HTML5, CSS3, ASP.NET, MVC, AJAX, Knockout.js, Web API and
much more. Get web development skills now with LearnDevNow -
350+ hours of step-by-step video tutorials by Microsoft MVPs and experts.
SALE $99.99 this month only -- learn more at:
http://p.sf.net/sfu/learnmore_122812
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] LinuxCNC documentation in russian

[Rafael Skodlar]

On 12/15/2012 08:33 PM, dave wrote:
 On Sat, 2012-12-15 at 22:31 -0600, Jon Elson wrote:
 andy pugh wrote:
 On 15 December 2012 20:07, Rafael Skodlar ra...@linwin.com wrote:


 One thing that Americans are terribly conservative about is standards.
 They will not adopt metric system, not even at the gun point.


 Not even using a 9mm?

 Yeah, the military has been totally metric for quite some time.  All
 guns were metric
 during the Vietnam war, now everything is metric, tools, maps, parts.

 Autos are all metric except for wheel lugnuts.

 Aircraft manufacturing is all metric.  Lots of consumer products are made
 in metric measure, with the US measure also on the package.  Sodas, for
 instance.

 Jon

 Even when I was a GI, Berlin crisis. the maps were partially metric. X,Y
 grid was Km but the heights were still in feet. ;-)

 Dave

That's because road signs were in metric so you did not need to convert 
is my guess. With exception for bridges, there are no signs for heights 
on the roads.

It's a SHAME (on Burma, Liberia and the United States) that with all 
microcontrollers in any electronic device these days we still don't use 
SI in daily life. American machine shops rarely use metric in my 
experience. Same goes for hardware stores, metal suppliers, food 
containers, cookie recipes, weather, GPS, etc.

http://en.wikipedia.org/wiki/SI
and when they do use it, it's not correct in most cases, see specs for 
stepper motor torque in kg-cm instead of Nm by many online sellers.

Local HW store sells metric screws and bolts but they are very 
expensive. What's weird is that they sell bolts in packs of 5 and nuts 
in packs of 3. Go figure. People don't realize how expensive it is to 
keep two systems in parallel in modern time with growing international 
trade and Internet.

What is interesting is that Americans having European roots did not 
adopt metric system while Japanese did long time ago.

-- 
Rafael


--
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
Remotely access PCs and mobile devices and provide instant support
Improve your efficiency, and focus on delivering more value-add services
Discover what IT Professionals Know. Rescue delivers
http://p.sf.net/sfu/logmein_12329d2d
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] LinuxCNC documentation in russian

[Rafael Skodlar]

Hi Peter,

On 12/15/2012 01:56 AM, Peter Blodow wrote:
 Przemek,
 when I want to build (or also keep) up a business I have to be reliable
 for the customers and grant continuity for my products. Therefore, I
 have to demand just the same from my suppliers. I will have to put a lot
 of money into my business, in most cases from loans, and have to keep
 the risk of loss as low as possible. I will need banks and insurances
 which want to calculate the risk of lending money to me and at which
 cost.  All this calls for professional solutions, as well in machinery
 as in computer hard- and software, be they ingeniuous, modern, inventive
 or not, they have to be reliable and calculable in the first place. A
 large company with an international name gives me more of this security
 than the best and most ingenious free product. This is why Linux can't
 break though on the professional office market, and this is why LinuxCNC
 can't make it on our professional production market. A lot simplified,
 as the saying goes if it costs nothing, it's worth nothing.

That's not true about Linux. My career depends on companies using Linux. 
It's professional software that makes (Internet) world go around and 
Linux plays a huge role in this. Virtualization would not happen without 
need to run multiple operating systems on the same hardware at the same 
time for example. Whole new innovations came out of that, supercomputing 
in cheap hardware, cloud computing, better HW utilization, etc.

Here it doesn't matter how many university degrees you have as long as 
you know how to make systems run. Many of us simply learn on the job to 
become productive Linux Meisters.


 Another thing is that there are legal regulations to be licenced and
 authorized to run a commercial machine shop.  To be allowed open up a
 production of any kind you have to have a certain degree (Meister in
 German) from the Chamber of Crafts and Industry, preceded by a three
 years apprenticeship with final exam, several years of professional
 experience and a Meister school degree exam in the end. During all this

Not long ago, public TV had an interesting program about that German 
apprenticeship system in practically every kind of business or trade, 
from chimney sweeper, to butcher, pastry maker, etc. Impressive indeed.

Traditional trades do evolve into highly specialized fields in the 
industry. Some is due to regulations as a reaction to accidents or 
disasters. Other is simply a tradition or to protect the union which 
prevents innovation in many cases. Meister wants to keep doing it the 
old way, while there might be a better way.

Countries without industrial tradition make poor quality products. We 
(regretfully) end up buying cheap machines from China instead of quality 
from elsewhere. My small lathe/mill is an example of that. It's such a 
sloppy work and you can't improve it. Chinese !CRAP! for 1/3rd of what a 
German made would cost. Granted, it's only for a hobby; no way to make a 
living with it.

In the US you can start almost any kind of business regardless of your 
profession. If you don't know how to do things you try a few times until 
it works, hire somebody to make it work, or outsource to other 
businesses. Example is emerging private space industry. I know people 
who used to work in computer industry, made some money, and decided to 
go build rockets. They are successful.

Some use KickStarter to start new things or fill the need: 
http://joshondesign.com/2012/09/17/innovator_terencetam

And when things go bad, we have lawyers that line up to take care of 
the problem ;-)

 time, people get acquainted with professional equipment and want to rely
 on it in their own business later on. Large companies, as is
 understandable, do a lot of advertizing for these people in order to
 make them stay with their equipment later when they are on their own. I
 have experienced myself that shop workers insisted of buying a specific
 machine without which they would refuse to grant the quality of the
 products they were making.

 So, there is your German conservativism. Our dual educational system
 of schools and apprenticeship with its pursuit of quality and continuity
 has made Germany a blooming economy among a lot of declining countries

that's true, however, it would not be possible to keep going on it's 
own. Other countries buy goods from Germany or do labor intensive work 
for German companies, sometimes on borrowed money and that made things 
go ugly for a number of EU members when the economy went south.

 all around in Europe, and this only a few decades after a war that had
 destroyed three quarters of all buildings, virtually all means of
 production and millions of men to run this production.

 Peter


Well, that would not happen without help from the USA to rebuild West 
Germany after the war. At least the rate of reconstruction would be much 
slower for whole EU. Remember that huge number of men were killed in the 
war and had to 

Re: [Emc-users] LinuxCNC documentation in russian

[Rafael Skodlar]

On 12/15/2012 12:57 PM, andy pugh wrote:
 On 15 December 2012 20:07, Rafael Skodlar ra...@linwin.com wrote:

 One thing that Americans are terribly conservative about is standards.
 They will not adopt metric system, not even at the gun point.

 Not even using a 9mm?


they opt for .45 since 1800's. BTW, forgot a disclaimer in previous 
message. I tried everything else but a gun to make them adopt metric system.

Back to documentation. I too was wondering about translating sections of 
open source software in general. It all depends on where one needs to 
make changes. Messing with source code is not good approach as that can 
break it. On the other hand, if the text for UI comes from separate 
files it would be relatively easy to make translations. There are 
limitations of course, like GUI space for number of characters, fonts, 
icons, etc. but that too should not be too difficult to manage IMO.

It's my belief that LCNC would be easier to adopt around the world if 
the user interface supported other languages. People using CNC machines 
do not necessarily know English that well if at all. Smaller ethnic 
groups are always shorthanded in this regards.

Ideally, one would switch between languages with a click or two. 
Important for European theater of operations and others as well. I'm not 
sure why keyboards don't have a dedicated key for LANG. There is this 
underutilized SysRq key since DOS days.

I haven't looked into LCNC specifically so I can't comment on how hard 
it is to modify it for use with other languages. Standalone XML files 
for text use in GUI could be translated pretty fast IMO. There are not 
too many changes between application releases so the maintenance would 
not require much work.

-- 
Rafael

--
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
Remotely access PCs and mobile devices and provide instant support
Improve your efficiency, and focus on delivering more value-add services
Discover what IT Professionals Know. Rescue delivers
http://p.sf.net/sfu/logmein_12329d2d
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Power and data cables for CNC

[Rafael Skodlar]

On 05/22/2012 02:21 AM, andy pugh wrote:
 On 22 May 2012 07:53, Rafael Skodlarra...@linwin.com  wrote:

 Using Cat-5 was a surprise to me as it's a bit stiff unless each wire is
 made of even smaller wires, not common in general use.

 There is solid stranded for fixed installation and stranded for patch cables.
 http://uk.rs-online.com/web/c/cables-wires/network-communication-cable/cat5e-cable/?searchTerm=cat5
 The stranded would be very much preferred.

 There are many other types of multicore cable, it is just that CAT5 is
 readily available.


There are different kinds of stranded CAT-5 from my experience. However, 
stranded cat-5/6 were not designed for bending over and over thousands 
of times IMO. Connecting laptops is one thing, wiring sensors and 
electronics on moving mechanisms on CNC machines is another.

I've come across multi-wire cables with magic white powder inside 
which made bending much smoother. Insulated wires inside main cable 
jacket were sliding along each other easily. I did not pay attention to 
that at that time but now I suspect I know what that was all about.

-- 
Rafael

--
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Power and data cables for CNC

[Rafael Skodlar]

On 05/21/2012 09:15 AM, Jon Elson wrote:
 Rafael Skodlar wrote:
 On 05/20/2012 12:12 PM, Dave wrote:

 On 5/20/2012 12:40 PM, Rafael Skodlar wrote:

 I wonder what kind of power and data cables would you recommend for use
 in X-Y-Z CNC about 1.2mx1m size? Besides power, it's not clear to me how
 many data lines for sensors, encoders, motor control, etc. are needed in
 general.

 I'm well familiar with flexing and shielding issues in general since
 early HP plotter days. I was hoping to get a simple answer based on real
 life experience, get this or that cable with x number of shielded and
 stranded or twisted wires ;-) That's why I mentioned size as that would
 give one an idea of motor sizes and other requirements.

 Is it better to have one cable with x number of wires to take care of
 all needs or a number of smaller cables (y) with x/y wires?

 With non-filtered motor drives, it is pretty important to NOT put
 encoder or other signal-level
 wires in the same cable as the motor wires.  I usually separate encoder,
 home/limit sensors
 and motors on 3 cables per axis, even though my motor drives ARE filtered.
 A plain quadrature encoder needs 4 wires, if it has index then 5.  If
 differential,
 then 6 or 8 wires.  If brushless motors are used, those usually need
 Hall sensors,
 add 4 more wires.  Stepper motors need a minimum of 4 wires, brush servos
 need 2 plus maybe a safety ground, brushless would need 3 plus ground.

 You can check the catalogs for the number and size of the wire strands.
 The more fine
 wires there are, the better the cable will handle flexing.  The good
 stuff has #36 AWG
 or finer strands, thinner than hair.

 Jon

Thank you very much for detailed advice to all of you that responded to 
my question. I received enough material to spend a few evenings doing my 
homework. It's much easier to start knowing what others have tried and 
what works in different circumstances.

Using Cat-5 was a surprise to me as it's a bit stiff unless each wire is 
made of even smaller wires, not common in general use.

-- 
Rafael

--
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Power and data cables for CNC

[Rafael Skodlar]

I wonder what kind of power and data cables would you recommend for use 
in X-Y-Z CNC about 1.2mx1m size? Besides power, it's not clear to me how 
many data lines for sensors, encoders, motor control, etc. are needed in 
general.

Is it preferable to carry power and handful of control signals to the 
circuit on the gantry and preprocess some functions there or have 
everything wired into the central box? That would dictate the number of 
wires in the cable needed and what kind you need to buy.

Would cable carrier from McMaster-Carr # 55835K432 be sufficient?

Thanks,

-- 
Rafael

--
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Power and data cables for CNC

[Rafael Skodlar]

On 05/20/2012 12:12 PM, Dave wrote:
 On 5/20/2012 12:40 PM, Rafael Skodlar wrote:
 I wonder what kind of power and data cables would you recommend for use
 in X-Y-Z CNC about 1.2mx1m size? Besides power, it's not clear to me how
 many data lines for sensors, encoders, motor control, etc. are needed in
 general.

 Is it preferable to carry power and handful of control signals to the
 circuit on the gantry and preprocess some functions there or have
 everything wired into the central box? That would dictate the number of
 wires in the cable needed and what kind you need to buy.

 Would cable carrier from McMaster-Carr # 55835K432 be sufficient?

 Thanks,



 Figure out what cables you need to move back and forth and then choose
 the carrier.  Jamming cables into a cable carrier is not a good idea.

I'm well familiar with flexing and shielding issues in general since 
early HP plotter days. I was hoping to get a simple answer based on real 
life experience, get this or that cable with x number of shielded and 
stranded or twisted wires ;-) That's why I mentioned size as that would 
give one an idea of motor sizes and other requirements.

Is it better to have one cable with x number of wires to take care of 
all needs or a number of smaller cables (y) with x/y wires?

 If you don't do some figuring before you know what you are going to carry,
 you will either have too much, or not enough space.   Guessing usually
 doesn't work out nearly as well.


Agree, and that's why I tried to see what others are doing on this issue 
as it seem to be poorly documented in general. I try to capture this 
from numerous pictures but wires seem to generate little interest in 
general.

 As an alternative supplier; Igus sells direct in the US.
 http://www.igus.com/default.asp?c=usL=en

 Dave


Links are always encouraging. However, I'm still struggling with the 
variety of cables mentioned on Igus site: data cable, bus cable, servo 
cable, control cable. Lots of homework ahead I guess, and that's before 
the experiment can begin.

Thanks guys,

-- 
Rafael

--
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] DIY output driver

[Rafael Skodlar]

On 01/27/2012 10:23 PM, Viesturs Lācis wrote:
 2012/1/28 Rafael Skodlarra...@linwin.com:

 Of course that won't work with most if not all the suggestions I've seen
 so far. If I understand it correctly, your laser already has a circuit
 to drive it at reasonable current to do it's magic.

 Circuit I suggested earlier was under assumption you have a bare laser
 diode connected to it. You cannot daisy chain circuits one after another
 and expect laser to work properly.

 Relays are out of question IMO because they are too slow mechanical
 devices with many other drawbacks for this application. I see no reason
 to bring in solid state relays into the picture either. You are not
 driving high voltage stuff.

 You either need to modify your PCB that came with the laser diode or
 build a new circuit. Maybe you can reverse engineer that PCB and use one
 spot to inject signal from the EMC side.

 Ed, Kirk, Rafael, thank You!

 Given my lack of skills in electronics, I think that interfering and
 modifying the pcb that comes with laser is the last thing I want to
 do.
 Here are some pics of the laser and its pcb I took late in last night
 with my phone (it was late enough that I forgot to upload them
 yesterday evening):
 http://picpaste.com/2012-01-27_21.54.59-QzJqS24n.jpg
 http://picpaste.com/2012-01-27_21.55.16-mx4U2GiU.jpg
 http://picpaste.com/2012-01-27_21.55.36-wtANcKHw.jpg
 http://picpaste.com/2012-01-27_21.56.28-ySpITI8d.jpg

 I think that I will try the use a relay approach, because:
 1) it is simple enough for me to do it;
 2) it does not require modifying existing laser's board;
 3) it will be fast enough for me, because I am going to use the laser
 in a a la milling fashion - move laser along the line to be burned
 instead of moving it back and forth and switching it on, when
 necessary, because:
 a) the laser is weak, so by definition it can't burn quickly;

That will cause missed or late start unless you compensate that with 
stepper or servo waiting for the relay/laser delay. Photo sensor in a 
loop (PID) could fix that but that's more work than creating a simple 
current limit circuit with LM317 and a few other components as noted 
earlier.

If all else fails, you can burn a T-shirt for yourself :-)
http://blog.craftzine.com/archive/2011/07/diana_engs_laser_lace_tops.html

 b) machine is heavy and relatively slow, compared to normal laser
 engravers, so it would not be able to handle very powerful laser
 anyway;

Not necessarily true. Not knowing how it looks like it's hard to tell. 
Remember that light can be transmitted over a long distance around the 
corners using mirrors like this one:
http://madmodder.net/index.php?topic=4895.0 as long as it's not too foggy.

 c) g-code generation - I have no idea, how to generate g-code for
 normal laser engraving, but I know, how to do it for a la milling
 style.


 BTW client was happy, when he saw the first hand-burned lines in the
 wood that I managed to do last night in that small moment, when the
 laser was working...

 Viesturs

more links:
http://www.laoslaser.org/
these guys have a roadmap something suggested on this list:
http://wiki.laoslaser.org/index.php/Roadmap

I hope we see some pictures of the final system.

-- 
Rafael

--
Try before you buy = See our experts in action!
The most comprehensive online learning library for Microsoft developers
is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro Style Apps, more. Free future releases when you subscribe now!
http://p.sf.net/sfu/learndevnow-dev2
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] DIY output driver

[Rafael Skodlar]

On 01/27/2012 11:07 AM, Viesturs Lācis wrote:
 2012/1/27 gene heskettghesk...@wdtv.com:

 That depends.  Can, if you just short your device, burn wood?  If not, or
 only very much slower than you expected, then the wire is too small.

 I could. Now it seems that my diy stopped working again - laser
 receives ~3V DC regardless of the state of gpio output pin.

 I give up trying to get it working.

 However, this gives me another better idea, that of putting your DIY
 switching device right at the laser, essentially doing away with the wiring
 losses, and use the existing small wire going to it for the logic signal to
 control it.

 That seems like the most serviceable solution to me, quit trying to send
 the amps up and down a small wire, just send the controlling signal. I am
 assuming that the relatively small currents your DIY needs can be supplied
 by the lasers own supply, removing the need to also send your DIY a pair of
 power leads of its own.

 Thanks, sounds like a pretty good idea. Its only downside - I cannot
 implement it on the spot at client's site, because I have to redo my
 diy - it is on the same piece of pcb with 7 input optoisolators for
 home, limits and e-stop.

 2012/1/27 gene heskettghesk...@wdtv.com:
 On Friday, January 27, 2012 01:50:53 PM Kirk Wallace did opine:

 I suspect the power to the laser driver just needs to be switched as an
 Enable, with the driver's TTL input modulated by PWM/PDM to turn the
 bean on and control the strength.

 I have zero experience with lasers, so grains of salt are recommended.

 And I suspect you are spot on, and that we have managed to make a larger
 problem out of it than it is.


 As I wrote - there are only 2 connection terminals for laser board,
 labeled + and -

 Viesturs

Of course that won't work with most if not all the suggestions I've seen 
so far. If I understand it correctly, your laser already has a circuit 
to drive it at reasonable current to do it's magic.

Circuit I suggested earlier was under assumption you have a bare laser 
diode connected to it. You cannot daisy chain circuits one after another 
and expect laser to work properly.

Relays are out of question IMO because they are too slow mechanical 
devices with many other drawbacks for this application. I see no reason 
to bring in solid state relays into the picture either. You are not 
driving high voltage stuff.

You either need to modify your PCB that came with the laser diode or 
build a new circuit. Maybe you can reverse engineer that PCB and use one 
spot to inject signal from the EMC side.

This link would be a good starting point to get an idea what you are 
dealing with: http://www.rog8811.com/laserdriver.htm
Suggested circuit uses same IC regulator LM317 as you mention having in 
currently included PCB I believe.

Links that might help:
http://laserboy.org/
http://repairfaq.cis.upenn.edu/sam/laserssl.htm#ssltoc

You are practically dealing with the same issue as power LEDs except 
that laser diode provides coherent light:
http://www.instructables.com/id/Power-LED-s---simplest-light-with-constant-current/

-- 
Rafael

--
Try before you buy = See our experts in action!
The most comprehensive online learning library for Microsoft developers
is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro Style Apps, more. Free future releases when you subscribe now!
http://p.sf.net/sfu/learndevnow-dev2
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] [OT]Moveing pcb-gcode generated files to the milling machine=huge PIMA

[Rafael Skodlar]

On 01/25/2012 02:44 AM, gene heskett wrote:
 On Wednesday, January 25, 2012 05:29:19 AM Rafael Skodlar did opine:

 On 01/24/2012 11:21 AM, gene heskett wrote:
 On Tuesday, January 24, 2012 01:46:15 PM Rafael Skodlar did opine:

 [...]

 Why not try (g)awk? You can search, match strings, and do some math with
 it. Of course you could always use a combination of sed, awk and bash,
 or simply perl.

 I looked at the gawk man page, didn't see any mention of floating point
 math so I kept on looking.  Bash only does integer.  Didn't see any mention
 of sed  math or floating point.


Well, there's yet another Unix thing, bc man pages sayz:
bc - An arbitrary precision calculator language

The  most  basic element in bc is the number. Numbers are arbitrary 
precision numbers. This precision is both in the integer part and the 
fractional part.  All numbers are represented internally in decimal and 
  all  computation  is  done  in  decimal.

For your amusement: man pages come with EXAMPLES. How about that?

 It turns out the easiest way is add a G92 x2.195 before the first move in
 the top of the file, and a G92.1 to clear it at the bottom.


Not good at G-code. However,
myvar=2.195
result=$(echo $myvar | awk 'CONVFMT = %2.2f{printf (2.5 * $1)}');echo 
$result
5.49

result=$(echo $myvar | awk 'CONVFMT = %2.1f{printf (2.5 * $1)}');echo 
$result
5.5

result=$(echo $myvar | awk 'CONVFMT = %2.4f{printf (2.5 * $1)}');echo 
$result
5.4875

result=$(echo $myvar | awk 'CONVFMT = %2.8f{printf (2.5 * $1)}');echo 
$result
5.4875

seem to work. Note printf.
result=$(echo $myvar | awk 'CONVFMT = %2.8f{print (2.5 * $1)}');echo 
$result
5.4875

default print ignores formating.

 But I've changed the location of the tool change, so I'm now making a
 contact gage to sit on the table to set drill lengths and will add the
 probing code after each M6.  That's a heck of a lot better than having to
 edit 24k LOC line by line. :)


24 karat? We want pictures :-)

 I have nfs working both ways now too, which means I can put pcb-gcode
 output files directly on the mill from pcb-gcode.


Cool.

 From the properties list, it looks like about 3 hours to make one board
 plus bit changes  board remounting.  Needs more spindle rpms by at least
 10x.

 Question, what ipm feeds for a 60 degree sharp pointed carbide bit, running
 about 3 thou deep, would be recommended when 2500 revs is all you have?

 Cheers, Gene

Sorry can't help you here.

-- 
Rafael

--
Keep Your Developer Skills Current with LearnDevNow!
The most comprehensive online learning library for Microsoft developers
is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro Style Apps, more. Free future releases when you subscribe now!
http://p.sf.net/sfu/learndevnow-d2d
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] DIY output driver

[Rafael Skodlar]

On 01/25/2012 01:43 PM, Viesturs Lācis wrote:
 2012/1/25 andy pughbodge...@gmail.com:
 On 25 January 2012 21:07, Viesturs Lācisviesturs.la...@gmail.com  wrote:

 Kirk, I am getting totally confused...
 What is the difference in schematics between both images in Your last
 email? I cannot see any...

 The direction the arrow points in the transistor schematic symbol.

 Ok, thanks!
 I do not understand - that is inside the BD139, right? What can I do
 about it, if it is still the same BD139 transistor???

 I finally managed to get at least something working:
 Short-circuiting +5V to 4N25's pin5 turns the laser power on and off
 as required.
 I am left with an impression that 4N25 is not working correctly.

 Is that diode next to R1 mandatory? I will not have a time to obtain
 one and solder in, before going to client.
 I just hope that I have spare 4N25...

 How sensitive to soldering heat are things like 4N25 optoisolators?
 Maybe I have accidentally burned it?

 Viesturs

I'm following this thread for a while now and believe that solution lies 
in managed current supply, not voltage. I see no current check in the 
suggested circuit.

Specialized circuits will provide much longer life for your laser, 
better max current limit, inrush current, and you can drive it with PWM 
to run cooler.
Perhaps generic LED driver circuit is the way to go:
http://www.maxim-ic.com/datasheet/index.mvp/id/4510

While low power, same rules apply:
http://www.maxim-ic.com/app-notes/index.mvp/id/242
There are probably other ICs to drive laser diodes.

Laser you mentioned is in a way another type of LED but I might be wrong 
here when it comes to driving it properly. Still, I would be more 
confident to use any current source with a limit to drive the laser 
diode than use circuits discussed earlier.

-- 
Rafael

--
Keep Your Developer Skills Current with LearnDevNow!
The most comprehensive online learning library for Microsoft developers
is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro Style Apps, more. Free future releases when you subscribe now!
http://p.sf.net/sfu/learndevnow-d2d
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] [OT]Moveing pcb-gcode generated files to the milling machine=huge PIMA

[Rafael Skodlar]

On 01/24/2012 11:21 AM, gene heskett wrote:
 On Tuesday, January 24, 2012 01:46:15 PM Rafael Skodlar did opine:

 [...]


 Chuckle, I need that this morning (morning?  Duh, it's past 2pm), the 2nd
 cup hasn't kicked in yet. :(

 I've now been searching the package repo looking for a sed-like util that
 can do the additions.  4 hours wasted and I am only down the the middle of
 the p's.  Sigh.

 Cheers, Gene

Why not try (g)awk? You can search, match strings, and do some math with 
it. Of course you could always use a combination of sed, awk and bash, 
or simply perl.

-- 
Rafael

--
Keep Your Developer Skills Current with LearnDevNow!
The most comprehensive online learning library for Microsoft developers
is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro Style Apps, more. Free future releases when you subscribe now!
http://p.sf.net/sfu/learndevnow-d2d
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] [OT]Moveing pcb-gcode generated files to the milling machine=huge PIMA

[Rafael Skodlar]

On 01/23/2012 06:31 AM, gene heskett wrote:
 On Monday, January 23, 2012 08:49:08 AM Rafael Skodlar did opine:

 There is security, and there is Pain in the Ass obnoxiousness, this is
 the latter.

 I'd be much appreciative of an idiot-proof (and I'm apparently the
 idiot)

 Don't do that too often as we might assume something :-) Your emails are
 amusing and educational also.

 Like most old fart farts, I do find the mental processes are slowing,
 particularly the short term memory aspects.  I will be the first to admit
 that the tested IQ I had at age 12, 65 years ago, 147, has faded enough
 that its starting to get my attention.  Now I have to grab the eagle book
 Ed was kind enough to print and send me and dbl-check just about everything
 I do if its been 20 minutes since I last did it.  TBH, that hurts, a lot.

 How much of that can be thrown at the number of calendars I've thrown away,
 and how much to the blood vessel damages that type 2 diabetes is doing to
 me, is probably open for debate.  I really worry when the glucose hits
 200+, but its been years since I tested under 135 2 checks in a row.

 method of send this stuff around on my home, private as I can make it,
 network.

 Thanks all.

 Cheers, Gene

 The following needs to be done as root or use sudo:

 * assume one side is the NFS server:
 This machine
 - install package autofs
 done
 - edit /etc/exports with something like
 /home/gene(rw,sync,no_subtree_check)
 done
 - restart autofs
 this worked, it wasn't running
 service nfs-kernel-server restart
 no such service on a pclos machine
 test with
 showmount -e- to get
 Export list forservername:
 /home/gene *
 [root@coyote gene]# showmount -e
 Export list for coyote.coyote.den:
 /home/gene *

 * on the workstation side:
 - install package autofs
 - edit /etc/auto.master to enable auto.net function
 /net-hosts- line that likely needs to be uncommented
 service autofs restart
 returns:
   autofs start/running, process 24811

 that will let you see (automount) servers' export under
 /net/servername
 whereservername  is your other PC. ls /net shows nothing while 'ls
 /net/servername   should show exported files in your home directory.
 Now let's test this setup:
 touch /net/servername/xxx
 ls -l /net/servername/xxx
 gene@shop:/etc$ ls /net/coyote
 home
 and:
 gene@shop:/etc$ ls /net/coyote/home/gene/eagle
 lathe-encoder  ulp
 gene@shop:/etc$ ls /net/coyote/home/gene/eagle/lathe-encoder
 eagle.epf  lathe-encoder.b#2lathe-encoder.bot.etch.tap
 lathe-encoder.pro  lathe-encoder.top.drill.tap
 lathe-encoder.b##  lathe-encoder.b#3lathe-encoder.bot.mill.tap
 lathe-encoder.s#1  lathe-encoder.top.etch.tap
 lathe-encoder.b#1  lathe-encoder.bot.drill.tap  lathe-encoder.brd
 lathe-encoder.sch  lathe-encoder.top.mill.tap
 gene@shop:/etc$


Good, you are making a progress.

 Amazingly (to me) it looks like it is working.  A first since I switched
 from amigados to linux in 1997.


There were some bugs in autofs scripts which caused a lot of grief in 
the industry for a while.

 One more test from the shop box
 gene@shop:/etc$ touch /net/coyote/home/gene/killroy-was-here
 touch: cannot touch `/net/coyote/home/gene/killroy-was-here': Permission
 denied

 So its a one way deal at best.  I can fire up mc and copy a directories
 contents from here to the shop, and just did, but I can't move anything
 back.  To get things like the drill files in sync, it needs to be a 2 way
 deal.  Is this as simple as making matching entries for exports and such on
 both machines so that each machine is both client and server?


Yes, make similar config on another side, restart NFS, and it should work.

 Thank you very much, Rafael.  Nothing beats clear  concise instructions.

 Cheers, Gene

What I have in all my Linux workstations is this:
- in konsole (my favorite GUI terminal) I open a number of tabs. First 
one is always reserved for root. I label it root and to use it I normally do
sudo su -

that gives me full root environment needed to manage packages, hard 
mount files, ISO images, etc. manually. That way I can easily remember 
what I'm doing in each tab.

- in screen utility I also create first text screen for use as root. 
Others are named by function or remote host.

If you are going to rename or change UID or GID it's best to do it in 
text terminal, i.e. not under KDE or Gnome for yourself as you'll pull 
the rug under your feet. You could create a new user with desired 
UID/GID in GUI but you'll need to make different login account name. The 
easiest IMO is to do the following:

- Assuming you are at GUI login prompt, don't login, use Ctrl-Alt-F1 to 
go to text mode terminal,

- login and become root with 'sudo su -' (ubuntu and some other distros)

- edit /etc/passwd to change UID and GID
user:x:1000:1000:user name,,,:/home/user:/bin/bash
   ^^   ^^

- run command pwconv

- edit /etc/group
user:x:1000:
   ^^
- run
chown -R user.user /home/user
to change

Re: [Emc-users] [OT]Moveing pcb-gcode generated files to the milling machine=huge PIMA

[Rafael Skodlar]

Hi Gene,

On 01/22/2012 03:18 PM, gene heskett wrote:
 Greets everybody;

 I am going slowly berzakers here with this bs of having different gid's on
 these two machines.


Didn't we discuss this some time back?

 I am at that stage where I have files ready to rename and load into emc
 (linuxcnc) to see what they look like in axis.

 But I'll be damned if I can get scp to move the files.  Something has been
 done to cifs so I cannot mount the shares defined from here, regardless of
 which of the 10,000 monkeys output I try, its 'no permission.

 This is the line in my rc.local that has been mounting that box as a

mistake number one. No need to force that manually. statement in 
/etc/fstab would be better. However, see automounter section bellow.

 read/write share at /mnt/shop, and which now fails, and was failing even
 before I built the new box for the shop/mill:

 mount -t cifs -o

mistake number two. Why the heck are you torturing Linux with protocols 
mainly used for windows?

 user=gene,passwd=gh10041934,uid=1000,forceuid,gid=1000,forcegid,noserverino
 //shop.coyote.den/shop-slash /mnt/shop


mistake number 3. Don't use /mnt as that's traditionally used for quick 
mounts like CDs, or USB thingies.

 response:
 mount error(13): Permission denied
 Refer to the mount.cifs(8) manual page (e.g. man mount.cifs)

 And of course in my man pages, error 13 is never mentioned.


Agree. Man pages are historically very poorly documented. First was lack 
of disk space, now it's a tradition not to show any examples of how to 
use some freaking exotic command.

 But it worked for years before.  From testparm's output on the shop box:
 [shop-slash]
  comment = The root directory of coyote's FC6 install
  path = /
  read only = No


that seem to be all windows stuff. Handle it Unix style and you'll be 
better off (or it should on?).

 So I have resorted to copying the files to a publicly invisible web page
 directory so that I can maybe fire up firefox out on that box and save the
 files to it that way, but dammit there ought to be a better way.

Why not simply scp files to other box? That will change ownership. You 
could use rsync -av -e ssh local_file userx@remotehost:~/some_subdir
If you copy ssh public key to the other side you won't need to enter 
password each time.

Better yet, setup NFS, see bellow.

 There is security, and there is Pain in the Ass obnoxiousness, this is the
 latter.

 I'd be much appreciative of an idiot-proof (and I'm apparently the idiot)

Don't do that too often as we might assume something :-) Your emails are 
amusing and educational also.

 method of send this stuff around on my home, private as I can make it,
 network.

 Thanks all.

 Cheers, Gene

The following needs to be done as root or use sudo:

* assume one side is the NFS server:
- install package autofs
- edit /etc/exports with something like
/home/gene(rw,sync,no_subtree_check)

- restart autofs
service nfs-kernel-server restart
test with
showmount -e- to get
Export list for servername:
/home/gene *

* on the workstation side:
- install package autofs
- edit /etc/auto.master to enable auto.net function
/net-hosts   - line that likely needs to be uncommented

that will let you see (automount) servers' export under
/net/servername
where servername is your other PC. ls /net shows nothing while 'ls 
/net/servername  should show exported files in your home directory. 
Now let's test this setup:
touch /net/servername/xxx
ls -l /net/servername/xxx

2 minutes, no public exposure, assuming both sides have the same 
distribution, (k)ubuntu in my case, and same UID,GID. If not I suggest 
you do that with changing it on one machine to match the other:
- edit /etc/passwd
- pwconv
- edit /etc/group
chown -R user.group /home/user

Note that automounter will disconnect after a timeout, 15min I believe, 
unless you have files open or you 'cd' into that space. That's normal. 
It will be visible next time you access the files.

Other option is to use sshfs which will mount directories in user space 
over ssh, see man pages.

I do make house calls on occasion :-)

Man it's noisy outside. Chinese/Vietnamese New Year. Firecrackers are 
chasing bad spirits away, I think.

-- 
Rafael

--
Try before you buy = See our experts in action!
The most comprehensive online learning library for Microsoft developers
is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro Style Apps, more. Free future releases when you subscribe now!
http://p.sf.net/sfu/learndevnow-dev2
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] LinuxCNC: The new name of the Enhanced Machine Controller

[Rafael Skodlar]

On 01/18/2012 07:30 AM, Matt Shaver wrote:
 On Wed, 18 Jan 2012 20:50:41 +1100
 Erik Christiansendva...@internode.on.net  wrote:

 I'm only having trouble pretending that their case has any
 legitimacy.

 It's not legitimate. Unfortunately, many recent activities originating
 in the U.S. are governed by a might makes right attitude rather than
 one of fairness, morality, or even basic human decency. For this I can

That's not recent but an ongoing human activity going on since Adam and 
Eve. However, it's more likely you'll win the case in the US than 
anywhere else. At least we have freedom of speech, open Internet (so 
far), and can always (?) find a lawyer for your side.

 only apologize and beg your patience while we in the U.S. try to
 recover our traditional common sense.

 Thanks,
 Matt

That will take a while but don't hold your breath. As a matter of 
principle, not all options about EMC2 have been exhausted IMO. EMCII 
(Roman numerals) is one of them. Moving domain to other country would be 
another.

So we all know, or speculate at this point, their case has no 
legitimacy, but it's impossible to tell unless we see what was presented 
to the LinuxCNC board by the Extremely Miserable Company.

Questions arise: who are they going to sue or who they were threatening 
to sue? Developers? Board members? Mailing list members? Companies that 
use or plan to use EMC2? I see no reason to keep that correspondence 
private. As a matter of fact, it should be open to the members to see 
what's in it? Or is it going to be Nancy way Pass the (health care) 
bill so we can see what's in it.

While the name to me doesn't matter as much as the well proven quality 
of the software, it's annoying to see such legal test cases or 
practice to go on as that will be used for more intimidation in the 
future. We've seen too much of that going on in the last few years.

Remember Lindows? Another victim of corporate bullying by a convicted OS 
manufacturer. Lindows was a real company 
http://en.wikipedia.org/wiki/Microsoft_vs._Lindows

I'm all for $20 mil to be paid to LinuxCNC nonprofit for further 
software/hardware development. Like the case above, EMC company should 
pay the time that people need to spend to make changes on this open 
source project. Renaming everything including IRC channels is just crazy 
and very time consuming!

My first reaction would be to fight fire with fire, i.e. have 
http://www.fsf.org/ experts look into this. If nothing else, they should 
be aware of this bullying case.

-- 
Rafael
http://www.linwin.com/rafael/get_up_stand_up.mp3
Linux Is Not WINdows . com

--
Keep Your Developer Skills Current with LearnDevNow!
The most comprehensive online learning library for Microsoft developers
is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro Style Apps, more. Free future releases when you subscribe now!
http://p.sf.net/sfu/learndevnow-d2d
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] LinuxCNC: The new name of the Enhanced Machine Controller

[Rafael Skodlar]

On 01/17/2012 08:12 PM, Erik Christiansen wrote:
 On 17.01.12 21:31, Jack Coats wrote:
 IMHO, we need to keep references to EMC and EMC2 on the web site as
 'historical artifacts', and also reference the NIST project that
 started the initial 'Enhanced Machine Controller' project and named
 it. Like in http://www.isd.mel.nist.gov/projects/rcslib/emc_links.html

 It will also allow links from around the internet and search engines
 to still logically link back to LinuxCNC in the future.

 I doubt EMC Corp would have issue with that, since it is for
 'educational and historical' purposes.

 IANAL
 It seems unimaginable that even under perverse American law, one
 corporation can own all rights to every use of an abbreviation.
 In most countries, it is sufficient for the activities of the separate
 users to be significantly distinguishable for courts to dismiss a
 litigant's claim to be the only one on the planet with the name Bob.

 If our current full name conflicted, then there would be grounds for
 discussion, but coincidence of abbreviation (EMC) is no ground for
 complaint. And both parties have adapted Albert's little equation, so
 neither can claim originality or exclusivity there.

 That said, the board's decision is the board's to make, and it avoids
 distractions other than whether or not to make any changes in the wiki.
 If we place on the home page LinuxCNC was historically known as EMC2,
 then the job's done, innit?
 /IANAL

 As always, genuine thanks for all the work done on our behalf.

 Erik


I join this opinion based on the facts that this (EMC2) project has a 
long history under its current name. It's ridiculous to see a not 
exceptionally successful company (http://finviz.com/quote.ashx?t=EMC) 
terrorize nonprofit organization especially when the product has 
absolutely nothing to do with their core business.

My feeling is that their lawyers have nothing else to do, like protect 
employees of their spinoff VMware etc.

Quick search for EMC2 on Google brings LinuxCNC.org in the third place. 
That's due to the fact that EMC Corp is advertising on the top.

Google also delivers other links that one would expect to be more 
offensive to EMC Corp lawyers than LinuxCNC open source product.

http://www.emc2.com/
http://www.emc2fusion.org/
http://emc-squared.net/
http://www.aimprogram.com/
http://www.emc-sq.com
http://www.emc2architects.com/
http://www.emctwo.com/
http://www.emcsquared.com/
http://www.emc2acne.com/
http://www.emc2interiors.com/
http://emc2payouts.com/
http://www.facebook.com/pages/EMC2/174903925885899?v=info

Other (Google) Searches related to emc2:
emc2 cnc software
linux emc2
e=mc2
emc2 wiki
emc2 engineering
emc2 fusion
emc2 architects
emc2 bikes

so EMC2 CNC software comes up twice in this section alone. Fact: search 
for emc2 returns more links related to LinuxCNC EMC2 on the first 3 
pages than EMC company.

I respect LinuxCNC board decision but wonder about EMC legal department. 
My guess is that generic search for their products turns up little or 
nothing and they want more free advertising.

http://www.emc.com/about/investor-relations/faqs.htm
Q: What does EMC stand for?
A: The founders of EMC are Richard Egan and Roger Marino, the E and 
M behind the naming of EMC. EMC Corporation is the Company's full name.

So they were sitting on the fact, and did nothing, all these years that 
a nonprofit org openly used EMC2 (Enhanced Machine Controller 2) for the 
project name. There is statue of limitation somewhere, even in the US 
law one would think. Perhaps they need some http://www.emcpic.info

Imagine how many links are going to be broken if LinuxCNC wiki page 
changes every instance of EMC2? That's censorship IMO and I'm against it.

-- 
Rafael

--
Keep Your Developer Skills Current with LearnDevNow!
The most comprehensive online learning library for Microsoft developers
is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro Style Apps, more. Free future releases when you subscribe now!
http://p.sf.net/sfu/learndevnow-d2d
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Live CD

[Rafael Skodlar]

On 12/28/2011 09:06 AM, Tim James wrote:
 I have downloaded and burnt the Live CD 10.04.

 It has worked on one desktop I tried but not on another or my
 laptop.

 I only have the laptop with me at the moment so I would like to
 install Ubuntu and EMC to find out more in conjunction with reading
 all the EMC manuals I have printed out.

 The laptop is a Panansonic Toughbook CF29,Windows XP,Pentium 4,756Mb
 ram.

what's more important to know is which chipset video and CPU are in there.


 From the users emails over the last few days I tried the
 ctrl/alt/f1and have experimented with the f6 boot options.

 The laptop appears to boot to a certain point and then the screen
 goes blank and then 'hangs' with the power off button being the only
 option.

 If I delete splash-- and replace it with vga=771 then the display
 remains on  and the laptop 'hangs' on pulseaudio configured for
 per-user sessions line.

 I have tried some of the other f6 suggestions but not all as yet (I
 will wear the power button out soon).

 I don't know if I can add the options  sequentially to the boot line
 or only use one at a time.

 Does anybody have any suggestions rather than me working through all
 the f6 options without really knowing which maybe causing the boot
 problem ?

 Tim

First try to see if you can install it in text mode only. Next would be 
Ubuntu 11.x to see if it works better or you could try perhaps kubuntu 
version. If not then try a different distribution altogether. I would 
try debian as ubuntu is based on it.

One other option is to try virtual installation inside XP to see if that 
works. RT will not work of course, but you can still test basic EMC 
functionality. Installing virtualbox from Oracle takes only a few minutes.

-- 
Rafael

--
Ridiculously easy VDI. With Citrix VDI-in-a-Box, you don't need a complex
infrastructure or vast IT resources to deliver seamless, secure access to
virtual desktops. With this all-in-one solution, easily deploy virtual 
desktops for less than the cost of PCs and save 60% on VDI infrastructure 
costs. Try it free! http://p.sf.net/sfu/Citrix-VDIinabox
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] fail2ban default setup gotcha

[Rafael Skodlar]

On 12/23/2011 08:08 PM, gene heskett wrote:
 On Friday, December 23, 2011 10:18:29 PM Jon Elson did opine:

 gene heskett wrote:
 That has been done long ago Mark.  The problem is that on pclos (this
 box) gene is the first user, with a userid of 500.  On ubuntu, gene
 is also the first user 1000, so when user 500 tries to copy a file to
 /home/user=1000 on ubuntu, its 100% no permissions.

 Now if the copy utilities used the username, and it was the same $name
 on both machines, there is no clash.

 Cheers, Gene

 You should be able to create an alternate user (like gene2) and then
 create a group that allows
 access to both the 500 and 1000 users.  I may have missed the start of
 this thread, I'm guessing
 this is a problem with a NFS file system?  Seems like that would be the
 only time such
 cross-system IDs would matter.

 I have the *buntu box mounted at /mnt/shop, a samba share I believe.

 From mounts output:
 //shop.coyote.den/shop-slash on /mnt/shop type cifs (rw,mand)

 What I would like to be able to do, and which requires scp or sftp to do,
 is fire up mc, send one pane to the *buntu box, the other to wherever I
 have downloaded an emc useful file to here on this box, and just hit an F5
 to copy or an F6 to move it.  I fail to see why such a simple operation,
 where I am the user gene on both machines, has to be such a %$#@()^ pain
 in the ass.


There is a number of ways to fix your problem but the following is 
likely the easiest way to do it. My setup: kubuntu workstation with 
openbox virtual machine(s). For test purposes I created a different user 
in one of VMs and then used the following method:
* workstation
   - install sshfs
   - create a directory; for example ~/tmp/vm01
   - run the following commands:
VM=vm11   or you could use IP# 192.168.3.185
USR_REMOTE=rafaelx
sshfs $USER@VM:/home/${USR_REMOTE} tmp/vm01

* VM (virtual machine) or other Unix system
   - enable ssh connection, possibly use auto login with ssh key

Now I can copy files or dirs back and forth using cp, rsync, mc or 
whatever on my workstation side. I tried it both ways and the files 
changed ownership as expected so that I have right ownership on either 
side. No need to mess with passwd file or anything else.

If you want gui, install krusader which has the same functionality as mc 
with a lot of excellent candy! krusader (from KDE) is standalone and 
does not need sshfs I believe.

I employ these three methods securely between the systems on LANs and 
the Internet: Linux, BSD, Solaris.

 Why can't there be an option in these file management utility's to tell
 them, not to use the user number for the perms checking, but the user name
 instead?  All this bs would disappear in a puff of invisible smoke instead
 of all the blue smoke I generate because it takes me 10 minutes to reread
 the manpages several times, and likely 20 tries to get the proper command
 line syntax constructed from the totally obtuse man pages of scp and sftp.


It's not BS but I agree with what you say about the man pages. Too many 
man pages suck because they don't give you any examples of how to use 
the command. Old Unix problem.

Still, the way things are is important for security reasons. It keeps 
improving for the most part but you cannot make too drastic changes as 
that breaks too many home grown utilities in large installations.

What you could do is to setup a user on one system to be in the same 
group as the user in the other system and/or vice versa. In addition, 
you would need to change umask (002) to have users create group writable 
directories and files.

 Could this be such a matter as security=user in the cifs.conf files on
 both machines?  On checking, that option is set on this box.  And now is
 set on shop.coyote.den too, it was share before on that machine.

Why bother with mosquito carrying viruses as it's inherently insecure 
and messy when you can fly in fortress? While samba can provide 
ownership change for the files when you copy them between the systems, 
it's something I will NEVER use between Unix systems when NFS is superior!

You can setup automounter which will let you mount directories from any 
system with NFS. Check /etc/auto.* files. After it's setup, you can use 
autofs as a regular user, no root intervention needed. For example:
in /etc/auto.master  enable
/net-hosts

Sometimes you need to change /etc/auto.net because some implementations 
were broken in the past.

/etc/exports   --- file tells what to export.
/home/rafaelx  192.168.3.0/24(rw,sync)

Put IP# and hostname in /etc/hosts. Restart NFS server daemon after you 
make changes

Use:
On workstation
ls /net/hostname
will give you names of directories exported by hostname. You can then 
do whatever depending on the permissions.

Install autofs on the client side and nfs-kernel-server on the serving 
side. You could do the same on both sides if you have enough resources 
and want to play with it.

ls /net/vm01
shows what's exported on that 

Re: [Emc-users] Next distribution after Ubuntu 10.04 LTS is no longer supported?

[Rafael Skodlar]

On 11/23/2011 07:52 PM, Jack Coats wrote:
 I'm not saying that the 10.04 is bloated, but some of the folks that
 have moved to 11.04 or 11.10 locally have noted several 'broke as
 designed' items and have complained about bloat in the newer versions
 as well as the new Ubiquity user interface.  I am guessing 12.04 will
 be the next LTS version.


Agree.

Kubuntu 11.04 upgraded to 11.10 on my home workstation is giving me a 
hell of a time I haven't seen in years. Nvidia drivers especially are 
unpredictable. Some utilities like Amarok drive CPU use to 103% 
according to top.

Stable Debian or minimal Ubuntu with XFCE4 would be more appropriate for 
EMC and better fit for smaller embedded systems IMO. I looked at XFCE4 
recently and find it provides good enough environment for most GUI needs 
with smaller footprint than Gnome or KDE.

 I am not complaining, just wondering what our outlook for the future
 is and if there are any suggested changes.  Trying to be a bit
 proactive and forward thinking.


Agree.

 Historically I have seen software enlarge to meet the capability of
 the hardware it runs on.  Using its features is OK, but leaving some
 of the capability for other software to use (like EMC2, CAD/CAM
 packages, etc) rather than trying to soak it all up in the OS and
 support software is good too.

 ... Jack


There is one more issue at this time in history, transition from 32bit 
to 64bit. Mix of 64bit and 32bit libraries and other code is causing all 
kinds of strange problems if you run some Java or other apps that were 
written for 32 bit release.

-- 
Rafael

--
All the data continuously generated in your IT infrastructure 
contains a definitive record of customers, application performance, 
security threats, fraudulent activity, and more. Splunk takes this 
data and makes sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-novd2d
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Dust-proof case?

[Rafael Skodlar]

On 11/06/2011 10:37 AM, Viesturs Lācis wrote:
 2011/11/6 Slavko Kocjancicesla...@gmail.com:

 I agree that filtering is good thing to do. But in this case (running
 near 1 hour without problem) the EMI isnt the source of trouble.

 Thanks, guys, for suggestions!
 I will try to figure out, how to test, if PC is ok and if the
 spindles/VFD does mess the thing for the starting point, because I
 also think that the tendency for machine to work correctly some time
 after turning it makes the whole situation frustrating.

 Viesturs

Just reading this thread takes longer than troubleshooting should take. 
Lots of good suggestions, with a number of repeats, but there's one 
thing that I don't remember reading: cost of troubleshooting.

It's one thing to troubleshoot a hobby machine, it's another to do it on 
a production system. I cannot imagine running test programs on 
production PC for hours to troubleshoot memory or anything else for that 
matter. Correct me if I'm wrong, but production loss in time, ruined 
parts, and delivery in one day is likely (times) more than the cost of 
whole PC and possibly other electronic components comprising the CNC system.

Question: why is it that the end user does not have spare kit(s) (whole 
PC or a motherboard, memory, disk drive, sensors, etc.) on site? And if 
not, why is the supplier not providing spares to replace them during 
first troubleshooting session? Spares are cheap compared to production 
loss especially when travel is involved!

When you have intermittent problems that cannot be found in an hour by 
troubleshooting means suggested in this thread, it's time to start 
changing parts depending on their (statistical or experienced) 
possibility of failure, ease of access to the components, or complexity 
of changing them. That way you take care of two problems: bad component 
or subassembly, and intermittent connections between those components.
You have to be careful not to introduce new problems of course.

It might turn out that the replaced component was not bad at all and it 
could be used as a spare or for troubleshooting purposes next time. I 
used to test field replaced boards back in office on test computers. 
Good boards were labeled and returned to our stock. That 80's practice 
should work today too.

One possibility would be to connect a second computer without removing 
the original assuming cabling could be taken care of. And it should if 
designed properly. Also, a number of messages addressed interference and 
noise issues. What hasn't been mentioned is how to measure this. My 
suggestion would be to bring an oscilloscope (I depend on my old 
Tektronix 2445 and little Velleman Personal Scope) on-site to see the 
quality of electric signals on sensor and power lines, power supplies, 
and PCB boards. You can easily buy cheap digital scopes with sufficient 
capabilities for field service these days.

Good DMM also needs to be part of tech or engineers tool box. As the 
saying goes: show me your tool box and I'll tell you how good an 
engineer or technician you are.

Scope will tell you if you really need a line filter or not. Just 
throwing new components into the system is likely a waste of time and $ 
and might introduce new problems as well.

Anxious to read the rest of the story,

-- 
Rafael

--
RSA(R) Conference 2012
Save $700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Dust-proof case?

[Rafael Skodlar]

On 11/04/2011 08:32 AM, Viesturs Lācis wrote:
 2011/11/4 Les Newellles.new...@fastmail.co.uk:
 In general wood dust does not affect computers that much. I have seen
 machines with piles of dust inside, still working fine day in day out.
 Heat is an issue so do as much as possible to keep it cool. Possibly fit
 an over sized CPU cooler.

 The thing is that machine works correctly for 40-50 mins, then
 malfunctions begin to appear - it does not run a file, when button is
 pressed and some other. Or it ruins a part, just as if motor had lost
 steps, but the trick is that next part if good without rehoming or any
 other activity taken - just change material and run the file.

If the OS is not crashing then your problems are elsewhere as others 
have suggested. Depending on your motherboard, it's components 
temperature can be monitored with utilities like digitemp fancontrol 
hddtemp lm-sensors. I cannot tell how well they work with RT kernel and 
EMC as I haven't tried it.

I put a thermometer with memory for high and low records in areas of 
interest many times. It helped me relate server crashes to environment 
temperatures. You might put such a thermometer in the control box for 24 
hours to see what the extremes are.

Temperature logger would be even better. You may want to check in 
Arduino world for such a sensor/shield solution.

 And the more machine is running the more often these malfunctions
 appear. That tells me that some parts in machine are overheating and
 thus not functioning properly. I would like to blame the dust that
 gets in all the narrow places on the boards and disturb normal heat
 dissipation.


It's possible you have a poor connection somewhere. Connectors are metal 
which expands and contracts with a temperature and also act as unwanted 
vibration sensors. Check the connections first.

It's also possible you have a component with a cold solder joint which 
also reacts to temperature and vibration. That would explain why your 
CNC system resumes working without errors at some point.

Is it possible that the problem appears when the machine (tool, router 
etc.) is in a particular position? That could indicate a broken cable, 
more likely on older machines.

To troubleshot such problems I use a screwdriver handle or some other 
insulated material and tap on different components or wiggle the wires 
to see if that triggers a problem.

You might want to check screws that hold wires in connectors or terminal 
blocks. They tend to get lose or corrode sometimes.

While wood is not causing electric problems most of the time, it's dust 
could bring moisture into the box that would cause connection problems 
or corrosion.

And let's not forget grounding again. Make sure all is well grounded all 
the time.


 If you want to enclose the computer in a box, fit a fan on the side
 pushing air into the box. In other words the box becomes slightly
 pressurized. That stops dust from creeping in through cracks or door
 seals. Filter the air coming in with an air filter from an older car,
 the type that is just a big round ring that sits in a frying pan shaped
 housing. Clamp it to the side of your box with a round disc of wood or
 metal. You will have to replace the filter fairly regularly but they are
 quite cheap.

 That is why I would like to put them in totally sealed box.
 In that case I could implement water cooling for PC components.

I like this idea but you need to monitor/check plumbing every once in a 
while which in my experience almost never happens. All equipment I ever 
supported in industrial environment come with instructions for regular 
maintenance. Some even include nice log books or sheets to keep track of 
it. I've seen empty pages too many times so that needs to be taken into 
consideration.

Another option is to use heat-pipes which are frequently used to cool 
components inside sealed boxes in industrial and military systems.
Search for DIY heat-pipes

Heat-pipes are not going to cool every component of course, just those 
that touch it directly. Trouble is that they need to be custom made most 
of the time.

Let me give you one link http://www.silentmods.com/section2/item230/

 What should I do about monitor?


Modern industrial monitors work well in wide temp range but cost more. 
Use either that or check how they solve this problem as others have 
suggested. Google heat exchanger design for tons of links.

 Many bigger commercial machines use air to air heat exchangers so the
 clean air inside the box and dirty air outside never mix.

 Any idea, how to do that in a cost-effective way?

 Viesturs

See above.

-- 
Rafael

--
RSA(R) Conference 2012
Save $700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] hello people NEED your help

[Rafael Skodlar]

On 09/22/2011 01:47 PM, dave wrote:
 AFIK Linux/Unix does not accept spaces in a filename.

All Unix systems accept spaces and other (strange) characters in file 
names with few exceptions. Here's one way to take care of it:
rename 's/ /_/g' file\ name\ with\ spaces.xxx
Use Tab to expand the file name automatically as the shell will insert 
backslashes for you.

The main reason we have issues with spaces in file names is that shell 
is using spaces to separate things on the command line. It interprets 
words as different options or separate file names given to commands. In 
addition, a number of utilities and parsers use space as the default 
separator so spaces are not good idea.

Spaces in file names are as ugly as top posting in email.

 Note that emc thinks you have a file named intento and not the whole
 filename including the .ini
 Eliminating the spaces or joining with hyphen or underscore should solve
 the problem.

 HTH

 Dave
 On Thu, 2011-09-22 at 15:18 -0500, Stuart Stevenson wrote:
 I would change the directory and file names.
 I would remove the spaces.


-- 
Rafael

--
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2dcopy2
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Parallel port card that just works?

[Rafael Skodlar]

On 09/13/2011 11:00 AM, Jon Elson wrote:

 On Tuesday, September 13, 2011 12:04:09 PM doug metzler did opine:


 I'll ship the machine to WV with a $50 bill taped to the top for beer
 acquisition :-)

 How do you know the card doesn't work?  You mentioned Ubuntu, did you
 try printing
 with it?  EMC doesn't use any Ubuntu drivers when using a parallel port
 card for controlling
 steppers, so it is NOT Ubuntu's fault.


While not using the default drivers, it's possible that

 Finding the CORRECT I/O port address for these cards is often a guessing
 game, as the

Not necessarily. Linux commands will help you with this. Here's how I 
would approach this problem:
- use commands
dmesg | less
or
dmesg |grep parport

which in my case results in

[8.181311] parport_pc 00:09: reported by Plug and Play ACPI
[8.181369] parport0: PC-style at 0x378 (0x778), irq 7, dma 3 
[PCSPP,TRISTATE,COMPAT,ECP,DMA]
[8.256909] lp0: using parport0 (interrupt-driven).

This tells me that my parallel port uses physical address 0x378, 
interrupt 7, and DMA channel 3. It also tells me what the BIOS setup for 
parallel port is. The same information ends up in /var/log/syslog I believe.

Let me mention the following link on this list again:
http://www.epanorama.net/circuits/parallel_output.html

it might help you actively troubleshoot the problem faster than have any 
of us drive or fly to Seattle area. Circuit to test the pport on that 
page is simple enough to put it together in a few minutes. Compiling 
related code should be trivial also.

 chips have 2 parallel and 2 serial ports on them, the difference between
 the options is
 which ports are actually brought out to connectors.  The addresses
 change depending
 on what other cards are in the system, so the manufacturer can't give
 static numbers,
 either.

 Jon

Doug, you may want to send us related segments of dmesg or 
/var/log/kern.log.

We'll collect beer later.

-- 
Rafael

--
BlackBerryreg; DevCon Americas, Oct. 18-20, San Francisco, CA
Learn about the latest advances in developing for the 
BlackBerryreg; mobile platform with sessions, labs  more.
See new tools and technologies. Register for BlackBerryreg; DevCon today!
http://p.sf.net/sfu/rim-devcon-copy1 
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] D510 EPP problem

[Rafael Skodlar]

On 08/27/2011 09:15 AM, Jon Elson wrote:
 Viesturs Lācis wrote:
 So what can I do now? Peter mentioned some low level poking. How can I
 do that and what tools do I need for that?


The following page might be of some help to troubleshoot parallel port 
issues: http://www.epanorama.net/circuits/parallel_output.html


 And I would appreciate, if any user would share their experience
 suggestions for EPP-capable PCI add-in card.

 You should be able to run an EMC config for software stepper output and
 see if it will
 even start up. If it does, then use a voltmeter, earphone with series
 resistor or, if you
 have one, an oscilloscope, to see if the step and direction pins work as
 expected.
 Some of the reports you have shown seem to indicate the port is not
 working at all
 (where all registers read as 255). I'm thinking that the BIOS may have
 disabled the
 port, but it could also be that the parport chip has gone bad.

 Jon

--
Rafael

--
EMC VNX: the world's simplest storage, starting under $10K
The only unified storage solution that offers unified management 
Up to 160% more powerful than alternatives and 25% more efficient. 
Guaranteed. http://p.sf.net/sfu/emc-vnx-dev2dev
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] A possible new gotcha for emc users.

[Rafael Skodlar]

On 04/13/2011 07:52 PM, Przemek Klosowski wrote:
 On Wed, Apr 13, 2011 at 12:52 AM, Rafael Skodlarra...@linwin.com  wrote:

 /boot was added because of crappy BIOS that was not able to handle
 cylinders beyond 1024 years ago. That's not needed anymore and makes no
 sense either. What good is it booting kernel from /boot and then fail to
 access core utilities (fdisk, fsck, df, etc.) on another partition to
 fix the system.

 /boot never contains end-user executables. Besides the 1024 cyl

where did you see such a claim?

 limitation, another reason was multibooting: you could keep boot files
 for several Linux/Unix and Windows systems in one place; that's why
 /boot usually uses the FAT filesystem, which is understood by all
 systems without having to install foreign filesystem modules.

Not true. You do not need /boot to boot any of multiple OSes installed 
on the same system. Grub or LILO for that matter handled that just fine. 
It was the BIOS limitation that required to have boot stuff under the 
cylinder number 1024. Even that was bogus in a way as the firmware on 
hard drives translates cyl, hd, sec/track for the BIOS default values. 
I rarely setup /boot partition in last 15+ years on PCs.

Special boot partition was never a requirement on most (DEC, HP) 
mainframe or Unix minicomputers to boot up either.

--
Rafael

--
Benefiting from Server Virtualization: Beyond Initial Workload 
Consolidation -- Increasing the use of server virtualization is a top
priority.Virtualization can reduce costs, simplify management, and improve 
application availability and disaster protection. Learn more about boosting 
the value of server virtualization. http://p.sf.net/sfu/vmware-sfdev2dev
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users