Re: [Emc-users] air pressure sensors for up to 10 psi

[Danny Miller]

Here ya go, choose whatever pressure range you want.  But here's the 
thing: most technical parts are now using BPT, aka "G"/DNx threads (not 
to be confused with GHT "garden hose thread").  This is G1/4 aka DN8.  
All the USA hardware store stuff is NPT (MPT,FPT) and they have the same 
diameter and thread pitch but G thread profile does NOT seal with NPT, 
even with thread tape. Well, TBH we've sealed them before with like 10 
attempts with way too much thread tape but in general no.  Of course G 
seals just fine with G, with thread tape.


I just bit the bullet and started building everything with G threads 
until I have no choice but to switch to NPT.


https://es.aliexpress.com/item/2255800789691504.html?spm=a2g0o.order_list.0.0.5430194d4pWxuF=glo2esp4itemAdapt

On 5/28/2022 10:40 AM, gene heskett wrote:

Greetings all;

I'm getting frustrated, but I know it has to be out there, someplace.
I need a differntial air pressure sensor, differential so it references
local barometric as its reference zero, and up to 10 psi, with a 3 wire
analog output.
I've solved the control aspect, but not how to detect a pressure to
accomplish the control.

automotive EGR types seem seem the most promising as they are affordable
and 2 port gismo's but all I can find just lists the vehicles it fits and
works with, but zero info on how to hook it up and get an analog reading
out of it.  "MAP" sensors whose internal reference is a good vacumm can
have their signals clipped and massaged into something usable, but so far
I've bought 4 w/o finding a working one that actually reads the 14.7 psia
of normal air.

So how are you folks solving that problem?

Or are you burning up the shop compressor feeding your misters? I can't
see that as being an acceptible solution since none of those have an
intake muffler so you can still hear yourself think.

What I have cobbled up at the moment is giving me more than enough air to
run a mister, and I can't hear it running over the 2 fans in this 6040
mill. One in the vfd is really noisy.

Thank you.

Cheers, Gene Heskett.



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

Re: [Emc-users] Can't compile MODBUS VFD in 2.8.2

[Danny Miller]

OK, I tried putty with 2 & 3 tied together and it loops back and shows 
keystrokes while tied.  So the NUC's RS232 port IS functional.


I found an FTDI USB RS232 but it ends in female pins, I need male.

Found a knockoff PL2303 USB-to-DB9 cable, I hope it's RS232 level, it 
does end in a DB9.  No luck there either.  Same "Modbus timed out" error.


There is an isolated RS232-Modbus bridge there in between.  I know it 
works because it's fine with the existing Dell and 2.7.  I did speculate 
if the electrical signal levels are somehow different it might not work 
on the new machine, but that doesn't make much sense.


I guess I need to bring the oscilloscope out- if the baud rate is wrong, 
that would do it, but I don't see how 2.8 would change that.  Same 
code.  But I'm running out of other things to check.


Danny

On 4/10/2022 1:19 PM, andy pugh wrote:

On Sun, 10 Apr 2022 at 18:29, Danny Miller  wrote:

Nothing to unplug- it's a built-in RS232 DB9.  Nothing enumerates across it.

https://superuser.com/questions/131044/how-do-i-know-which-dev-ttys-is-my-serial-port

suggests "cat /proc/tty/driver/serial" will tell you which ttyS* has
hardware in it.




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

Re: [Emc-users] Can't compile MODBUS VFD in 2.8.2

[Danny Miller]

Nothing to unplug- it's a built-in RS232 DB9.  Nothing enumerates across it.

I'm baffled where to go.  I see there's a port /ttyS0 from the command line.

Launching LinuxCNC from the command line shows text warnings about 
Modbus timing out.  It's just not communicating.  I played with the 
loadusr command and recompiled the code with /ttyS0 as the new default 
if the command line option didn't take somehow.


I'm pretty sure the x200 code and libmodbus are not the prob- something 
about config?  I don't know where to look.  I suppose the RS232 port 
itself could physically be busted, I could look at it with a logic 
anaylzer.  Be nice to see if anything was trying to come out.  I don't 
think it could be putting out valid traffic, if it did it I think the 
VFD would turn on.


Hmm, it is going through an isolated RS232-MODBUS bridge.  But that and 
the cable and VFD are known working.  I put the cable back on the other 
machine and it works.


I guess I could install a putty terminal and see if I can send UART 
traffic in and out to verify the port is working.


Maybe I could give up on the built-in RS232 and try a USB FTDI to RS232?

Danny

On 4/9/2022 3:16 PM, andy pugh wrote:

On Sat, 9 Apr 2022 at 16:44, Danny Miller  wrote:


I tried specifying the port loadusr x200 device="/dev/ttyS0". I'm not
sure how my serial ports are named.  Would "/dev/ttyUSB0" be if that was
an FTDI USB to serial device?

Plug and unplug it, see what appears and disappears in /dev/




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

Re: [Emc-users] Can't compile MODBUS VFD in 2.8.2

[Danny Miller]

Ah, found it:

https://forum.linuxcnc.org/25-classicladder/33130-hitachi-wj200-setup

That worked, cut and pasted the gcc lines.

LinuxCNC boots and loads it.  But no modbus communication with VFD.

I've been using a Dell with a built-in RS232 successfully before.

This is on a new fanless NUC with a built-in DB9 RS232 serial port.

I tried specifying the port loadusr x200 device="/dev/ttyS0". I'm not 
sure how my serial ports are named.  Would "/dev/ttyUSB0" be if that was 
an FTDI USB to serial device?


Not sure where to go from here.

Danny

On 4/8/2022 1:42 PM, gene heskett wrote:

On Friday, 8 April 2022 14:26:21 EDT Danny Miller wrote:

I used "sudo apt-get install" and libmodbus5 was already there.

halcompile --install x200_vfd.comp does:

gcc -I/usr/include -I/usr/include/linuxcnc -URTAPI -U__MODULE__ -DULPAI
-Os -o x200_vfd /tmp/tmp4bdQIY/x200_vfd.c -Wl,-rpath,/lib -L/lib
-llinxcnchal

Then, no error about missing typedef (and function prototypes)- which
would have been a lack of a header file.  It's got the header file, now
"undefined reference" for the functions.  So the header prototyped the
function now, but it's not seeing the actual compiled code
"modbus_write_registers" etc

Danny


So I'm lost. Sorry for the noise.

On 4/8/2022 12:25 PM, gene heskett wrote:

On Friday, 8 April 2022 13:06:57 EDT Danny Miller wrote:

So, I think libmodbus-dev is correct,  the typedef and function
names
match.

I installed it, now the file in  /usr/include/modbus/modbus.h should
work, and it looks like halcompile definitely uses it when given an
explicit path, since the error message changes.

That directory just has the header file, it looks like the error is
that it can't actually find the compiled definitions of the
functions
called out in the header.

Danny

That sounds like you do not have libmodbus5 installed. And that
puzzles me a bit because libmodbus-dev should have pulled it in as a
dependency.

What utility are you useing to install this stuff?


On 4/7/2022 9:01 PM, Danny Miller wrote:

Thanks Andy!  Closer...  it's in /usr/include/modbus/

I specified that.  Now it doesn't fail on a missing type modbus_t,
but
rather "undefined reference" to any modbus_write_bit etc functions.
So even though I did get that install for libmodbus to work, it
just
has the header file.  It's not able to find the definitions in a
compiled file.

More ideas?

Danny

On 4/7/2022 5:23 AM, andy pugh wrote:

On Thu, 7 Apr 2022 at 03:27, Danny Miller 

wrote:

And same prob.  It's looking for stuff in the libmodbus's
modbus.h
but
it's not seeing it somehow.

You could try giving it a complete path:

#include "/usr/share/modbus/modbus.h"

(quotes rather than braces, I think, for an explicit path)

Or, maybe

#include 

As libmodbus installs into a directory.

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

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

Cheers, Gene Heskett.

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


Cheers, Gene Heskett.



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

Re: [Emc-users] Can't compile MODBUS VFD in 2.8.2

[Danny Miller]

I used "sudo apt-get install" and libmodbus5 was already there.

halcompile --install x200_vfd.comp does:

gcc -I/usr/include -I/usr/include/linuxcnc -URTAPI -U__MODULE__ -DULPAI 
-Os -o x200_vfd /tmp/tmp4bdQIY/x200_vfd.c -Wl,-rpath,/lib -L/lib 
-llinxcnchal


Then, no error about missing typedef (and function prototypes)- which 
would have been a lack of a header file.  It's got the header file, now 
"undefined reference" for the functions.  So the header prototyped the 
function now, but it's not seeing the actual compiled code 
"modbus_write_registers" etc


Danny

On 4/8/2022 12:25 PM, gene heskett wrote:

On Friday, 8 April 2022 13:06:57 EDT Danny Miller wrote:

So, I think libmodbus-dev is correct,  the typedef and function names
match.

I installed it, now the file in  /usr/include/modbus/modbus.h should
work, and it looks like halcompile definitely uses it when given an
explicit path, since the error message changes.

That directory just has the header file, it looks like the error is
that it can't actually find the compiled definitions of the functions
called out in the header.

Danny

That sounds like you do not have libmodbus5 installed. And that puzzles
me a bit because libmodbus-dev should have pulled it in as a dependency.

What utility are you useing to install this stuff?


On 4/7/2022 9:01 PM, Danny Miller wrote:

Thanks Andy!  Closer...  it's in /usr/include/modbus/

I specified that.  Now it doesn't fail on a missing type modbus_t,
but
rather "undefined reference" to any modbus_write_bit etc functions.
So even though I did get that install for libmodbus to work, it just
has the header file.  It's not able to find the definitions in a
compiled file.

More ideas?

Danny

On 4/7/2022 5:23 AM, andy pugh wrote:

On Thu, 7 Apr 2022 at 03:27, Danny Miller 

wrote:

And same prob.  It's looking for stuff in the libmodbus's modbus.h
but
it's not seeing it somehow.

You could try giving it a complete path:

#include "/usr/share/modbus/modbus.h"

(quotes rather than braces, I think, for an explicit path)

Or, maybe

#include 

As libmodbus installs into a directory.

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

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


Cheers, Gene Heskett.



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

Re: [Emc-users] Can't compile MODBUS VFD in 2.8.2

[Danny Miller]

So, I think libmodbus-dev is correct,  the typedef and function names match.

I installed it, now the file in  /usr/include/modbus/modbus.h should 
work, and it looks like halcompile definitely uses it when given an 
explicit path, since the error message changes.


That directory just has the header file, it looks like the error is that 
it can't actually find the compiled definitions of the functions called 
out in the header.


Danny

On 4/7/2022 9:01 PM, Danny Miller wrote:

Thanks Andy!  Closer...  it's in /usr/include/modbus/

I specified that.  Now it doesn't fail on a missing type modbus_t, but 
rather "undefined reference" to any modbus_write_bit etc functions.  
So even though I did get that install for libmodbus to work, it just 
has the header file.  It's not able to find the definitions in a 
compiled file.


More ideas?

Danny

On 4/7/2022 5:23 AM, andy pugh wrote:

On Thu, 7 Apr 2022 at 03:27, Danny Miller  wrote:


And same prob.  It's looking for stuff in the libmodbus's modbus.h but
it's not seeing it somehow.

You could try giving it a complete path:

#include "/usr/share/modbus/modbus.h"

(quotes rather than braces, I think, for an explicit path)

Or, maybe

#include 

As libmodbus installs into a directory.





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



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

Re: [Emc-users] Can't compile MODBUS VFD in 2.8.2

[Danny Miller]

Thanks Andy!  Closer...  it's in /usr/include/modbus/

I specified that.  Now it doesn't fail on a missing type modbus_t, but 
rather "undefined reference" to any modbus_write_bit etc functions.  So 
even though I did get that install for libmodbus to work, it just has 
the header file.  It's not able to find the definitions in a compiled file.


More ideas?

Danny

On 4/7/2022 5:23 AM, andy pugh wrote:

On Thu, 7 Apr 2022 at 03:27, Danny Miller  wrote:


And same prob.  It's looking for stuff in the libmodbus's modbus.h but
it's not seeing it somehow.

You could try giving it a complete path:

#include "/usr/share/modbus/modbus.h"

(quotes rather than braces, I think, for an explicit path)

Or, maybe

#include 

As libmodbus installs into a directory.





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

Re: [Emc-users] Can't compile MODBUS VFD in 2.8.2

[Danny Miller]

Nevermind, fixed the "update" part and libmodbus installed.

Went back and tried sudo halcompile --install x200_vfd.comp

And same prob.  It's looking for stuff in the libmodbus's modbus.h but 
it's not seeing it somehow.


Danny

On 4/6/2022 7:31 PM, Danny Miller wrote:

OK I get:

Err:1 http://deb.debian.org/debian buster/main amd64 libmodbus-dev 
amd64 3.1.4-2 404 Not Found [IP: 151.101.50.132 80]


E: Failed to fetch 
http://deb.debian.org/debian/pool/main/libm/libmodbus/libmodbus-dev_3.1.3-2_amd64.deb 
404 Not Found [IP: 151.101.50.132 80]


E: Unable to fetch some archives, maybe run apt-get update or try with 
--fix-missing?


Danny

On 4/6/2022 9:29 AM, andy pugh wrote:

On Wed, 6 Apr 2022 at 03:09, Danny Miller  wrote:


https://github.com/stephane/libmodbus IS the correct one.

wj_200_vfd.comp is a "comp" file, which means that it should compile
and install with just "sudo halcompile --install filename.comp"

it contains "#include "  which will be looking in
/usr/include on your system.

I think that what you probably need to do is install libmodbus:

sudo apt-get install libmodbus-dev




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



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

Re: [Emc-users] Can't compile MODBUS VFD in 2.8.2

[Danny Miller]

OK I get:

Err:1 http://deb.debian.org/debian buster/main amd64 libmodbus-dev amd64 
3.1.4-2 404 Not Found [IP: 151.101.50.132 80]


E: Failed to fetch 
http://deb.debian.org/debian/pool/main/libm/libmodbus/libmodbus-dev_3.1.3-2_amd64.deb 
404 Not Found [IP: 151.101.50.132 80]


E: Unable to fetch some archives, maybe run apt-get update or try with 
--fix-missing?


Danny

On 4/6/2022 9:29 AM, andy pugh wrote:

On Wed, 6 Apr 2022 at 03:09, Danny Miller  wrote:


https://github.com/stephane/libmodbus IS the correct one.

wj_200_vfd.comp is a "comp" file, which means that it should compile
and install with just "sudo halcompile --install filename.comp"

it contains "#include "  which will be looking in
/usr/include on your system.

I think that what you probably need to do is install libmodbus:

sudo apt-get install libmodbus-dev




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

[Emc-users] Can't compile MODBUS VFD in 2.8.2

[Danny Miller]

I'm doing a long-awaited upgrade from 2.7 RT-Preempt to 2.8.2 Debian.

I have an X200 VFD.  It is only superficially different than the 
WJ200_VFD included.  Like 2 coil numbers change.  I modified it and 
compiled and included it in my Linux 2.7 installation- I stumbled all 
there there, but eventually got it compiled and going.  But the steps I 
followed under 2.7 don't work. "./configure" isn't any task it can find.


I brought in my existing X200_VFD code into hal/user_comps.  It includes 
modbus.h.  It (and WJ200_VFD) fail compile for "modbus_t" being 
undefined (should be in header file).  I found modbus.h that I think 
came with 2.8.2, but it's an empty file.  I think the compile path is ok 
to find modbus.h, but why is the file empty?


I did look at github- there is supposed to be a modbus.h with contents 
there, but it's a modbus.h that contains "modbus_param_t".  The 
WJ200_VFD file uses a different modbus.h that defines "modbus_t" (and 
all the other content is different). I found nothing in the installation 
of LinuxCNC on the drive, but I did fine 
https://github.com/stephane/libmodbus IS the correct one.


How do I include this so it compiles?   What's actually broken here- why 
does the  WJ200_VFD included with LinuxCNC not have the modbus.h file it 
needs to compile?


Full disclosure- I'm not a Linux guy.  So, really basic simplistic 
difficulties might be the problem.


Danny




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

Re: [Emc-users] MESA 7i92H Power

[Danny Miller]

What is the draw of the 7i92, really?

I have a 48V Maxwell supply.  I'm currently using a step-down buck 
converter to make 5V.  This means we have to power off the whole 120v 
line when idle, and every once in awhile a user gets confused and tries 
to book LinuxCNC without remembering to power up first and gets a face 
full of weird errors.


The supply HAS  a remote-turn on for the 48V high current out AND a 12v 
100mA always-on supply, which I could buck to 5v at roughly 200mA 
reliably.  I could use a Mesa pin for "Toggle Machine Power" hooked to 
the 48v remote-turn-on.


However the Mesa manual does say 250mA max draw.

That 12v 100mA supply would probably have to be the only supply for 
running the Mesa 7i92 during jobs, too.  Switching to one derived from 
the 48v source is kinda dicey.


Does the 7i92 actually ever draw that much, over 200mA, in max use?  If 
the Maxwell 12v standby line drew 125mA for a moment, will the 12v rail 
fold or just allow it?


I know, I could add a 5v 1A usb supply or whatever.  I'd rather avoid 
adding more hardware though, esp on the 120VAC line.


Danny




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

Re: [Emc-users] Spindle experiences/recommendations?

[Danny Miller]

Have used a cheap Chinese air cooled 18krpm 3kw spindle for years.  You 
must drive them off a VFD anyways.


At 3000 rpm you have much less power.  Well the torque is probably the 
same but power=torque * rpm.


The air cooled is cooled by a shaft-driven impeller, so fan rpm drops to 
3000 rpm too.  At high torque, low-rpm loads, it will still generate a 
lot of heat but get very little fan cooling.  You can swap with a DC 
computer fan or go with water cooled.


Danny

On 6/23/2019 5:41 PM, Andy Evans wrote:

Greetings everyone,

I wish to upgrade a dressing spindle in our EMC grinder.  We currently 
have a very small DC motor turning a one-inch diamond-studded disc at 
about 1500 RPM, with fairly decent results, but my research indicates 
we may be better served with a much larger disc and higher RPM.  We 
drive this dresser under the wheel with a contoured path to give us 
the form we desire.


There are available dresser spindles in the neighborhood of $12K that 
are very good looking but we wonder if we can build something that is 
sufficient for our needs for far less.


We are considering the common router/milling machine spindles on ebay, 
air or water cooled, in the 1-2 KW range.  Most of these seem to be 
equipped with either an ER11 or ER20 spindle, and most are advertised 
as 18,000 or 24,000 RPM.  Some of them tout multiple bearings.  Some 
claim a range of 8,000 to 24,000 RPM.


Since I would expect many on this list to have experience with these, 
I am seeking to find out if we can successfully utilize a 3,000 to 
10,000 RPM range, and also if the ER mounts are integral or if they 
can be removed.


Or, if there are better ideas for accomplishing this, or if we are 
crazy not to purchase the $12K solution.


Thank you!

Andy Evans
Evans Precision Tooling Incorporated
541.990.2122






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



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

Re: [Emc-users] Dual boot for WIN-XP and LinuxCNC

[Danny Miller]

No command line anything here.  I would not characterize LinuxCNC like 
that at all.  It ABSOLUTELY is a GUI tool.


We use the WHB04 wireless mpg for everything but loading the file, 
looking at the job, and the occasional rescue from broken bits and misc 
catastrophes (rolling the job back to a specific line/feature) sends 
people to the terminal.  Which is all GUI.


The MDI will allow you to command-line g-code if you like. Really mainly 
to g0 x... y... which few ever do.  Literally no one types in probe 
commands manually on this machine.


It is a community shop with MANY users.  LinuxCNC is friendly.  I have 
personally instructed about 100 people on it, the vast majority have not 
used any CNC before.  We do about 2.5 hrs of classroom theory/CAD/CAM 
and another 2hrs at the machine going through the 5x8 machine's LinuxCNC 
interface.  I teach a lot pretty fast.


The WHB04 has a Probe-Z cycle button, and Safe-Z to lift to a machine 
coord z-max.  Jogging is usually done via mpg and zeroing axes, 
start/pause/stop is pretty much always is done via mpg buttons.  Only a 
few people use GUI buttons and nobody uses keystrokes.


Everything about the WHB04 was built in the HAL.  Which was very 
esoteric, but powerful.


Danny

On 6/3/2019 12:27 AM, Andy Pugh wrote:



On 2 Jun 2019, at 23:41, John Dammeyer  wrote:

LinuxCNC is still very much a command line type application compared to a 
WYSIWYG graphical application.

I don’t think that is entirely true. Admittedly there are a lot of scripts in 
the background handling the config, but you don’t deal with those on a day to 
day basis...

Now with MACH3 I can clip my little PC board to the router bit, set the PCB 
under the router bit where I want my zero.  Click on a button  on the screen

... Though I admit that I still use MDI for Z probing, (G38.2 F20 Z-10) which 
is a bit command-live-ey.

I have installed a nest probe screen but never got round to setting up the INI 
parameters for it.
https://forum.linuxcnc.org/49-basic-configuration/29187-work-with-probe#58620




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



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

Re: [Emc-users] Dual boot for WIN-XP and LinuxCNC

[Danny Miller]

I'm not a big Linux guy in general.  I use PC for everything but the CNC 
machine.


It's been fine.  I don't want to put CAD software on the machine running 
the CNC.


Both LinuxCNC and Linux RT-Preempt are vastly more desirable than 
Mach/Windows, even with Ethernet Smoothstepper.


The ONLY thing I miss about Mach3 was that it gave accurate, real 
runtimes upon loading a file.  Outside of running the actual job, 
LinuxCNC can only give a useless fake time estimate that doesn't involve 
acceleration.


Danny

On 6/2/2019 3:12 PM, John Dammeyer wrote:

I've figured out a way to get a 7192H to a friend who will be staying at a 
hotel in San Diego mid-June.  He'll bring it back.  Then I don't have to pay 
$38US for shipping.  Only $8.60 for USPS.  Really too bad there isn't a driver 
for MACH3 for this board to make the dual boot option easier.

John



-Original Message-
From: Danny Miller [mailto:dan...@austin.rr.com]
Sent: June-01-19 7:09 PM
To: emc-users@lists.sourceforge.net
Subject: Re: [Emc-users] Dual boot for WIN-XP and LinuxCNC

Longtime user of the 7i92.� Former user of Mach3/Smoothstepper.

It is very good all around, perfect high performance, and frees you from
almost any PC latency probs.

It can configure as a plain-jane parallel port that you'd plug and play
into PMDX-126, but also other things.

Combine with Linux RT-Preempt and it's SO solid

Danny


On 6/1/2019 3:01 PM, Andy Pugh wrote:

On 1 Jun 2019, at 20:52, John Dammeyer  wrote:

Therefore under LinuxCNC I need something with say an Ethernet interface

that has two ribbon cables compatible with PC Parallel port pinout so the
latency isn't an issue
http://store.mesanet.com/index.php?route=product/product=69_62
duct_id=306

(Or also look at the 7i80)
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users



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



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



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

Re: [Emc-users] Dual boot for WIN-XP and LinuxCNC

[Danny Miller]

Longtime user of the 7i92.  Former user of Mach3/Smoothstepper.

It is very good all around, perfect high performance, and frees you from 
almost any PC latency probs.


It can configure as a plain-jane parallel port that you'd plug and play 
into PMDX-126, but also other things.


Combine with Linux RT-Preempt and it's SO solid

Danny


On 6/1/2019 3:01 PM, Andy Pugh wrote:



On 1 Jun 2019, at 20:52, John Dammeyer  wrote:

Therefore under LinuxCNC I need something with say an Ethernet interface that 
has two ribbon cables compatible with PC Parallel port pinout so the latency 
isn't an issue

http://store.mesanet.com/index.php?route=product/product=69_62_id=306

(Or also look at the 7i80)
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users




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

Re: [Emc-users] Hardware emcmot? Real time delay

[Danny Miller]

On 4/13/2019 3:46 PM, Nicklas SB Karlsson wrote:

Even th D-525-MW motherboard? I can latency-test for an hour or 2 on this
machine I'm building up right now, and not show any worse than 5.1
microseconds for base thread latency, and If I run without that thread,
which I don't use anyway, and the worst case latency is 4850ns.  Thats
right alongside the figures I've seen quoted for the STM-32.

On an STM32 it will probably be less than 4850ns on the highest priority 
interrupt. Others however depends on threads with higher priority. Ideally all 
threads should be known and have an upper bound on execution time then it might 
under correct conditions be possible to calculate beforehand all dead lines 
will be met.

Then I last checked real time seems to run really well on ordinary linux with 
real time kernel. Do anybody know if it is possible to list running real time 
threads in linux?


I did a calc and got *10ns* latency on a Cortex m7 400MHz, with only 2ns 
of that as actual jitter, which would be avoidable.


Danny



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

Re: [Emc-users] Hardware emcmot?

[Danny Miller]

Well, what I would ideally picture is a realtime hardware motion 
controlled coupled to a Raspberry PI running Linux non-RT, the PI is 
hooked to a PC and has an embedded http server.  At that point you 
access it via browser and load and control control your job from the PC 
interface.  estop etc is still on the hardware.  At that point it does 
not matter if you use Windows or Linux.


Danny

On 4/11/2019 7:19 PM, bari wrote:

I disagree. The problem is making the machine configuration easy. If
LCNC ran on Winders or Androids, many would still complain about how
difficult it is to configure. Active High vs Low for Limit switches,
Home switches, Step and Dir signals, encoder pulses per rev. etc etc is
far to challenging for many to deal with.

On 4/11/19 2:41 PM, Chris Albertson wrote:

If the machine controler could run on a me=aintstream OS (like Windows 10)
and on a normal notebook PC it would be orders of magnitude more popular
and widly used



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




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

Re: [Emc-users] For PCW

[Danny Miller]

I have used an RS232-to-RS485 adapter and just plugged it into the PC 
serial port.  It was not ideal to add another cable from the PC but it 
works fine.  There are MODBUS drivers for LinuxCNC.


Danny

On 4/11/2019 5:30 PM, Peter C. Wallace wrote:

On Thu, 11 Apr 2019, Gene Heskett wrote:


Date: Thu, 11 Apr 2019 17:56:45 -0400
From: Gene Heskett 
Reply-To: "Enhanced Machine Controller (EMC)"
    
To: "Enhanced Machine Controller (EMC)" 


Subject: [Emc-users] For PCW

Greetings Peter;

The new vfd materialized here today, and I see it actually has what it
calls an RS485 interface, a teeny little white 3 pin socket on its
motherboard.

And I have one or two 422 to 485 translators, the $2 variety, that I
believe are duplex capable as I had to wire them to a solid logic level
on the 5i25 end of them before they would take the balanced signals from
the encoder I put on the spindle motor of my G0704, and make nice 5 volt
rail to rail square waves to feed the 5i25.

If I jury rig the separate 422 lines end of that to the 422 terminals on
that 7i76D, can I talk to a huan yueng lookalike?  That is what I
supposedly bought but the instruction booklet makes no brand claims.

It is a YL-620-A, and it also has the usual analog interface too.
It claims no ascii, but supports the modbus RTU protocol, and mentions
modbus several times..  Am I better off useing the analog, or is the
RS-485 the better control?

Unfortunately you cannot do this currently since MODBUS is not 
supported by the Hostmot2 UART driver



Analog control may be better for things like rigid tapping since the 
normal MODBUS driver is a userland task (and I'm not sure how real 
time the MODBUS implementation in the VFD is.



I also see an AO output, but its reads like 0-5 volts, which doesn't
sound like it would work to indicate the achieved speed via one of the
A/D's on the 7i76D when the 7i76 goes to about 38 volts full scale.



If you have a spindle encoder, that a better way to get speed feedback

This is the first vfd I've had that actually has braking R terminals, 
and

says optional, but if used s/b 100 ohms, 100 watt.  What advantage would
hooking one of those up amount to?



Fater slow-down and faster reversals if tapping (better control of depth)


Thanks for any advice Peter.

Cheers, Gene Heskett
--
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page 



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



Peter Wallace
Mesa Electronics

(\__/)
(='.'=) This is Bunny. Copy and paste bunny into your
(")_(") signature to help him gain world domination.



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




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

Re: [Emc-users] Hardware emcmot?

[Danny Miller]

On 4/10/2019 11:31 AM, andy pugh wrote:

On Wed, 10 Apr 2019 at 17:12, Danny Miller  wrote:


I have been wondering- can emcmot be separated from the HAl and emctask
and become a true dedicated realtime stage to control the joints?

I suspect not as it stands. This is based on the observation that
motmod is a HAL module.


?

http://wiki.linuxcnc.org/cgi-bin/wiki.pl?EMC_Components

Shows EMCMOT as a discrete module, and emcmot.c and emcmot.h are 
supposed to be discrete files.  The motor PWMs and switch IO is under 
the realtime veil, which is easy to integrate into hw.


I see the shared memory (inherently fast) with EMCTASK, but I doubt 
ethernet latency would be a dealbreaker.  Correct me if I'm wrong.


I also have this diagram:

https://www.researchgate.net/figure/EMCMOT-Structure_fig2_265199849

which depicts the HAL with servo feedback (not essential as we use 
steppers) and fault/limit switches going to EMCMOT on a separate port.  
That seems functionally identical just a different artistic style.


I'd like to be able to integrate 2.8's circular arc blending trajectory 
planning into laser cutter and 3d printer control HW. This might 
eventually mean EMCTASK is also on hardware and Windows machines or 
anything else with whatever bastard config they happen to have installed 
can keep it fed with G-code.  The subset of common g-codes may be 
preferable, you don't need G73 high speed peck drills or G74 left-hand 
tapping cycle for 3D printers or laser cutters.  I don't think any of 
the CAM pkgs would even generate subroutines which is best to leave 
unsupported as it's nonlinear.


Ideally, a bit ambitious, and not entirely within G-code, but I'd like 
to take a stab at nontrapezoidal acceleration profiles and multiple 
acceleration models.  This would be useful- in transitioning from the 
end of a laser cut path into a rapid, it's "as fast as the machine can 
go" once the beam is off, whereas in starting a cut you may want a 
smoother stop because of belt stretch causing some oscillation that 
would persist briefly as the beam turns on.  With lasers you typically 
want to fullstop briefly at the start and end of every cut, but then why 
slam it to a rushed G0 stop only to do a 
G0-to-cold-dwell-then-powerup-dwell-then-go.




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

[Emc-users] Hardware emcmot?

[Danny Miller]

I have been wondering- can emcmot be separated from the HAl and emctask 
and become a true dedicated realtime stage to control the joints?


The shared memory FIFOs could be replaced with like an ethernet link to 
the hardware RAM.  You'd probably load a DMA buffer and just let it 
flow, there's some minor latency but there's not a realtime requirement.


There are cheap high freq MCUs with single-cycle hardware double 
multiply/add and gobs of memory.  It would have to be coded in non-OOP C 
and I greatly prefer hardware interrupt-driven strategies over RTOS.


Danny






___
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.) --> CPU + Ethercat + FPGA/CPLD but no PRU

[Danny Miller]

One common solution is you have a sizable FPGA and just add a soft core 
into it.


RISC-V is an open-source, royalty-free core specifically architected to 
efficiently implement as FPGA gates that Keil can compile for, and the 
common Segger JTAG programmer can program AND debug the code.


At that point, you're pretty seamless.  You can write your own 
peripherals that just memory map onto the data bus without touching the 
core.  You can have a custom hardware neural network that has registers 
and DMA that you just let run independent of the instruction core.


Danny

On 3/31/2019 3:14 AM, Nicklas Karlsson wrote:

A CPU with Ethernet and a builtin FPGA/CPLD to implement the SPI ports would 
probably be a very good device for an SPI router, probably also very cheap. 
Packets could routed to ordinary computer running Linuxcnc as is today or split 
so that real time part is in the simple device. It is not to different from a 
PLC I have on my workplace, it is mounted on a din rail without any user 
interface except a reset start/stop/reset button, there is an Ethernet 
connecter used for the programming, it is also possible to connect user 
interface if needed but I have not investigated how it communicate.


I remember someone here talked about someone implementing PWM on a PRU. I did 
on an ordinary Micro controller once, I guess code could be reused but I do not 
want to do it again. A timer with compare registers plus a few other things 
usually available in micro controllers is really good at this.


On Thu, 28 Mar 2019 08:05:26 +0700
TJoseph Powderly  wrote:


Gene hello



On 03/27/2019 08:13 PM, Gene Heskett wrote:

On Wednesday 27 March 2019 08:12:48 TJoseph Powderly wrote:


( list apology, I seem to have replied to posters rather than to the
list, and several times :-(

re SPI comms Linuxcnc SBC microcontrollers ...

it has been said on this mail list,
that SPI is a good candidate for a bus technology to work _with_
realtime.

heres some work in that vein (maybe overlooked)

yeltrow's work:
a generic spi hal module so you can use other SPI devices ( rpi,
arduino, mcp23s17, ENC28J60, theres a lot of spi stuff to hang onto
such a bus )
https://www.forum.linuxcnc.org/24-hal-components/28851-spi-bus-generic
-driver-and-st-l6480


Interesting driver for the big stuffs. However, I wonder if yeltrow is
aware of rpspi.ko, now part of LCNC.  Not parport based but gpio,
written specificly for the rpi3b.  And I'm using it, writing to a Mesa
7i90 at 42 megabaud, and reading back from the 7i90 at 25 megabaud
useing only 4 gpio pins for 2 or 3  target devices.

I dont find source rpspi.c in my sources.
cant google it except to find messages from you.

My RIP source tree is DGarr's external offset branch.

I dont have raspi.kp either.

Where are these files?

The files I was speaking of were generic spi utilities,
not RPi to Mesa communication modules,
They were attempts to make SPI available on any hardware platform to any
SPI device.
They are attempts only ( one uses a parport but i guess that could be
altered to gpio )
But from the parport, the author connects to many different SPI devices.

What file builds rpspi.ko ?

Maybe its private/unpublished work from Matsche ?

I cant find it on https://github.com/LinuxCNC/linuxcnc either.

Plz lemmeno
TomP

Perusal of that code might be of help to yeltrow. gpio seems to be about
10x faster than trying to simulate SPI over a parport.  The 7i90 has
both modes depending on the firmware loaded.


the files are on the forum for members linuxcnc-upload-2015-12-03.tar

So this work handily predates rpspi.ko. Still, theres obviously things to
be learned from the rpi version.


erste's work:
for ethernet circumventing usb ( via spi
interesting as spi and ethernet seem to be future avenues
http://erste.de/ethraw/README

theres other efforts but less coupled to the spi grail

tomp


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

Cheers, Gene Heskett



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


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




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

[Emc-users] Fusion360 gcode= "Radius to end arc differs from radius start"

[Danny Miller]

I'm managing a community shop with the CNC I designed and built. It's 
running on LinuxCNC 2.7.4


I was told people brought in some Fusion360-generated code that created 
an error "Radius to end arc differs from radius start". Nobody has 
provided me that gcode, so I have no further details. Google saysa this 
was a common issue after a q4 2018 patch to fusion360


People cited this:

https://forums.autodesk.com/t5/fusion-360-computer-aided/help-with-radius-vs-ijk/m-p/7234440#M34397

That the .ini file should assign a new #TOLERANCE_INCH and #TOLERANCE_MM

Does this make sense as a fix?

Danny




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

Re: [Emc-users] Tool touch-off

[Danny Miller]

So, I did see that earlier, it CAN be used to zero Work Coords- but ONLY 
if you use Bottom Datum.  Many times, we use Top Datum.


If you just reassigned WC Z, then LinuxCNC should be able to see 
lowest-Z is beyond the allowed Machine Coords automatically, but I have 
doubts it could work- the file may not get reevaluted automatically when 
WC Z is reassigned between the time Run is pressed and we actually start 
user G-code.


However, Top Datum, that wouldn't apply at all.  You can't reassign WC Z.

Danny


On 12/16/2017 3:02 PM, Jon Elson wrote:

On 12/16/2017 12:53 PM, Danny Miller wrote:


Every time you hit Start, before it starts the spindle or actually 
runs any line of the user G-code, it jogs over to a tool height 
sensor (THS) and probes for the installed tool height


If you did this like a homing move, only the tool height sensor sets 
the Z home position, and the axis soft limits are such that MIN_Z = 
0.0, then it would be impossible to cut into the spoil board.


Jon

-- 


Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users




--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Tool touch-off

[Danny Miller]

I have a lot of users on my CNC router now and, predictably, they keep 
chewing into the spoilboard.


I had some thoughts about what if:

Every time you hit Start, before it starts the spindle or actually runs 
any line of the user G-code, it jogs over to a tool height sensor (THS) 
and probes for the installed tool height


If you factor in the lowest Z anywhere in the file, vs the difference in 
currently predefined spoilboard surface and where the THS was mounted, 
vs where it found the tool tip, means that at some point it goes more 
than 0.05" into the spoilboard, it just refuses to run.


This can't be in the G-code file.  I don't have control over how people 
create G-code they bring.


I can't do this as a manually invoked step prior to running. Because if 
I could rely on users to do the right thing I wouldn't have a problem in 
the first place.  No, they might press the "do the probe" thing 
correctly, then change bits and forget to redo it, and run the job too 
low and gouge.


How would that work?  I have G-code for a wireless toolsetter that uses 
the probe command.  But I have doubts that'll work.  Is there any access 
to a variable for "the lowest Z value found when the file was loaded, 
accounting for the Work Offset=lowest MACHINE Coord Z found in the 
current G-code"??


Danny


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Tech Shop goes out of business

[Danny Miller]

I saw this info, including a non-public letter to their investors.

It declared an intent, yes.  BUT lemme say it's "out there".

It does not look like a professional business move.  It doesn't look 
like it's even in accordance with acceptable business practices.


They didn't file for chapter 13, which companies sometimes use to 
restructure debt and stay in business.  They filed chapter 7- the 
business assets and its whole existence is handed over to a third-party 
trustee who liquidates it and uses the cash to partially pay back 
creditors.  The business is then gone-gone.


Danny

On 12/4/2017 8:39 PM, Jon Elson wrote:
Well, there's an update.  Apparently an investor has come in and will 
reopen the Tech Shops fairly soon.  I think it will be called Tech 
Shop2.  There will almost certainly be some reorganization of the 
business model.


Jon

-- 


Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users




--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Interactive g-code routine?

[Danny Miller]

I have a simple .ngc file I run to run the dust shoe across the entire 
table.  I'd like to add a UI button to run this.


But there are other users of this machine.  I would like to avoid just 
loading the file and having to run it manually.  However, pressing a 
"Clean Table Now" button with no further prompting (regardless of 
whether it auto-runs or not) is kinda dangerous. i.e. "I didn't realize 
I had to remove the bit before starting it" is predictably gonna come up 
over and over.


I imagine after pressing "Clean Table Now" UI button, you'd get a popup 
"Ensure bit is removed, dust shoe is attached without the spacer, and 
press (some key) to run or (another key) to abort". And ideally return 
to the original file without reloading (like the tool probe routine).


I know error messages keep popping up in small red text in the lower 
right hand corner, and remain until scrolled by other messages.  I don't 
think that's adequate, honestly.   It's a poor way to represent error 
messages and worse for trying to be interactive.


What's the best way to do that?

Danny


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] How to add additional condition for enabling machine

[Danny Miller]

Electrically you need a way to get this signal on a pin.

Past that, it well within the capabilities of the HAL to code up that 
this pin reading "TRUE" is a condition for running.  It will not matter 
if "TRUE" is high or low.


The HAL can be used to integrate this signal, in like a dozen different 
ways.  It will only require like <5 lines.


Danny


On 7/29/2017 3:30 PM, Tomaz T. wrote:

Some more explanation


pressure switch is a simple NO switch which is triggered at certain pressure 
(5.5bar in my case) and that is minimum I need to operate my machine (Z-axis 
counter weight cylinder). So every time I want to use my machine I (manually) 
have to open valve for air.

I doesn't need to stop (or pause) machine if pressure drops, as my air compressor is 
always on "stand by".

So this should be one of conditions needed to power up machine and start homing 
axes.

I don't like the idea to wire it in same circuit as e-stop, I think it would be 
nice to receive warning for this specific error, if it is possible?



Two important questions:
1.  Do you want the machine to pause if you lose pressure mid-run?
(almost certainly so)
2.  Do you want the machine to be able to auto-resume if pressure drops
but comes back?




--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users




--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] How to add additional condition for enabling machine

[Danny Miller]

Two important questions:

1.  Do you want the machine to pause if you lose pressure mid-run? 
(almost certainly so)


2.  Do you want the machine to be able to auto-resume if pressure drops 
but comes back?


Danny


On 7/29/2017 1:36 PM, David Berndt wrote:
If the air pressure is related to movement of an axis (or you want to 
say it is...) you can use axis.N.amp-fault-in as a place to input an 
your error condition. The machine won't be forced to e-stop, but will 
not allow the power button to be pressed/stay on.


Same thing should be possible with motion.enable in bit.

For the spindle, maybe motion.spindle-inhibit in bit is what you're 
looking for.


If you want the machine to be totally useless then perhaps wiring it 
into the e-stop circuit is the way to go. Personally I'm not on board 
with that. But about half of the standard conventions in linuxcnc seem 
to rub me the wrong way, so it might be just fine with you.



Dave

On Sat, 29 Jul 2017 12:10:31 -0400, Tomaz T.  
wrote:


I have air pressure switch to control if there is enough air 
pressure. Now I would like that this is also condition for enabling 
machine to run.


I would need a little help how to correctly add this to hal.
-- 


Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users







--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] change of subject, brass or copper?

[Danny Miller]

Look up CA954 aluminum bronze.

Very good sliding surface on stainless.

Danny


On 7/15/2017 10:05 PM, Gene Heskett wrote:

On Saturday 15 July 2017 22:00:44 Trent Hejazi wrote:


I have brass contact fingers and they work well.

I'll get a brass bar yet tonight then, thanks.


When supporting the
barrel tennon, they ride on the threads without damaging them.

Threads on a p-17 are square, 10 tpi. I already have a tool ground for
that.


Otherwise, they will always leave an annoying ring on a barrel.  Only
way to circumvent is to make a collar to slip over barrel, then skim
cut concentric with the bore.  Other option is to make a set of
inboard and outboard cat-heads so you can work near the head stock.
This really depends on the spindle bore dia and barrel dimensions you
are working with.

Fortunately, the spindle will pass the barrel nicely all the way.

It looks like I'll have to make the spider bolts, so I may as well get a
foot of 1/2" brass hex stock while I'm at it.

Currently working on a piece of 3/4" steel, crooked as a dogs hind leg,
need to fit the action on it between centers so I can face it square.
But first I had to shim the tailstock for elevation, it was about 25
thou low.

Then figure out a way to spin the bolt as its face is about 3.5 thou out
of square.  Thats hard on brass. Then measuring the tenon after getting
it out, it doesn't match the drawing I have by as much as 37 thou in one
measurement.  And I need to make a headspace gauge as I've not managed
to find one for the 6.5 C.M. Yet. Lots of work before I actually chuck
up the barrel.

And I found today, that the scroll in this 3 jaw chuck is an absolute
piece of crap. So I am tightening it, measuring the runnout, loosening
the mount bolts and tapping it gently to make it run true at the size
its currently tightened on. This chuck, a Polish Bison that looks brand
new, must be a very poor factory second to be that far off. I already
cut the registration in the backplate down some so I could move it
around.

Honeydo's call. :)


Marshall

Thanks Marshal.


On 7/15/17, 8:46 PM, "Gene Heskett"  wrote:

Greetings all;

My steady rest has steel sliders, which will not do for working on
finished parts.  So I need to order a foot of 1/2x1" brass or copper
to make some more better shoes.

Which will do the least damage to a pre-finished part, such as the
fairly high polished SS barrel?  Brass or copper?

Thanks everybody.

Cheers, Gene Heskett
--
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page 

-
- 
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

--
 Check out the vibrant tech community on one of the world's
most engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users


Cheers, Gene Heskett



--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Multi-meter recommendations

[Danny Miller]

Best deals of out of grey market China sites:

http://www.dx.com/s/fluke?cateId=0=All%20Categories

Fluke cut China a deal.  17B+ is over $300 in US.  But China will sell 
you one of their discounts- same meter- for $130.


It IS a genuine Fluke.  Same product.  Only confusion is, on eBay, 
they're selling cheap VC97 meters spamming the keyword "fluke" in the 
title.  But it obviously isn't a Fluke.



Danny


On 7/15/2017 11:54 AM, Nicklas Karlsson wrote:

Thes HP bench multimeters I think have really good accuracy, they do not run 
away but are a little bit harder to move.


On Sat, 15 Jul 2017 09:47:03 -0700
Chris Albertson  wrote:


I agree about Fluke build quality.   I have a couple older Fluke portable
DMMs.  They drift out of Cailbration and typically you need to adjust
serval pots with a mini-screwdriver.My HP calibrates with push buttons
and no disassembly is required.

Here is an example, this meter is really common and there are always a few
on eBay every day
www.ebay.com/itm/HP-Hewlett-Packard-3478A


On Mon, Jul 10, 2017 at 5:16 PM, Christopher  wrote:


What is also great about Fluke products is that you can get just about
every spare part you could want and can actually get them calibrated if
need be.  I have a really nice Fluke 87 that I bought used.  It was pretty
beat up when I got it, but I was able to purchase all the parts I needed to
get it working / looking just about new.

N. C. Perry


On 2017-Jul-10, at 04:53, andy pugh  wrote:

On 10 July 2017 at 05:56, Bruce Layne 

wrote:

I think the best value in a quality DMM is a Chinese Fluke.

I am happy with mine (12E+) bought soon after seeing that AvE video.

I have also had good service from a Uni-T
http://www.uni-trend.com/Productslist_1106_1106_1106_1106.html
But that has become a little floppy in the lead-sockets, and also has
the weird feature of over-estimating voltages if the internal battery
is low (so, if it is showing a low battery warning, you can't trust
the voltages. They end up doubled, just before it dies). Swapping the
battery is awkward too, you need to dismantle the meter, there isn't a
battery compartment as such.
However, it does have some measurments (capacitance, inductance,
frequency) missing from the Fluke.

--
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916



--

Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users



--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users




--

Chris Albertson
Redondo Beach, California
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users



--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] OT: Waterjet cutter

[Danny Miller]

There was a kickstarter for the "Wazer", a home/low-end waterjet cutter.

Problem is, it is very slow and consumes a LOT of expensive abrasive 
media per inch, relative to a full-size waterjet.


A waterjet's key component is the intensifier pump- it is FAR larger 
than a pressure washer.


Now, granted, IMHO there IS a solid niche for a slower waterjet cutter 
if the cost and facility requirements are much lower.


But the key is being efficient with the abrasive consumption, on a 
per-inch basis.  So far, "not good".


Though you might think "I'll just buy blasting sand and use that", 
sorry, it's my understanding this won't work.


Unless someone DOES invent a low-end waterjet that just uses blasting 
sand, in which case, "f'ing A, man!"  Yes, that would be a game-changer, 
but the market already moved away from sand as basically useless long 
ago so you'd have to have some new, overlooked tech never thought of to 
change that game.


http://www.thefabricator.com/article/waterjetcutting/selecting-the-right-waterjet-abrasive

Currently the standard media is almandine garnet particles.  They have 
the hardness, toughness, density, and consistent particle size to do the 
job.


Danny


On 6/17/2017 11:56 AM, Roland Jollivet wrote:

Hi All

I came across this video of a home-brew waterjet cutter. I don't know if
this guy is on this forum, or if this was posted before, but;

here is quite an interesting video
 on a home made water-jet
cutter.
It's a pity he does not give more data on pressure/cutting speed for
various materials, including steel.

https://www.youtube.com/watch?v=Lg__B6Ca3jc

Roland
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users




--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Mesa 7i47S issue

[Danny Miller]

That's not a true differential output.  It's an optoisolator's two 
output terminals.  Since they're isolated, neither terminal is tied to a 
gnd or + so you can connect it high or low.


When tripped, it conducts current (supplied externally) from + to -.  It 
cannot generate its own voltage, only conduct.


So all you need to do is hook it to a pin with an external pullup.  
AFAIK all the 7i47 have a pullup to +5v.   Alarm+ to pin, Alarm- to 
gnd.   The pin will go low when the Alarm signal is active.


Similarly the inputs aren't true differential, they're an optoisolator's 
input diode with a series resistor to produce the rated trigger current 
when +5v is applied + to -.  Due to some esoteric details in how the bus 
switch drive output works, these work best if you tie the + to +5v and 
use the pin to drive the - to gnd.  The signal is weaker if you do it 
the other way.



Danny


On 6/17/2017 9:30 AM, Andrew wrote:

Hello,

The stepper drive has open collector alarm output.
Can I use a differential RX input on 7i47s to read it?

Thanks
Andrew
​
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users



--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Fiber optic control for CNC

[Danny Miller]

Leadshine drives input/output are not just opto, they're differential.  
i.e. the step signal has 2 wires and dir has 2 wires, there's no common 
ground or common anode.  So they're not just galvoisolated from the 
pulse source, but from each other.

When run as twisted pairs or just pairs, the noise immunity is greater 
than you'd ever need.

You are still going to tie them together as common-gnd or common-anode 
at the pulse source, but the differential-pair part that creates is 
important.

Fiber optic requires serialization, added a lot of new complexity and 
latency, for no real benefit.

Danny


On 4/20/2017 3:10 PM, Chris Albertson wrote:
> Yes.  The linked device acts just like 1/2 of a 3.3 volt opto isolator.
>
> However there are devices LIKE it that cost 1/10 as much, run on 5
> volts are faster and use consumer grade cables.
>
> But the ability to use un-terminated fiber cable cut from a bulk
> spool.  It is worth paying more for that.
>
> I'm wondering WHY?  Is it for galvanic isolation?  A small opto device
> does that.  Is it for noise impunity, shielded differential twisted
> pair cable does that really well.  Normal 100BaseT  Ethernet is
> already differential and isolated.  The normal engineering method is
> to first spec a requirement THEN look for parts.  This saves a lot of
> work compared to doing it the other way around.
>
>
>
>
> On Thu, Apr 20, 2017 at 12:22 PM, Ken Strauss  
> wrote:
>> Could something like
>> https://www.digikey.com/catalog/en/partgroup/afbr-1528cz/66628 be useful?
>> They are about $12 each but operate at 5Mbps so perhaps you could multiplex
>> some of the signals. There is also an eval kit
>> https://www.digikey.com/products/en/development-boards-kits-programmers/eval
>> uation-and-demonstration-boards-and-kits/787?k=hfbr-1506amz  You can get
>> 10-foot fibre patch cables with SMA terminations to avoid making your own.
>>
>>> -Original Message-
>>> From: Gene Heskett [mailto:ghesk...@shentel.net]
>>> Sent: Thursday, April 20, 2017 2:51 PM
>>> To: emc-users@lists.sourceforge.net
>>> Subject: Re: [Emc-users] Fiber optic control for CNC
>>>
>>> On Thursday 20 April 2017 11:22:03 Todd Zuercher wrote:
>>>
 Fanuc has been using fiber optic connections for more than 30 years.
 And I work with an old SCM machine with a Num 1040 control that has a
 bank of fiber optic remote io that is one of the things holding me
 back from trying to convert it to Linuxcnc.

>>> This has been a wish of mine for a long time.  We need some cheap stuff
>> that
>>> can be the equ of an opto-isolator, but with a piece of glass or plastic
>> fiber up
>>> to several feet long as the optical media between them.
>>> With power on the rx end, signal losses in the plastic fiber could be
>> easily
>>> compensated. I can visualize stepper drivers incorporating it, probably at
>> no
>>> more cost for the pair of fiber sockets than they cost for the full opto-
>>> isolation BOM right now.
>>>
>>> All we would need would be an HE IR LED with a molded in funnel to guide
>> the
>>> fiber tip to the chip face in the LED.  Some sort of a jacket on the fiber
>> to
>>> prolong its life if rubbing on something, or to prevent optical crosstalk
>> would
>>> be needed. The major design problem AISI is in gripping the fiber to
>> anchor it
>>> at both ends, with long term gripping pressure imprinting itself on the
>> fiber
>>> creating radiation leaks.  Even glass will do that, but usually over time
>> frames
>>> that exceed the life of the rest of the machine since technically, glass
>> is a
>>> super-cooled liquid, flowing visibly over the lifetime of the observer.
>>>
>>> The led makers have now had 40+ years to design such a package, and I fail
>> to
>>> understand why it has not happened.
>>>
>>> Or has it, and I missed the announcement?  Thats a plausible excuse given
>> my
>>> age and retired status. If anybody would have it on this side of the pond,
>>> digikey, and I've only found one candidate so far. Intended to be a
>> remoteable
>>> indicator, the led is conventional shaped, available in several colors.
>> Pricing
>>> starts at $2.15 with a 6" light pipe, Available up to 3940" long. :) With
>> the far
>>> end of that pipe facing into an avalanche mode transistor, mounted exactly
>>> the same as the led, and substituted for the BOM that puts the
>> opto-isolation
>>> into the stepper drivers input circuits, it ought to be essentially free!
>> What
>>> sort of speed would have TBD.  But I'd certainly have to think its faster
>> than
>>> the 200 to 300 kilohertz we can now drive a $35 M542 stepper driver.
>>>
>>> Look at 
>>>
>>> And see what you think of it.
>>>
>>> One of those in red led, looking at one of these:
>>>
>>> 
>>>
>>> Would I think, be a good test bed. Need some smd bypassing of the rx
>> supply,
>>> and I'm not sure what 

Re: [Emc-users] Changing arrow key bindings

[Danny Miller]

I thought "well, wouldn't it be a matter of just telling axis.py that 
"Rotated Top View" means -90 instead of +90?"

I looked through.  It looks like this is "z2", or "view_type.set(2)".  
It looks like view_type=2 means "use rotated top view".

Presumably that's just gonna mean to view the XY plane, fix the origin, 
set the rotation=+90, and if you click on "rotate view" button it'll 
leave those parameters as-is for a starting point but un-"sink" the 
"Rotated top view" button and sink the "rotate view" button.

But I don't see anywhere where it implements the z2/view_type=2 mode.  
Does the AXIS frontend not do this, it's in Linuxcnc itself??

Danny


On 3/19/2017 3:21 PM, Danny Miller wrote:
> I should clarify (as assumed anyhow) that I'm using AXIS.
>
> Actually, isn't "Rotated Top View" just that, anyhow?
>
> Can I make it the default?
>
> Crap...  looks like the "Rotate Top View" is going to rotate the wrong
> way, making it 180 deg from what I need.  The machine's at the shop but
> that's what it looks like it says.  Any way to change that??
>
> Danny
>
>
> On 3/19/2017 2:37 PM, Jeff Epler wrote:
>> In master branch with the AXIS user interface, you can control how the
>> jog keys are assigned with [DISPLAY]JOG_AXES and [DISPLAY]JOG_INVERT, as
>> documented here:
>> 
>> http://linuxcnc.org/docs/master/html/config/ini-config.html#_display_section
>> In 2.7 and in other UIs there may not be a way to do this without
>> modifying the source code and rebuilding.
>>
>> This does not modify the orientation of the preview display, so that's
>> another possibility for future enhancement, i.e., the ability to control
>> what the preset views are.  I can certainly understand why you would
>> want them to be related to how the operator sees the machine from where
>> she operates the control panel.
>>
>> Jeff
>>
>> --
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>> ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>>
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Changing arrow key bindings

[Danny Miller]

http://linuxcnc.org/docs/2.6/html/config/ini_config.html

"Geometry has no effect without a rotary axis."

I think that's the answer there.

Danny


On 3/19/2017 4:35 PM, andy pugh wrote:
> On 19 March 2017 at 20:21, Danny Miller <dan...@austin.rr.com> wrote:
>> Any way to change that??
> I am wondering if you can do that with GEOMETRY = CXYZ.
>
> The bit I haven't figured out is how to make glcanon.py think that the
> mythical C = 90
>

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Changing arrow key bindings

[Danny Miller]

I should clarify (as assumed anyhow) that I'm using AXIS.

Actually, isn't "Rotated Top View" just that, anyhow?

Can I make it the default?

Crap...  looks like the "Rotate Top View" is going to rotate the wrong 
way, making it 180 deg from what I need.  The machine's at the shop but 
that's what it looks like it says.  Any way to change that??

Danny


On 3/19/2017 2:37 PM, Jeff Epler wrote:
> In master branch with the AXIS user interface, you can control how the
> jog keys are assigned with [DISPLAY]JOG_AXES and [DISPLAY]JOG_INVERT, as
> documented here:
>
> http://linuxcnc.org/docs/master/html/config/ini-config.html#_display_section
> In 2.7 and in other UIs there may not be a way to do this without
> modifying the source code and rebuilding.
>
> This does not modify the orientation of the preview display, so that's
> another possibility for future enhancement, i.e., the ability to control
> what the preset views are.  I can certainly understand why you would
> want them to be related to how the operator sees the machine from where
> she operates the control panel.
>
> Jeff
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Changing arrow key bindings

[Danny Miller]

I set up my 5x8 machine with the X in the long direction.  It is loaded 
from the narrow side.  Originally the terminal was on a table parallel 
to the long X axis with the router behind the monitor. That seems 
correct to me, VCarve CAD/CAM will always put the X-axis horizontal, the 
long side of your CAD/CAM monitor, and origin in the lower right to keep 
all XY coords positive is also the most natural thing.

Somewhere along the way the keyboard and monitor were moved to the right 
of the narrow side, putting the monitor parallel with the Y axis.  
Someone complained this was wrong.

For the most part, everything's done with an MPG, but if you do fall 
back to the arrow keys, "UP" sends the spindle to the left.  And the 
screen's orientation is off by 90 deg.  I never even thought about it.

LinuxCNC's toolpath view IS already square.  No case for saying the 
monitor would need to physically be mounted in a landscape orientation 
to match the bed length.  Which is good because rotating whole graphical 
front end sounds complicated.  Can we tell it to just orient the preview 
window XY so X+=up, Y+=left, and origin in in the lower right by 
default?  How would I do that?

I googled changing the key mappings so "up arrow"  means X+ and it was 
"you just need to change axis.py", but then "hey ok but that's this huge 
program to edit".  Is that currently the only way to do it or you just 
need to change something in the .ini file now or what?

Danny





--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Using a homing switch as a limit when not homed

[Danny Miller]

On 2/8/2017 8:15 PM, Jon Elson wrote:
> On 02/08/2017 11:34 AM, dan...@austin.rr.com wrote:
>> I don't have a physical stop on my Z axis, putting one in would be a PITA.  
>> The drive is exceptionally strong and it would be problematic to stop the 
>> axis that way.  If you're an idiot and jog it that way prior to homing, you 
>> can dismount the axis.
>>
>> There is a homing switch at Z+.  It sounds desirable to halt + movement when 
>> that switch is tripped, AND not homed, AND not in the actual homing 
>> sequence.  But NOT stop - movement, otherwise you'd jog it upwards, get it 
>> stuck, and no way to jog down.
>>
>> How difficult would it be to do that?
>>
>>
> LinuxCNC includes provisions for this.  Set
> HOME_IGNORE_LIMITS = YES in the .ini file for that axis.
> The trick, of course, is to make sure the home position is
> moved off that switch.  So, you have to set a
> HOME_OFFSET = -0.1
>
> or something like that, so it jogs off the home switch
> sufficiently to UN-trip the switch.
> Then, set your MAX_LIMIT to less than that distance (or zero).
>
> This way, it will home up to the switch, move down by the
> HOME_OFFSET amount and then set soft limits so you cannot
> reach the switch.
>
> Jon
>
Hmm, could I also AND the input pin with NOT IS_HOMED for Z before tying 
it in as  a z+ limit?  So the limit switch can only apply prior to homing?

Thing is, I like to preserve the capacity to travel all the way up.

Danny


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Finally figured out how to halve an axis!

[Danny Miller]

This has troubled me for awhile.  I have the XHC-HB04 pendant. On it is 
a "1/2" button.  It is very useful when your CNC job uses center datum 
and you want to place it in the center of your stock.  e.g. you've got a 
X=10" wide board, you do X=0 at one side, move to the other size and 
X=10, hit "1/2" and you want the current work coords to shift so X=5 
there and now X=0 is the center of the board.

Troubles are, LinuxCNC doesn't provide clear access to absolute value of 
current work coords.  It can't be work offsets, and it needs to apply to 
the currently selected work coord system.  I THOUGHT I had a solution 
but I later discovered it was bugged and useless.

But here's the solution!

#18
MDI_COMMAND = G10 L20 P0 X[#5420/2]
#19
MDI_COMMAND = G10 L20 P0 Y[#5421/2]

G10 L20 = "set to a calculated value that makes the current coordinates 
become the given value"

P0 = "in the active coordinate system"

#5420= "Current relative position in the active coordinate system 
including all offsets and in the current program units for X, Y, Z, A, 
B, C, U, V & W, volatile."

Boom.  That was way too hard to figure out.

Danny




--
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Will a PCI serial / parallel card cure port issues?

[Danny Miller]

A card doesn't fix much.  Basic problem is you're timing EACH step, 
which can be over 6KHz.  A single step being significantly out of place 
can create a stall.  The main source of jitter is in the CPU and 
chipset, not the parallel port.

The Mesa stuff, I believe mine is a 1ms period, and it's actually an 
overall plan, the FPGA will carry out the timing plan smoothly even if 
the packet is early or late, making it mostly insensitive to jitter 
problems.

Danny


On 1/27/2017 6:18 PM, andy pugh wrote:
> On 27 January 2017 at 23:56, Gregg Eshelman  wrote:
>> If a PCs built in ports are too iffy to use with LCNC or other CNC software, 
>> will a PCI port card work?
> Given the price of a 5i25 / 6i25, why bother with an actual parallel
> port cards? They can drive P-port breakout boards just like a p-port,
> but can work with much higher latency and can also step in the MHz
> frequency range.
>
> The difference in price is a handful of broken milling cutters.
>


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] nema 24 motors.

[Danny Miller]

What drive and system voltage are you using here?

Toroids don't make any sense anymore.  Switching power supplies DO 
perform great, unless you  overload 'em with BEMF, but a toroid is 
vulnerable to that too.  Specifically Meanwell switching supplies 
perform reliably.

Resonance seems to be resolved AFAIK like 10 yrs ago with GeckoDrive, 
then GeckoDrive rendered obsolete by this DSP stuff, then THAT was 
rendered obsolete by DSP+3phase.  Of course this doesn't mean a stepper 
can do 10K RPM, but the torque does perform to the spec profile at least.

Mechanical dampeners are obsolete, unless you want a retro "steampunk" 
aesthetic.

There's one exception that DSP drives can't tolerate a 4th order system, 
i.e. a springy connection between the motor and load. Unless the motor 
has its own dampening.   But no one does a springy thing like that.  
Ordinary spring couplers don't count as springy for this purpose.

Yep full disclosure I drank one too many at the moment, but I'm 
instinctively compelled to get right to the point and help you with your 
resonance thing.  Get some 3ph and the 3DM683 and I think you'll start 
to cry.

Danny

> Its amazing the papers one can come across from a google search.  For
> instance, this one bears testing out:
>
> 
>
> Look at the last wiring diagram.  I may try this tomorrow with the small
> motor just for S  The theory is that it rotates the resting position
> by 1/2 of whatever the microstep angle is, with the real effect coming
> from there never being a condition where one winding is completely off.
>
> This should result in a smoother step whose absolute movement angle is
> more consistent from step to step regardless of the current setting,
> where with the upper two configurations, smoothness and equal angle per
> step performance is much more dependent on the current mapping in the
> driver matching the magnetic characteristics of the motor. I am going to
> put the nema 24 motor in but that depends on whether I had the great
> good sense to buy two of those 20 tooth pulleys, so I can bore the 2nd
> one to 8mm. It will be a bit under driven, the m542T driver does 4.2
> amps peak at full song, but this motor says 4.5 amp rms for max twist.
>
> Might have to buy a heavier driver & toroid power tranny to feed it.
> This one is heating some feeding the M542T 46 volts at its full current
> of 4.2 amps peak.
>
> News tomorrow night maybe.
>
>
> Thanks Danny.
>
> Cheers, Gene Heskett


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] nema 24 motors.

[Danny Miller]

Dude, drop the 2-phase bipolar stuff and move to 3 phase!  Or at least 
DSP drives.

http://www.ebay.com/itm/Nema-23-3phase-1-8N-m-255ozf-in-stepper-Motor-57mm-frame-8mm-shaft-57J1285-658-/262562493149?hash=item3d21f1b2dd:g:fMsAAOSwCfdXor7F

http://www.ebay.com/itm/LeadShine-3DM683-3-Phase-Digital-Stepper-Motor-Driver-20-60VDC-0-5-8-3A-NEW-/222183145504?hash=item33bb25c020:g:JU4AAOSw0kNXg2ZV

That's a programmable DSP drive.  That's a great combo.  3 phase's 
torque does not drop off until significantly later than 2-phase.  They 
run smoother and don't need dampening.  Torque numbers are not entirely 
equivalent since the 2-phase is typically limited by where the 
high-speed torque drops off, not the base torque.  There are larger ones 
available.

But programmable DSP drives alone do great things with 2-phase:

http://www.ebay.com/itm/1PC-Leadshine-AM882-Digital-Stepper-motor-Drive-80VDC-0-1-8-2A-protect-function-/252635288934?hash=item3ad23c9566:g:MUQAAOSwT6pVmTam

You go through a setup, tell it to auto-tune for that specific motor, 
and it's unlikely you'll ever have vibration issues.

Danny

On 1/23/2017 11:45 PM, Gene Heskett wrote:
> Greetings all;
>
> I see some "Nema 24" motors at quite reasonable prices showing up on
> fleabay, in fairly high torque ratings, so I bought one 10 days back,
> since I need a stronger one, with perhaps a damper on it on my x axis. I
> bought the 8 wire model so I can drive it in parallel where I ought to
> be able to get 3k rpms out of it. The mount bolt pattern appears to be a
> duplicated of nema 23, but the body is just a wee bit fatter, the shaft
> is double ended 8mm, and the whole thing looks like it will fit behind
> the new apron with about 5mm of clearance between the back shaft end and
> the face of the bed.  So its perhaps 5mm longer than the double ended
> nema 23 rated at 235 oz/in in there now.
>
> So I need to locate a 20 tooth, 8m bore pulley before I make the swap.
>
> Now, I don't believe this is going to help with the resonance stalls I am
> getting at about 30 ipm, so I'm thinking of building a viscous damper to
> fit on the rear of this puppy.  But unlike the ones on my hf micromill,
> which are long steel spools carrying a 2+ inch stack of heavy fender
> washers with elastomer sheets between the washers, which are true shock
> absorbers as the resonance is killed by the frictional losses as the
> washers walk on the talcum covered elastomer, but I don't have room for
> a 2+" stack of fender washers in this spot. 5/8" axially at best.
>
> I do have room for a larger diameter assembly on the back, so I am
> thinking in terms of a 1/2" thick by maybe 4" in diameter alu wheel,
> drilling in from the side at a low angle to put a couple set screws in
> to clamp it to the motor shaft, and knock down 1 or 2 of my 00 buck 12
> gauge rounds, measure them, and bore as many pockets in the rim as I can
> fit leaving about 20 to 30 thou for the balls to move within the pockets
> filled with grease, and sealed over. A thin gasket to contain the
> grease, and a 1/8" thick cap ring with a pair of 0-80 cap screws into
> the fillets between the pockets should make a good torsional vibration
> damper.  Or plow a groove for a 1/16" o-ring on the outside of the cap,
> and the inside radius of the cap ring on one side of it. Useing the
> o-rings would tend to make the grease self distributing, and the small
> clearance might even aid in the damping if I intentionally plow a
> shallow 3rd groove between the grooves for the sealing o-rings would
> further enhance the viscous losses of the grease moving back and forth
> between pockets thru that narrow passage.
>
> What do the vibration engineers in this crowd think of this idea?
>
> Cheers, Gene Heskett


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] New successor to the HB04 wireless mpg?

[Danny Miller]

I have all sorts of love for the HB04 wireless mpg.  I saw there's a new 
successor:

http://www.ebay.com/itm/CNC-MACH3-Wireless-Electronic-Handwheel-4-Axis-Manual-Controller-USB-Handle-MPG-/201698782859?hash=item2ef62f2e8b:g:whUAAOSwUEVYDeBn

Got mixed feelings.  A dot matrix LCD is AMAZING, if LinuxCNC could 
address it arbitrarily.  In the HB04, the display is limited, and half 
of it is taken up by displaying Machine Coords which is of no real 
value, and can't display the useful messages. e.g. if you press "Run" 
and the homing hasn't been done it'll mysteriously fail to react and we 
need to display that on the mpg, but there's no way to do that.

But man the button layout is crap.  This looks like a deal-breaker to 
me.  There's no symbols and color-coding except "reset" and it's very 
hard to read what they DID put on it.  I'm baffled my "feed +/-" and 
"spindle +/-" are buttons and not a handwheel mode like on the HB04.

And a "shift" key for overloading a second function on the mpg is just 
flat-out asking for trouble.  All too easy to hit the wrong function and 
that can easily break a job.

I wondered about making my own key overlay with new functions and 
symbols, and make it outstanding.  But I don't think I could make one 
really reliable, just print something out and put packing tape over it.

Still can't find a formal name for it, it's NOT "HB05" or anything.  In 
some places they say "UCONTRO"  but that's a vendor for breakout boards 
and stuff.  Has anyone made a LinuxCNC driver for it?


I did see where, in a different direction, there's "OpenXHC", where 
people made their own mpg hardware using the HB04's link, but the cheap 
handwheel hardware on the HB04 is "the good stuff" already.  I kinda 
feel like hacking the HB04 with the same hardware except a dot-matrix 
screen would be awesome, but realistically the work involved in 
designing custom hardware and a new protocol doesn't look like a 
rational use of time.

Danny



--
Developer Access Program for Intel Xeon Phi Processors
Access to Intel Xeon Phi processor-based developer platforms.
With one year of Intel Parallel Studio XE.
Training and support from Colfax.
Order your platform today. http://sdm.link/xeonphi
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Noise problem found, but not fixed.

[Danny Miller]

I can tell you this with confidence- the reliable, quality brand is 
"Meanwell".

They all look pretty much the same, they're not.

http://www.ebay.com/itm/Mean-Well-MW-5V-3A-15W-AC-DC-Switching-Power-Supply-NES-15-5-/351760561983?hash=item51e690373f:g:NbMAAOSwepJXYhR4

http://www.ebay.com/itm/GENUINE-MEAN-WELL-POWER-SUPPLY-ADAPTER-ES18A05-050-5V-3A-15W-max-100240v-/112229820233?hash=item1a216b2749:g:nssAAOSwA3dYGUnC

Meanwell isn't even notably more expensive.

Danny


On 12/28/2016 4:42 PM, John Thornton wrote:
> Hi Gene,
>
> Want to test one of my 5v 3a regulated power supplies?
>
> JT
>
>
> On 12/28/2016 4:33 PM, Gene Heskett wrote:
>> Greetings everybody;
>>
>> Now I am asking a question that I should know the answer to.
>>
>> It looks like, from the evidence so far, that this 5v 4a little
>> switchmode psu should be sent to the fcc and let them ban it from ever
>> coming off the boat or airplane on US soil.
>>
>> Its "static" ground terminal is common to the case, and I have now
>> grounded it by way of a couple flat braided ground straps about 5/16"
>> inch wide, back to the ground bolt in the bottom of the box. Ditto the
>> static ground on the z motors 60 volt psu. House static coming in is now
>> also bolted at that stud.  The scope probes ground lead is on that stud,
>> the clipon probe is clipped to one of the ioconnectors ground pins.
>>
>> That point is bouncing around 2.5v p-p with the switching activity in
>> that psu.  That bolt s/b common to the lathes bed casting thru the bolts
>> that mount it on a post bolted to the chip pan, which is braced by an
>> angle iron bolted to this post, with the end of the angle, about 8"
>> away, bolted solidly to the bed casting at the other end.
>>
>> I purposely did not connect anything in the at667 encoder wireup, to the
>> frame anyplace.  If I could find a good, real dirt ground for the scope
>> probe, I take bets that the whole latch is bouncing at least 25  volts
>> p-p to that real ground.
>>
>> The only way I can see ATM, is to install some "longitudinal" filters,
>> therefore chokeing that noise back into the psu instead of giving it
>> something to work against in the form of a solid connection to the
>> powerline.  I can source such a filter from Dave at the tv station as I
>> know he has a pile of failed monitor pcb's, all of which will have good
>> filters we can excavate.
>>
>> Or should I take the chance that the next one I buy on fleabay, would be
>> quiet?  Your call.
>>
>>
>> Cheers, Gene Heskett
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Calling MDI from HAL

[Danny Miller]

Say I want to use a button to call a one-time G-code op via MDI. 
Literally once.  Repeat calls will break things.

The button is debounced, but will remain pressed for more than one 
thread cycle.  In fact there's debouncing logic in HAL that yields a 
true/false.

Specifically it's a  "half" button to split the work coordinate by 1/2 
to set the zero at the center.  If you do it again, it's at 1/4, 1/8, etc.

MDI_COMMAND = G10 L20 P0 X[<#_x>/2]

How does this work?   Can I assign halui.mdi-command to booleans?  Do I 
assign the debounce output to "halui.mdi-command-18", or will that 
"break" as feared due to repeat calls unless it's also got a one-shot 
component?

Danny



--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] To BJT & PCW

[Danny Miller]

BTW, I DID grab this too:

https://www.amazon.com/Professional-Compression-Ratcheting-Wire-electrode-Die-Orange/dp/B00OMM4YUY/ref=s9_simh_gw_g469_i4_r?_encoding=UTF8=fresh_rd_m=ATVPDKIKX0DER_rd_s=_rd_r=0BMEGC66PM184FX0CHKE_rd_t=36701_rd_p=f58829ef-ff4e-4868-b5c8-fe8c3bd9f769_rd_i=desktop

And that seems like a much better tool, right?  Ratcheting and does both 
the wire and insulation crimps at the same time?  And it even SHOWS that 
same type of pin.

But I couldn't get it to work for CRAP on these Molex C-GRID III pins.  
Not sure why.  I suspect it wasn't intended for the smaller Molex C-GRID 
III, except it claims to work for "28ga".  IWISS is a great brand.

Danny


On 12/3/2016 7:33 PM, Danny Miller wrote:
> For those micro-pins, I had NO success with any crimper but this one,
> which worked GREAT:
>
> https://www.amazon.com/gp/product/B002AVVO7K/ref=oh_aui_detailpage_o05_s00?ie=UTF8=1
>
> Yeah it's stupid expensive for what it is,  but it worked for me, very
> well.  Perfect crimp on the copper and the insulation-grabber.
>
> Danny
>
> On 12/3/2016 12:47 PM, andy pugh wrote:
>> On 3 December 2016 at 18:18, Gene Heskett <ghesk...@shentel.net> wrote:
>>> Better, nearly as compact options? Suggestions aimed for limited overhead
>>> situations?  I'm all eyes.
>> http://www.digikey.com/product-detail/en/harwin-inc/M20-1071000/952-2027-ND/3727991
>> plus
>> http://www.digikey.com/product-detail/en/harwin-inc/M20-1160042/952-2157-2-ND/3728123
>>
>> You will need a rather small W-crimper. But then you can extract and
>> re-arrange the wires pretty easily.
>>
>> That's a 30-way. I tend to cut them to size, or glue together smaller ones.
>>
>> An example in use, but I don't normally use ribbon cable, that just
>> happened to be handy at the time:
>> https://goo.gl/photos/2pAhdHNDjoAD3Xg6A
>>
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] To BJT & PCW

[Danny Miller]

For those micro-pins, I had NO success with any crimper but this one, 
which worked GREAT:

https://www.amazon.com/gp/product/B002AVVO7K/ref=oh_aui_detailpage_o05_s00?ie=UTF8=1

Yeah it's stupid expensive for what it is,  but it worked for me, very 
well.  Perfect crimp on the copper and the insulation-grabber.

Danny

On 12/3/2016 12:47 PM, andy pugh wrote:
> On 3 December 2016 at 18:18, Gene Heskett  wrote:
>> Better, nearly as compact options? Suggestions aimed for limited overhead
>> situations?  I'm all eyes.
> http://www.digikey.com/product-detail/en/harwin-inc/M20-1071000/952-2027-ND/3727991
> plus
> http://www.digikey.com/product-detail/en/harwin-inc/M20-1160042/952-2157-2-ND/3728123
>
> You will need a rather small W-crimper. But then you can extract and
> re-arrange the wires pretty easily.
>
> That's a 30-way. I tend to cut them to size, or glue together smaller ones.
>
> An example in use, but I don't normally use ribbon cable, that just
> happened to be handy at the time:
> https://goo.gl/photos/2pAhdHNDjoAD3Xg6A
>


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] max_velocity during execution - multiple axis

[Danny Miller]

How does inverse-time work?

Say you specify 1 sec to do 1" for a single vector not connected to 
other cuts, so it starts/stop fully at the ends.

Is that the same as F=60ipm, which will take longer due to accel/decel?


Or does it guarantee that the whole cut will take 1 sec, which will go 
faster than 60 ipm in the middle?

Danny


On 11/6/2016 2:23 PM, andy pugh wrote:
> On 6 November 2016 at 20:09, Danny Miller <dan...@austin.rr.com> wrote:
>> Huh, ok, yeah inverse-time does make sense.
> A little less sense than not-inverse time. In my opinion.
>


--
Developer Access Program for Intel Xeon Phi Processors
Access to Intel Xeon Phi processor-based developer platforms.
With one year of Intel Parallel Studio XE.
Training and support from Colfax.
Order your platform today. http://sdm.link/xeonphi
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] max_velocity during execution - multiple axis

[Danny Miller]

On 11/6/2016 1:57 PM, Robert Ellenberg wrote:
> On Sun, Nov 6, 2016 at 2:46 PM Danny Miller <dan...@austin.rr.com> wrote:
>
>> I did some rotary stuff on Mach3 and was baffled by similar issues.
>>
>> Seems like it'd be CAM's job to manage the feedrate, and calculate for
>> the work radius.  That would make sense if you were cutting a cylinder
>> and the G-code move was "X moves 3 inches, A rotates 100 times, feed X
>> at 1.5 inches/min"
>>
>> But if the cartesians aren't moving- which is common- the F value has no
>> meaning at all if it's cartesian.  There is no way for G-code language
>> to communicate "polar Feed =  200 deg/sec" and it's nonsense to combine
>> polar and cartesian vectors into a single scalar for a feedrate.
>>
> The F word is overloaded. If there's no cartesian motion, then it treats
> the number as degrees per min instead of in per min.
That sounds like a bad idea.  The interp of F jumps around to a totally 
different thing if there's no cartesian motion?
So if you MDI in  "G1 X1 A200 F3" it does one thing if you're at X=2 and 
a totally different thing if you're at X=1?  Or does the inclusion of 
any linear axis mode it over to cartesian?  Well, but, you couldn't do 
that- if you're already at X=1, F3 is all polar and it's impossible to 
interpret as a linear feedrate since there's no linear motion.

What if X moves 0.001" and that gets optimized out?
>
> Seems like the logical answer is amending G-code Feedrate with syntax
>> for an angular vector in addition to cartesian, but no CAM pkg would
>> support it.
>
> The "proper" way as I understand it is to use inverse time when doing both
> linear and angular moves. It's perfectly sensible to say "move X inches and
> A degress within T seconds".

Huh, ok, yeah inverse-time does make sense.
>


--
Developer Access Program for Intel Xeon Phi Processors
Access to Intel Xeon Phi processor-based developer platforms.
With one year of Intel Parallel Studio XE.
Training and support from Colfax.
Order your platform today. http://sdm.link/xeonphi
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] max_velocity during execution - multiple axis

[Danny Miller]

I did some rotary stuff on Mach3 and was baffled by similar issues.

Seems like it'd be CAM's job to manage the feedrate, and calculate for 
the work radius.  That would make sense if you were cutting a cylinder 
and the G-code move was "X moves 3 inches, A rotates 100 times, feed X 
at 1.5 inches/min"

But if the cartesians aren't moving- which is common- the F value has no 
meaning at all if it's cartesian.  There is no way for G-code language 
to communicate "polar Feed =  200 deg/sec" and it's nonsense to combine 
polar and cartesian vectors into a single scalar for a feedrate.

Seems like the logical answer is amending G-code Feedrate with syntax 
for an angular vector in addition to cartesian, but no CAM pkg would 
support it.

Then again, the alternative is soft-declaring an axis and calculating 
surface speed and interpret the F as surface speed. Which would also not 
be supported by CAM AFAIK.  Well how DO commercial 4/5 axis CAM tools 
declare feedrates?   I presume they have a solution, as the tech's been 
in wide commercial use for a long time.

Danny


On 11/6/2016 11:48 AM, Robert Ellenberg wrote:
> As of now, circular arc blending doesn't work with the ABC axes (which is
> why it's running so slowly for you with short segments), but I'm working on
> a fix. I had an extensive discussion with Andy Pugh a while back, which led
> to some great ideas on how to solve this problem. There are two big reasons
> it doesn't work now:
>
> 1) ABC units are degrees, whereas linear axes (XYZUVW) are in distance. A
> spherical arc doesn't make physical sense between axes of different units.
>
> 2) "Velocity" as defined by the TP is actually velocity in either XYZ, ABC,
> or UVW axes, depending on the context, but a spherical arc in general would
> involve change in all 9 axes, so the arc length (and it's derivatives, TP
> velocity and acceleration) needs to be expressed in terms of all axes.
>
> What I'm working on now is to treat ABC axis motion as an equivalent tool
> tip motion in distance units. For example, if the tool is 10 inches above
> the physical A axis, then the A axis has an effective radius of 10 inches.
> Therefore it's easy to convert the angular motion to an equivalent tool top
> motion. Ex: a 90 deg rotation would be similar to a linear move of pi / 2 *
> 10 ~= 15.71 in.
>
> The good news is, if we assume some constant effective radius for all ABC
> axes, I think it's possible to implement with the same basic structures as
> the ones in my my experimental UVW axes blending branch
> 
> .
>
> The hard part is making the effective radius change with tool tip position.
> That's my long-term goal with this fix, but it may still be a big
> improvement to assume a constant effective radius.
>
> Best,
> Rob
>
> On Sun, Nov 6, 2016 at 10:51 AM Tomaz T.  wrote:
>
> So for now there is no solution to speed this up till TP will be modified
> in the way that it will take and "optimize" also moves with rotary axis ...?
>
>
> PS. yes I'm from Slovenia.
>
>
> 
> From: klemenzivko...@gmail.com  on behalf of
> Klemen Zivkovic 
> Sent: Sunday, November 6, 2016 2:10 PM
> To: tomaz_...@hotmail.com
> Subject: [Emc-users] max_velocity during execution - multiple axis
>
> Hello,
>
>
> I think your problem is that linuxcnc have x,y,z trajectory planning. As
> soon as you add rotary axis to have combined move you end up with single
> lookahead in tp, so this limits velocity.
> Check this out:
> https://forum.linuxcnc.org/20-g-code/29662-coordinated-motion-involving-rotary-axis
>
>
> According to your name - I need to ask you are you native slovenian speaker?
>
>
> regards
> KLemen
> --
> Developer Access Program for Intel Xeon Phi Processors
> Access to Intel Xeon Phi processor-based developer platforms.
> With one year of Intel Parallel Studio XE.
> Training and support from Colfax.
> Order your platform today. http://sdm.link/xeonphi
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
> --
> Developer Access Program for Intel Xeon Phi Processors
> Access to Intel Xeon Phi processor-based developer platforms.
> With one year of Intel Parallel Studio XE.
> Training and support from Colfax.
> Order your platform today. http://sdm.link/xeonphi
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Developer Access Program for Intel Xeon Phi Processors
Access to Intel Xeon Phi 

Re: [Emc-users] An observation on digital calipers

[Danny Miller]

I tried to buy a pair of HFT calipers a few years ago, found they didn't 
read right, lost large distances.

Went to exchange them but thought "hey let's make sure I get a pair that 
works..."

NONE OF THEM.  Went through like 10 boxes on the shelf.  None read 
correctly.

Now I've had an HFT I bought like 12 yrs ago that worked GREAT.

But, anyhow, Neiko calipers are just as cheap and work great now:

https://www.amazon.com/Neiko-01407A-Electronic-Digital-Caliper/dp/B000GSLKIW/ref=sr_1_1?ie=UTF8=1478415113=8-1=neiko

Danny

On 11/6/2016 12:01 AM, Chris Albertson wrote:
> On Tue, Nov 1, 2016 at 1:54 AM, andy pugh  wrote:
>
>> How cheap are HF?
>> http://www.machine-dro.co.uk/left-handed-digital-calipers-150mm-6.html
>
> HF sells the right hand version of that exact same caliper for $19 and they
> always have discount coupons.  I paid $15 for one.
>
> I've never had a problem with handedness, you can always flip them over,
> hold them upside down or whatever


--
Developer Access Program for Intel Xeon Phi Processors
Access to Intel Xeon Phi processor-based developer platforms.
With one year of Intel Parallel Studio XE.
Training and support from Colfax.
Order your platform today. http://sdm.link/xeonphi
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] What is usually used in place of rotating memory on one of these "Pi" boards?

[Danny Miller]

You wouldn't likely save the g-code on the SD card at all.

You'd have a wifi connection and load the g-code from your local network.

Those SD cards are used for storing photos and video quickly. 
Photographers and videographers use TONS of memory over and over., and 
quite frankly I never heard of someone wearing one out.

This isn't a drop in the bucket compared to those high-stress uses. It 
will go obsolete long before the SD card wears out.

They do employ load-leveling where it won't keep resaving over the same 
block of memory.  The file structure is unaware of it but if you keep 
rewriting a cache file, it moves it up and down across the whole SD card 
slowly, so it never wears out.

Danny

On 10/27/2016 9:35 PM, Chris Albertson wrote:
> The SD card will work for file storage but just remember a class 10 SD card
> will write at 10MB per second or 80Mbps.
>
> The write speed to the cache on a normal rotating disk is 6Gb or 6000Mbps.
> That is about two orders of magnitude better but that is "cheating" as it
> is only the interface speed.  The sustained sped of the hard drive is only
> about six times faster than the SD card.   Either way the I/O performance
> of a Pi is rather disappointing.   What people do to work around this is
> keep the files on the hard drives of a "real" computer.  They edit those
> files using the real computer so it is as fast as you are used to.   Then
> when it comes time to run those files they are NFS mounted.  So basically
> you do all the work on the real computer and keep the data there too.
>
> These Pis are powerful computers for many uses.  They can stream a movie to
> your TV set. But look at what that means, it takes an hour and a half to
> transcode and stream a movie.  That is so long you could watch a movie
> while you wait.   My point is not to expect a fast interactive experience
> but when you go to cut metal it will run fast enough to make the mill run
> at full speed.
>
> On Thu, Oct 27, 2016 at 5:36 PM, Gene Heskett  wrote:
>
>> On Thursday 27 October 2016 18:57:15 W. Martinjak wrote:
>>
>>> On 2016-10-27 23:07, Gene Heskett wrote:
 I am getting the impression that a micro-sd would have both poor
 loading performace, call a surveyor to measure write speeds. And
 poor life in a filesystem environment. SSD w/sata would be good, but
 the sata on a pi is a bad kludge from what I'm reading.

 So, Bari, Charles S., etc, what are you folks using?
>>> Hi Gene,
>>>
>>> it depends on which performance records you want break. ;)
>>>
>>> I suggest going to the next mall, buy a raspberry pi 3, three fast
>>> micro-sd cards (1st for testing, 2nd for backup and the 3rd for the
>>> running end version)
>> Ya convinced me, but the next mall, from out here in the puckerbrush of
>> WV, is amazon.com.  Its on the way, with free 2 day shipping. 3 32Gb
>> class 10 micro-sd's too.  That should wear level for a while. :)
>>
>>> with 16GB (space for wear leveling) and a fast sd card reader.
>>> As I've heard the 7i90 is on the way.
>>> Equipped with these arms you can start cnc'ing with pi sized
>>> ordinateure. :) Then I/we can assist you to setup the whole stuff.
>> I'll yelp when it has arrived, probably Monday now as I am tied up and
>> being held for ransom tomorrow afternoon at the dealer's shop where I
>> bought a 2007 Toy Rav4, about 5 years and 40k miles back & the brakes
>> are getting funkity from stuck calipers at 65k miles.  And they will
>> want a fee (the ransom) for that.  And I've a window regulator
>> (electric) to replace in my GMC before it rains seriously.  Sunshine
>> promised over the weekend. :)  Now, if my back will co-operate...
>>
>> Thanks.
>>
>> Cheers, Gene Heskett
>> --
>> "There are four boxes to be used in defense of liberty:
>>   soap, ballot, jury, and ammo. Please use in that order."
>> -Ed Howdershelt (Author)
>> Genes Web page 
>>
>> 
>> --
>> The Command Line: Reinvented for Modern Developers
>> Did the resurgence of CLI tooling catch you by surprise?
>> Reconnect with the command line and become more productive.
>> Learn the new .NET and ASP.NET CLI. Get your free copy!
>> http://sdm.link/telerik
>> ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>>
>
>


--
The Command Line: Reinvented for Modern Developers
Did the resurgence of CLI tooling catch you by surprise?
Reconnect with the command line and become more productive. 
Learn the new .NET and ASP.NET CLI. Get your free copy!
http://sdm.link/telerik
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Moving past Home Switch

[Danny Miller]

There are multiple parameters:

1.  You can declare the homing switch to be a coordinate other than 0.  
e.g. "after we settle on the switch, this axis is offset so this is now 
+1.25 inches"

2.  You can set axis limits at anything.  I probably would never set a 
negative coord as a limit but maybe that's just me.

3.  You can send the axis anywhere after homing.   e.g. "settle on the 
homing switch, call that +1.25", then move to 0.5" and leave it there."

Danny


On 10/25/2016 7:07 PM, Mark Johnsen wrote:
> Just guessing, but you might have software limits set to 0.  That'd be a
> configuration issue, I'd guess the .ini file or the main *.hal file would
> have those in it.
>
> You should be able to travel past the home switch.  On a side note, I have
> no home switches and will 'manually' home in Axis by clicking the home
> button, but I'm not sure it does any good and in my config, I ahve it setup
> w/ no homes as well as only 1 limit switch for each ends of an axis.  I do
> have software limits set and they're set w/ the idea that I'll home
> somewhere near the middle of my axes.  (except for z).  Anyway, look at
> that.
> Mark
>
> On Tue, Oct 25, 2016 at 4:04 PM, hubert  wrote:
>
>> Is there any way to move past home position.
>>
>> Each of my Axis, X, Y, and Z use three sensors to provide inputs for min
>> and max limits and Homing.  We have the Home switch set about 1 inch
>> inside the max limit switch.  However, We need to raise the Z axis above
>> the home switch but inside the limit switch for the head to clear the
>> tool changer.   Also for some parts we need to make use of more of the
>> table than the Home switch allows.  Is this a miss-configuration problem
>> or a deficiency? When we try to move past the home switch the machine
>> refuses to move.
>>
>> Hubert
>>
> --
> The Command Line: Reinvented for Modern Developers
> Did the resurgence of CLI tooling catch you by surprise?
> Reconnect with the command line and become more productive.
> Learn the new .NET and ASP.NET CLI. Get your free copy!
> http://sdm.link/telerik
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
The Command Line: Reinvented for Modern Developers
Did the resurgence of CLI tooling catch you by surprise?
Reconnect with the command line and become more productive. 
Learn the new .NET and ASP.NET CLI. Get your free copy!
http://sdm.link/telerik
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Machinekit? --> LinuxCNC computer

[Danny Miller]

On 10/24/2016 3:23 AM, Gene Heskett wrote:
> Opto's on the bob are a waste of protection with modern
> stepper drivers as they all have their own optos for all inputs.
> Properly driven, a 2M542 driver can exceed 350 kilohertz step rates. Put
> an opto on the bob and it will often fall over, stalling the motor at
> half that. One of them in series with a 10 kilohertz pwm signal and the
> control linearity error over powers the PID's ability to keep the motor
> within 20% of the requested speed. I found which one it was, pulled it
> and bridged the 2 pins to bypass it, and now I get within 5% of the
> requested speed without a lincurve module in the signal path, and the
> PID handles that nicely.
>
> Cheers, Gene Heskett

I totally agree with you on optoisolation, actually.   Shoot, the whole 
Mesa board is only $79, and since ethernet is galvanically isolated by 
design, it CANNOT damage the computer.  Why would I spend $40 on an 
opto-bob on the theory that there's a 2% chance a "mystery power surge" 
would destroy it??  I can't really see where you'd even get a "power 
spike".

But the AM882 (GREAT drive at a bargain price, BTW) has 5v differential 
opto inputs, which makes them noise-immune to induction when no common 
is used, and ground noise, since there's no ground. True diff pair.  
Again, probably was never gonna be a problem anyways.

Second, I ended up with the NPN proximity sensors shown here on the top 
left:

https://oceancontrols.com.au/IBS-0051.html

Those are insidiously awkward to interface, honestly, simplest, best 
practice is an opto.  Anode w/series resistor to +V (I used 24v 
system),  cathode to "output".

Also you can form an AND/OR gate with optos or also fanouts, and I used 
that.  The e-stop switch not only pulls down the input pin, but also is 
in series with the output pin that feeds the drives' Enable pins.  If  
the estop is opened, the opto output stops conducting, and the drives 
disable.

Danny




--
The Command Line: Reinvented for Modern Developers
Did the resurgence of CLI tooling catch you by surprise?
Reconnect with the command line and become more productive. 
Learn the new .NET and ASP.NET CLI. Get your free copy!
http://sdm.link/telerik
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] following along in my what if musings about ther new oramge pi 2e

[Danny Miller]

It absolutely requires its own ethernet port, no router either. That's 
because it's a realtime device and having to arbitrate would only make 
unpredictable latency/jitter.

Like others are saying, the link for accessing your LAN probably won't 
even suffer if you use wifi and a USB2.0.

I just used a Edimax Wireless N150 Wifi Nano.   $10.  People use 'em all 
the time for Raspberry Pi.  Immediately worked on Linux.

No speed complaints.

Danny


On 10/24/2016 8:52 PM, Gene Heskett wrote:
> Greetings guys; I hope everyone has arrived back home without incidents
> involving bent sheet metal or worse.
>
> 1. someone said the 7i92H needs its own dedicated ethernet port,
> presumably because udp patckets are not subject to any attempts at error
> correction resends if other traffic walks on a udp packet.
>
> So that means I'd need to find an Orange Pi with 2 ethernet sockets so
> two independant paths/addresses can be setup.
>
> None of these low power use cards have that, none that I've found have a
> 2nd port.  Is this a case of just waiting till it does happen?  Space
> considerations for the rj45 socket says it not going to be at all
> likely.
>
> Comments anybody?
>
> 2. In playing with this phony vfd today after I got everything bolted
> down again, I find the lack of docs a good sized problem. My test code
> loop to exercise it has a direct reversal in it at one point and I'd
> like to decelerate it to about 25 hertz in 5 to 7.5 seconds from
> whatever speed its turning, switching on the dc brakes at that point to
> bring it to a smooth halt.  Then accelerate smoothly in the other
> direction. It acts like that is what its doing for a straight stop from
> either direction. and the booklet says in can go directly from one to
> the other. However, in trying to speed up the changes, if I just switch
> directions, it gets down to about 25 hz, then jumps off the table about
> an inch, and is then running the other direction at the set speed even
> if the set speed is 120 hz!  No accelleration softening ramp at all.
> This controller has about 8 registers that can be set for various speeds
> where they are in effect, but zero discussion about what they effect in
> this little booklet. So how to go about optimizing those settings?
>
> If anyone has any wisdom to share, I'd sure appreciate it.
>
> Thanks everybody.
>
> Cheers, Gene Heskett


--
The Command Line: Reinvented for Modern Developers
Did the resurgence of CLI tooling catch you by surprise?
Reconnect with the command line and become more productive. 
Learn the new .NET and ASP.NET CLI. Get your free copy!
http://sdm.link/telerik
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Machinekit? --> LinuxCNC computer

[Danny Miller]

I don't know why you'd want to use that when the 7i92 is available.  
You're using a parallel port or serial, which has significant bandwidth 
and latency issues compared to ethernet. 7i92's only $30 more so I doubt 
cost is a big factor here.


But, pretty sure it'll work with LinuxCNC.  For 7i92, you loadrt 
HOSTMOT2 which creates HAL pins for all the functions.  You do need to 
use a command-line tool "mesaflash" once to load the hardware bitfile 
image into the FPGA.  There are several bitfiles available, they 
dedicate specific pins to be stepgens, encoders, etc.  You can turn off 
specific functions and turn them into GPIO, but you can't take GPIO and 
assign a new stepgen to it, because that involves new FPGA code.

Danny


On 10/24/2016 12:18 AM, Chris Albertson wrote:
> Does the 7I90HD work with Linux CNC?   It looks like it could work
> It looks like it would be the board to use with an ARM Pi3.  It
> connects with the host using SPI which is easy to use in the Pi3
> 7I90HD is the lowest cost Mega FPGA board at only $59 so you could
> have the computer and Mesa card both for under $100 and the set of
> cards would fit in the palm of your hand.
>
> There is an Ethernet version of the FPGA card too but it is much more
> expensive and SPI might be even faster as there is far less overhead.
> SPI can't go very far, just a few inches but
>
> What about labor problems as the ports?  Spomthing like this would
> come in air freight. Almost all electronics
>
>
>
> On Sun, Oct 23, 2016 at 8:37 PM, Danny Miller <dan...@austin.rr.com> wrote:
>>
>> On 10/23/2016 4:28 PM, Gene Heskett wrote:
>>>> No biggie, on my CNC (not an SBC) I just added a wifi USB already.
>>>>
>>>> Mesa 7i92 are totally available from the mfg:
>>>>
>>>> http://store.mesanet.com/index.php?route=product/product_id=30
>>>> 2
>>> Not according to that web page 2 hours ago. Out of stock. But this link
>>> say 170 on the shelf? Left hand, please call right hand. :)  Either that
>>> or the boat finally docked, I understand they've got labor problems at
>>> some west side ports.


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Machinekit? --> LinuxCNC computer

[Danny Miller]

> Sounds like its the -H I'd want too. More conserving of horizontal real
> estate.  And while I'm not going to add yet another potential loose
> connection to the home switch setup, these bobs I have have such poor
> terminals that I will cobble up a distribution point for all the stuff
> that needs 5 volts. 2 bobs and 2 points on the opto so far as 5 volt
> loads. As for home/limits, these cheap bobs, intended for a Mach system,
> already have pullups so all they'll have to do is pull to ground with a
> microswitch.
>
> Thanks Danny, for calling that board to my attention.
>
> Cheers, Gene Heskett
Some FYI-

Mesa cards use a 3.3v FPGA, but you can enable a bidirectional bus 
switch which can convert to "mostly" 5v IO, but it's not a true 5v.

Also ALL IO has pullup resistors on the Mesa, which only have an effect 
as inputs.  Outputs are not open-collector, they're driven, but a bus 
switch's driven "high" is less than the 5v Vdd.

This worked quite well for me.  The AM882H has 5v opto-isolated IO, with 
integral resistors (so it's gotta be 5v to generate specified opto drive 
current), and they didn't common the anode or cathode. The 
step/dir/enable I fed by tying the anodes to 5v and the cathodes to the 
Mesa IO pin because drive-low is definitely 0v but drive-high isn't 
quite 5v.

And the IO is just pulldown.  My board's got a  bunch of optoisolators 
and their outputs are just pulldowns against the Mesa inputs' built-in 
pullups- which is simple.

BTW, about "loose connectors"-  this ferrule stuff CHANGED MY LIFE on 
clamp-down-on-stranded-wire:

https://www.amazon.com/uxcell-AWG10-Insulated-Ferrule-Terminal/dp/B00UBUR6E2/ref=sr_1_6?ie=UTF8=1477283614=8-6=ferrule

https://www.amazon.com/Signstek-Adjusting-Ratcheting-Crimping-AWG23-10/dp/B00HPRYIL8/ref=sr_1_5?ie=UTF8=1477283614=8-5=ferrule=1

Seriously, it's amazing.  I learned not to trust stranded-on-clamps.  
Strands fray and short, they break, they work out strand-by-strand.  And 
having to rework it damages the wire. I've tinned stranded wire in the 
past, only to have it get loose later and wires fell out, and Googled 
this- most people say "never tin, it's too soft and gets loose and your 
wire falls out".  Oh. Kay.

This stuff is cheap and quick, it makes a 100% solid strain-relieved 
bond.  Phoenix connectors hold them easily and they never let go.

Use that 4-jaw tool, NOT the 6, if they're going in clamp connectors.  
Only 2 surfaces get used, so square maximizes the surface area that gets 
used.

I showed this stuff to my hacker buds and they were all "OMG where has 
this been all my life???"

Danny





--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Machinekit? --> LinuxCNC computer

[Danny Miller]

On 10/23/2016 4:28 PM, Gene Heskett wrote:
>
>> No biggie, on my CNC (not an SBC) I just added a wifi USB already.
>>
>> Mesa 7i92 are totally available from the mfg:
>>
>> http://store.mesanet.com/index.php?route=product/product_id=30
>> 2
> Not according to that web page 2 hours ago. Out of stock. But this link
> say 170 on the shelf? Left hand, please call right hand. :)  Either that
> or the boat finally docked, I understand they've got labor problems at
> some west side ports.

The DB25 FEMALE (no suffix) version is out of stock.
They have plenty of "H" (two 26pin locking DIL) and "M" (DB25 male).

Personally I recommend the "H".  It's locking.  I'm not gonna use a 
remade round cable, the board's right there with the controller so I'd 
make up an IDC ribbon cable for that anyways.

Me, I built a motherboard that provided power, broke out into individual 
DIL 10-pins for the AM882 drives, and broke out pins for homing switches 
etc.

>
>> There are 3 versions which change the port on the end: male DB25,
>> female DB25, male 26-pin DIL header.  But the one in the middle is
>> always a male DIL26 header.
> And the 7i92M means its a male db25 on the end?
Yep!
>
 Gene Heskett


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Machinekit? --> LinuxCNC computer

[Danny Miller]

I use the Mesa 7i92.  It is EXCELLENT.  I prefer it over PCI solutions 
just because I can keep the controller separate from the PC without a 
lot of wiring.

It works fine with my 1ms servo-thread.  I have no reason to see if it 
goes further.  It offloads all the pulse generation for the steppers so 
it doesn't require a lot of the computer.  The recommended setup uses a 
PLL which compensates for jitter down the ethernet link, it's probably 
overkill but it's just a HAL component to add.

I don't know what other issue I'd run into with latency in a stepper 
system.  I had no problem at all.  I don't know how it'd go with servo 
encoders.

I discovered it does NOT work with all ethernet ports, my MB had a very 
odd onboard ethernet chipset and LinuxCNC didn't start at all.   Adding 
a PCI ethernet card was an easy answer for me, but not for someone  with 
that SBC with no PCI, but I doubt that will come up.

It does require you to dedicate that ethernet port.  No routers. Since 
that SBC has only one ethernet port, your network access would have to 
be through USB.

No biggie, on my CNC (not an SBC) I just added a wifi USB already.

Mesa 7i92 are totally available from the mfg:

http://store.mesanet.com/index.php?route=product/product_id=302

There are 3 versions which change the port on the end: male DB25, female 
DB25, male 26-pin DIL header.  But the one in the middle is always a 
male DIL26 header.

Danny


On 10/23/2016 2:07 PM, Gene Heskett wrote:
> On Sunday 23 October 2016 13:54:59 Dave Cole wrote:
>
>> Connect to a Mesa ethernet controller?
>>
>> Dave
> What about the latency?  Can it match the 1 to 4 kilohertz servo-thread
> loop I have used?
>
> It's no help that the 7i92 is out of stock on both Peters and Johns web
> pages.  How long to source one of those critters?
>
> Or that what I think would be the next step up, as in skipping the db25
> breakouts in favor of direct i/o is the 7i76E @ $200.
>
> So I guess I am stuck, drooling over the possibilities, but still using
> an old Dell I paid $75 for, with a single core P4 that burns 90 watts,
> and a 5i25, in a box that occupy's at least a cubic foot to run this
> lathe. There IS a limit to the relatively clean space under the drawers
> in this Sheldon cabinet, so the motor driver box I am building up from
> an old rusty box will have to live out where swarf will collect on top
> of it.
>
> I do like the ethernetted idea because theres little effect to moving the
> computer to a cleaner location, a most desirable condition.  Sigh.  And
> not helped that my giddyup got up and left me needing another nap this
> morning. But I think I can go cobble up some hole locations and get this
> stuff mounted yet today.
>
>> On 10/23/2016 1:46 PM, Gene Heskett wrote:
>>> On Sunday 23 October 2016 12:02:58 Roland Jollivet wrote:
 By the way, while you guys are discussing SBC's, there's also the
 UP board  too, which is x86.

 Here's a fun video  on
 it too. Worth looking at...
>>> Nice board, pricy tho at a 90 buck starting point. And no pci buss,
>>> so nothing from pico or mesa is usable. Whats available to plug onto
>>> the 40 pin pi header that can duplicate a mesa anything i/o
>>> functionality?
>>>
>>> C'mon crickets, speak up.  I looked thru that blog post link to see
>>> what might be found only to see that decently capable boards for the
>>> BBB like the k9 have been discontinued for lack of interest.
>>>
>>> This is not how we move linuxcnc to the next generation hardware.
>>>
>>> Cheers, Gene Heskett
>> --
>>  Check out the vibrant tech community on one of the world's
>> most engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>> ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>
> Cheers, Gene Heskett


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] glue question?

[Danny Miller]

On 10/23/2016 2:30 AM, Gene Heskett wrote:
> I considered dropping the face to be joined on my big bench belt sander
> for about long enough to make the surface good and "hairy", but haven't
> actually tried that. About a second maybe as the melting point is
> reached very quickly and I'd want it badly scuffed without melting.
>
> Cheers, Gene Heskett
I know far too much about this, from autism-inspired tests.

Sanding only increases the bondability SOME.  It won't be anywhere near 
enough to be "great", it will never approach the strength of the parent 
material.  Alcohol, thinner, MEK, acetone, etc have no real value except 
in cleaning off residue.  Those work great for ABS or acrylic, they 
partially dissolve the surface leaving it tacky, etched, and the 
polymers are "open".

But both ABS and acrylic are a "high surface energy plastic".  LOW 
surface energy plastics include nylon, teflon, HDPE, LDPE, 
polypropylene, UHMW, PET, Mylar, and others.  Glues and paints just 
don't stick to anything in this category and there's no easy answers 
except "just don't use glue".  Their hallmark is chemical resistance and 
things NOT sticking to them.

Danny





--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] glue question?

[Danny Miller]

You CANNOT expect any high-strength bond on HDPE that rivals the 
strength of the material, not with any glue.  Typically it is bolted.

E6000 (same an Amazing Goop except uses PERC solvent instead of toluene) 
is the highest strength.

Locktite makes a system for cyanoacrylate where you use a primer to 
activate it and then use this superglue stuff.

HDPE can be thermally/ultrasonically welded to more HDPE.  UHMW, that 
doesn't really work because it has a melt index of 0.  If you put it in 
an oven, UHMW becomes a semi-transparent floppy sheet- but it DOES NOT 
melt before it starts decomposing.  Its chains are so long and entangled 
it can't become a liquid and remain UHMW.

Now I do know this.  Society of Creative Anachronism uses UHMW blocks on 
fiberglass arrow shafts for live-combat arrows.  To do this, they bore a 
tight, deep hole that requires compression to seat the shaft and use 
E6000.  The joint never goes out of compression and has a lot of surface 
area in shear alone.  That holds for life, no matter what.  But you 
really won't have that scenario.

Danny


On 10/22/2016 5:22 PM, Bruce Layne wrote:
> In general, no matter what your glue question is, my glue answer is epoxy.
>
> HDPE is almost as difficult to glue as UHMW or Teflon.  Very low surface
> energy.  I'd recommend cleaning with isopropyl alcohol to remove every
> trace of oil from the surfaces (HPDE and panel), allow it to air dry,
> and then use a liberal amount of RTV silicone.  You seemed to be heading
> in that direction with your Goop musings. Don't be stingy.  Slop it up
> the sides of the HDPE if possible so it has more surface to grip and
> it's protected from shear loads, which are the enemy of adhesives.
>
> I've had good luck mounting stuff like PC power supplies and solid state
> hard drives inside a panel with 3M VHB double sided tape.  VHB = Very
> High Bond.  It's the stuff the auto industry uses to attach automotive
> trim strips to prevent automated car washes from yanking them off.  Even
> then, I try to mount these components on the base of the cabinet so
> gravity is helping the adhesive bond rather than trying to peel the
> device off the vertical side of the panel.  I use powder coated steel
> cabinets, and that's a slick surface that doesn't bond well.  Even
> worse, it's a crinkle finish so there's about half the surface area or
> less for the adhesive bond to grab.
>
>
>
> On 10/22/2016 05:47 PM, Gene Heskett wrote:
>> Greetings all;
>>
>> I am in the process of re-arranging the peripheral layout on the inside
>> face of the box that holds the BoB's, opto's etc on the boxes swinging
>> door plus a couple power supplies and stepper drivers in the box proper.
>>
>> As opposed to drilling umpty holes in this cover, I've sawn out some
>> strips of HDPE, and screwed the boards to those.  Its nominally 1/2"
>> thick and I have not, but could, sand the face to be glued.
>>
>> So now I need to fix them to the inside face of this boxes door, and plug
>> up the holes I have drilled already in a probably vain attempt to
>> exclude swarf. Except for some conduit holes in the bottom, its gasketed
>> and liquid tight.  So I need a glue that will stick to this HDPE.  G0-2
>> or Goop is one possibility, but is there something that will get an even
>> better grip on that stuff?
>>
>> Thanks.
>>
>> Cheers, Gene Heskett
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] New stepper-servo --> Phase delay

[Danny Miller]

The SPI comm of course has latency within a comm period.

However, the pulse output must be super-fast.  It supports up to 14500 
rpm, so if it's 2000steps/rev the whole step cycle is 2ms, realistically 
like maybe 1/4 of that for latency?

Danny

On 10/19/2016 11:34 PM, Nicklas Karlsson wrote:
>>> ... The lag also seemed to vary with velocity.
>>>
>>>
>> Most specifically, it was a lag in the encoder's loop to respond to
>> acceleration.
>> Velocity was fine, but when there is acceleration, the encoder's
>> velocity didn't change for a few milliseconds, then it had to produce
>> velocity greater than real for position to catch up.
>> I thought it was a really severe problem, but it seems it may not be all
>> that bad.  But, it can make servo tuning more tricky.
>>
>> Jon
> Delay is called phase in control theory and it is not good at all for fast 
> response.
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Machinekit?

[Danny Miller]

But if you're doing that, then a Mesa FPGA card makes much more sense 
for the realtime component.

Danny


On 10/19/2016 11:26 PM, Nicklas Karlsson wrote:
>> ...
>> While I use many BeagleBones to control various machines exactly the
>> way you describe (using an HDMI monitor and KB/Mouse connected to the
>> BBB), it is not nearly the same user experience as running on an x86
>> PC.  Everything is noticeably slower on the BBB, and graphics
>> performance is particularly horrid (to the point that the 3D preview
>> display is essentially unusable).
> Sound like the real time stuff on Raspberry or BBB and an ordinary computer 
> for user interface is the way to go for all systems with an ordinary user 
> interface.
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] New stepper-servo

[Danny Miller]

A friend showed me this today, not powered up yet:

https://www.kickstarter.com/projects/tropicallabs/mechaduino-powerful-open-source-industrial-servo-m

At first I was confused, it says it's a stepper, but with an encoder, 
and lowers or disables drive current when not needed.  Or you can 
freewheel it and it'll maintain the coordinate system.  It cannot stall 
without the system knowing, and a stall won't corrupt the coordinate system.

But there's no mechanical connection to the motor, it's this new AS5047D 
high speed, high resolution magnetic rotary position sensor:

http://ams.com/eng/Products/Magnetic-Position-Sensors/Angle-Position-On-Axis/AS5047D

You glue a magnet to the rear of the motor shaft, and keep the sensor 
like 1/4" away.  Note it not only counts delta, it knows absolute rotor 
position.

Gives 2000 counts/rev.  That's enough that it could effectively show 
rotor phase through a single fullstep, which would allow for more stable 
electrical control of stepping.  There's an initial calibration phase 
that drives a phase so it knows where the TDC of Coil A is in absolute 
position.

They set it up so it powers down and freewheels, except it will hold its 
position by powering coils when needed.  It can also reduce current to 
product only the torque needed to follow the step commands, instead of 
always operating at full current.

Well, my mind is blown!

Danny



--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] converting from mm to inches

[Danny Miller]

When you change the machine units to inches, you need to change all the 
velocity and accel parameters.  They're just numbers, the unit is 
assumed, and changing from mm to inches makes that number 25.4x slower 
outright.

Danny


On 10/17/2016 10:05 PM, hubert wrote:
> My machine was initially setup by my vendor to mm.  Personally I prefer
> inches.  When I tried to adjust things in the .ini file it results in
> the machine crawling.  Is there another place I need to make changes.  I
> changed units from mm to in.
> Thanks
> Hubert
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Hacking a sim runtime mode

[Danny Miller]

Ah, I was about to say "no, it's just 100x the accel, not 100x^2 the 
accel, we're just changing the time frame".

But I started to question myself and did the math logically. Duh.  YOU 
are right, I've had that all wrong.

That 25% gap seems troublesome.  Because, like I say, I need to use it 
to tune the params, but there's other factors at work here, which was 
kind of expected.  The concern is say I could sim  it and find trading 
off accel for velocity yields a 10% runtime improvement in sim, but in 
reality that's just a idiosyncrasy of stepping 100x faster and in 
reality maybe this change makes it 5% slower, not faster.  In that case 
it's foolish to try to tune with it.  Chasing unicorns.  You'd never 
know if you actually made it better or worse without a 1x run, making 
the sim useless.

Did you loopback the position command, or just disable the motors?  I 
wonder if the DIR_SETUP and DIR_HOLD are mucking with the accelerated 
timeframe.

Hey, wait- try changing the SERVO_PERIOD given to loadrt emcmot too.  
That should change *everything*.

So I heard the part about looping back the position command. Could we 
leave the vel and accel normal, let the motion component believe that 
it's running at its normal rate, but then run the motion component at a 
much faster rate than real time?  Hack the time base?

The motion component's EMCMOT, right?  My understanding is quite crude, 
but is it like a 1ms thread that just increments its own time counter by 
1ms every time it gets called?  If so then the answer would be to leave 
it believing it's on a 1ms time base but reschedule the thread execution 
for 0.01ms, or 0.001ms, or any number.  It will reliably sim could how 
many ms the program takes regardless of how often the thread executes.


I do see the loadrt loads emcmot thread at SERVO_PERIOD.  Of course 
there's no hook to change the concept of time inside emcmot from the 
scheduler, because it's not normally done.


I don't see how you could change it on the fly either.  I mean there'd 
be a VCP button for "switch to sim run" that could easily loopback the 
position command, but it would also have to tell the scheduler that 
emcmot thread needs to run at a much different rate and I don't know how 
you'd do that from VCP.

Danny


On 10/14/2016 1:32 PM, Todd Zuercher wrote:
> I seem to be having some success with this.  I have a simulation config set 
> up with max velocities set to 100x, and max accelerations set to 1x, 
> feed-overide set to 100x.  Using my scaled timer set to 100x the estimated 
> run time seems to be pretty consistently about 25% too long for a 20 minute 
> engraving file, but a slow 1 hour long single move tests just right.
>
> So I increased the acceleration still further, until I started to get 
> runtimes approximately the same as an unscaled simulation.  I ended up with 
> the acclerations set to 2x the original configuration that I am trying to 
> simulate.  It doesn't hardly seem right but is what seems to be working.
>
> - Original Message -
> From: "Todd Zuercher" 
> To: "Enhanced Machine Controller (EMC)" 
> Sent: Friday, October 14, 2016 9:52:18 AM
> Subject: Re: [Emc-users] Hacking a sim runtime mode
>
> A quick stupid question, if i'm scaling time 100x, velocities should increase 
> 100x, but what about acceleration, should that be increased 1x or 100x?
>
> - Original Message -
> From: "Todd Zuercher" 
> To: "Enhanced Machine Controller (EMC)" 
> Sent: Thursday, October 13, 2016 9:30:34 PM
> Subject: Re: [Emc-users] Hacking a sim runtime mode
>
> I think you are probably right.  Running the simulation at 100x was much more 
> accurate than 1000x.  Tomorrow I'll have to play with the servo thread speed 
> some and see if and how much improvement I can find.  But not sure how much 
> improvement will be possible in a sim config on a VM.
>
> - Original Message -
> From: "Stephen Dubovsky" 
> To: "Enhanced Machine Controller (EMC)" 
> Sent: Thursday, October 13, 2016 3:52:42 PM
> Subject: Re: [Emc-users] Hacking a sim runtime mode
>
> I think the error of this approach (using N times higher accel/vel) is
> coming from the piecewise linear approximations of the curves going bad.
> If you try to follow a curve N times faster, then there are N times fewer
> samples along that curve for a fixed time sampling period.  That introduces
> errors.  There may be others but thats at least part of it.  Its the same
> problem w/ discrete time vs continuous time systems.  There is divergence
> between the solutions as the sampling rate goes down.
>
>
> On Thu, Oct 13, 2016 at 3:08 PM, Todd Zuercher 
> wrote:
>
>> I was just working at this a little bit.   The first thing I did was to
>> modify time.comp, to make a new sim_time.comp.  The sim_time.comp 

Re: [Emc-users] Probe cycle switches to MDI

[Danny Miller]

On 10/8/2016 12:13 PM, Gene Heskett wrote:
> On Saturday 08 October 2016 12:25:57 Danny Miller wrote:
>
>> I have the wireless pendant button triggering a probe cycle in an
>> .ngc.
>>
>> Problem is, IF the cycle ends without tripping the probe (user error),
>> Axis UI ends up switching to MDI tab, and all the pendant function is
>> for Manual Control tab.  So it disables the pendant until you can get
>> back to the PC and fix it by clicking back to Manual Control tab.  Why
>> does it end up in MDI tab, and how can make LinuxCNC end up back in
>> the Manual Control tab?
>>
>> .ngc code below:
>>
>> 
>> o100 sub
>>
>> #1 = #[5203 + [20 * #5220] ]
>>
>> G38.3 F50 Z[-6.5-#1]
> Coding style. did you mean -6.5 - #1? I always use a space around a math
> function char, reducing the ambiguity when I go back and read it
> months/years later.
>
>> G10 L2 P0 Z1.826
> This also is a coding style preference, but I have gotten out of the
> habit of adjusting the G54 map.  It reduces the hunting while looking
> for a map that matches the one originally homed to mapping, for me at
> least.  So I wind up using the next 2 or 3 maps in most of my coding.
>
> This also is a disadvantage because, realizing the machine is going to
> hit a fixture or something I didn't intend, hitting the esc key restores
> G54, and unless you know what map was in use, you lost the data that
> would tell me my math was funkity or THAT map was funkity because my
> math was.
Yeah I wanted to refer to a .ini parameter.  Wasn't sure how, on my 
to-do list.  So I can refer to .ini parameters inside g-code? 
Z#<_probing.probethickness>  or what?
Point is, probe tool is 1.826 inches thick, so once it touches off it 
needs to fix the z-offset for that point to be 1.826 so z=0 is the 
bottom of the probe, i.e. the surface it's sitting on.
>
>> o101 if [#5070 EQ 1]
>>   G10 L20 P0 Z1.826
>>   G1 Z1.93 F3
>>   G38.3 F3 Z[1.6] (<- useless brackets. what effect? if any)
>>   G10 L20 P0 Z1.826
>>   G0 Z2.5
>> o101  else
>>   G53 G0 Z0
>>   (debug, FAILED TO FIND PROBE)
>> o101 endif
>>
>> o100 endsub
>> 
>>
>>
>> Thanks,
>>
>> Danny
>>
> I don't see anything above that would put you in the mdi mode unless the
> math equates to an F5 character somehow. Not seeing the rest of the
> code, I've no clue if the scope of #1 thru #30 has been violated.
It's if the probe cycle fails to trip.  If the probe gets found in the 
acceptable travel, all is well.
If the probe travel limit is reached, the error condition dumps out of 
the .ngc code and for some odd reason the Axis UI ends up with the MDI 
tab selected.  Well it's not incomprehensible, Axis tried to run g-code 
and got an exception condition.  But it's a problem, not a show-stopper 
but I did aim to make the pendant control essentially everything but 
file selection with minimal need to use the terminal.
>
> This to me is a very strong argument in favor of separating global data
> from local data by the use of #<_name> (note the underscore) for global
> variables, and # for local to this subroutine data.  In my own
> coding I find that results in far fewer surprises.  Either way works if
> you pay attention but I find that the leading underscore is a reminder
> that this is global data, and its lack as an indicator that this was
> intended to be local, vanishing with the end of the subroutine rather
> handy when I am reading it again later.
>
> Named data to me is at least as handy as sliced bread. :)
>
> I've no clue if this helps but I hope it does Danny.
>
> Cheers, Gene Heskett


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Probe cycle switches to MDI

[Danny Miller]

I have the wireless pendant button triggering a probe cycle in an .ngc.

Problem is, IF the cycle ends without tripping the probe (user error), 
Axis UI ends up switching to MDI tab, and all the pendant function is 
for Manual Control tab.  So it disables the pendant until you can get 
back to the PC and fix it by clicking back to Manual Control tab.  Why 
does it end up in MDI tab, and how can make LinuxCNC end up back in the 
Manual Control tab?

.ngc code below:


o100 sub

#1 = #[5203 + [20 * #5220] ]

G38.3 F50 Z[-6.5-#1]
G10 L2 P0 Z1.826

o101 if [#5070 EQ 1]
 G10 L20 P0 Z1.826
 G1 Z1.93 F3
 G38.3 F3 Z[1.6]
 G10 L20 P0 Z1.826
 G0 Z2.5
o101  else
 G53 G0 Z0
 (debug, FAILED TO FIND PROBE)
o101 endif

o100 endsub



Thanks,

Danny


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Can speed/accel be changed without rebooting?

[Danny Miller]

I was going through the HAL and noticed the per-axis ini parameters are 
there.

Can they be changed without rebooting?  The machine will be at a stop 
for this.

Reason being, I use the standard cutting profile but I also have a 3D 
carving profile which increases the acceleration much higher.  To stay 
within the motor's performance curve, the max speed of the axes is 
limited for 3D carving.  And the carving profile's acceleration shakes 
the table in a way you just don't need for regular cuts.

So I was thinking, can I just make a PYVCP button that changes to the 
alternate profile, faster accel but slower linear, and back?

I've done 3D carving and then a toolchange to switch to an endmill for 
cutout.  No real biggie, but I couldn't switch to the 2D cut profile 
without rebooting, so I didn't reboot.  It'd be an interesting feature 
if it was just a button.

Danny







--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Getting current RPM setting?

[Danny Miller]

On 10/2/2016 11:35 AM, Sebastian Kuzminsky wrote:
> On 10/02/2016 10:12 AM, Danny Miller wrote:
>> On 10/2/2016 11:05 AM, Sebastian Kuzminsky wrote:
>>> On 10/02/2016 02:32 AM, Danny Miller wrote:
>>>> I thought it would be a good idea to display what the RPM setting is
>>>> before the spindle is started.
>>>>
>>>> I see current_rps is right there WHILE the spindle is running.
>>>>
>>>> But when I enter "S1", it's set inside LinuxCNC even before M3.  I'd
>>>> like to display it in pyvcp before M3 uses it. current_rps remains at zero.
>>>>
>>>> Is there a way to access the commanded RPM when not actually sending
>>>> it?  I tried but couldn't find it.
>>> It's available on the pin 'motion.spindle-speed-out'.  It's 0 when the
>>> spindle is stopped (M2 or M5), and it's the commanded S-word when the
>>> spindle is running (M3).
>>>
>>> Details here:  http://linuxcnc.org/docs/html/man/man9/motion.9.html
>>>
>>>
>> Yeah I already have that.   It's useful, but I'd like to be able to
>> display what the spindle was set for when NOT running.
>>
>> Specifically I want the commanded S-word.
>>
>> It does show up under MDI "Active G-codes", but I don't know how to
>> access it in HAL.
> I see, i misread your original email.  I think unfortunately the S-word
> is not available in HAL currently, just in the "active g-codes" where
> you found it.
>
> It would probably be easy to add a motion pin with that information, i'd
> be happy to guide you through that if you like.
>
>
Sounds like a lot of complexity, but it's of great value to know how 
that works.  Shoot!

Danny


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Getting current RPM setting?

[Danny Miller]

On 10/2/2016 11:05 AM, Sebastian Kuzminsky wrote:
> On 10/02/2016 02:32 AM, Danny Miller wrote:
>> I thought it would be a good idea to display what the RPM setting is
>> before the spindle is started.
>>
>> I see current_rps is right there WHILE the spindle is running.
>>
>> But when I enter "S1", it's set inside LinuxCNC even before M3.  I'd
>> like to display it in pyvcp before M3 uses it. current_rps remains at zero.
>>
>> Is there a way to access the commanded RPM when not actually sending
>> it?  I tried but couldn't find it.
> It's available on the pin 'motion.spindle-speed-out'.  It's 0 when the
> spindle is stopped (M2 or M5), and it's the commanded S-word when the
> spindle is running (M3).
>
> Details here:  http://linuxcnc.org/docs/html/man/man9/motion.9.html
>
>

Yeah I already have that.   It's useful, but I'd like to be able to 
display what the spindle was set for when NOT running.

Specifically I want the commanded S-word.

It does show up under MDI "Active G-codes", but I don't know how to 
access it in HAL.

Danny


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Getting current RPM setting?

[Danny Miller]

I thought it would be a good idea to display what the RPM setting is 
before the spindle is started.

I see current_rps is right there WHILE the spindle is running.

But when I enter "S1", it's set inside LinuxCNC even before M3.  I'd 
like to display it in pyvcp before M3 uses it. current_rps remains at zero.

Is there a way to access the commanded RPM when not actually sending 
it?  I tried but couldn't find it.

Thanks,

Danny


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Estimated time remaining

[Danny Miller]

The estimator DOES just use distance & feedrate, not acceleration.  This 
is effective for estimating 2D cuts but junk for 3D carving, which 
hinges primarily on acceleration.

There is no one effective "factor".  I have 3D carvings which took 4x 
longer than estimated, others 6x.  If it is only a gentle slope then the 
factor could be 1x, all for the same profile.

When tuning a machine, you can reduce the axis max velocity in order to 
increase the max acceleration, and this can in fact make the machine run 
faster.  Note an arbitrary multiplier factor would be useless for 
tuning, as you need to know the actual effect of the parameter changes.

But the feature I proposed- an ETA timer- I don't see how to do it, 
because the required info isn't on the interface that I can find. 
Specifically "total line count from the G-code" and "current line number 
in the G-code".  The concept would fail for sure on G-code bearing 
subroutines but that's not a feature of any of the 3d carving I'm doing.

The ETA-by-line-count would be inaccurate on 2D cutting, even without 
subroutines, because there's no telling how long a vector is.  A curved 
cut can be 10 or 1000 lines but only take a second or two, while a 
single straight line can be a long cutting time.

Danny

On 10/1/2016 8:39 AM, Todd Zuercher wrote:
> I suppose you might be able to make something like that.
>
> But I think I have a good idea for improving the run time estimator.  Right 
> now how does it work, does it just use the feed rates X distance to be 
> traveled?  What if it took that and added to it a factor(derived from the 
> machines acceleration rate) X the number of lines in the file.
>
> - Original Message -
> From: dan...@austin.rr.com
> To: "Enhanced Machine Controller (EMC)" 
> Sent: Friday, September 30, 2016 11:30:51 PM
> Subject: [Emc-users] Estimated time remaining
>
> I am doing 3D carving, where the Properties analysis is regrettably useless 
> for coming up with time estimates due to not taking into account the 
> acceleration aspect of trajectory planning.
>
> I did install the cycle timer pvcp and it does certainly help.
>
> But one thing I noticed- these carvings are "mostly" consistent in how much 
> time they're taking per-line.  It would be accurate enough to be helpful to 
> calculate:
>
> Time Remaining=(total # of gcode lines/gcode lines done so far)*cycletime so 
> far
>
> Is there any way to do that?  All I can see is access to "time".
>
> Danny
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Anybody have an idea how...

[Danny Miller]

I did have a theory it was potmetal.  Wasn't as bright and metallic as 
aluminum.

Danny


On 9/25/2016 1:58 PM, Jon Elson wrote:
> On 09/25/2016 01:11 PM, Danny Miller wrote:
>> Some of our folks built a metal kiln and we tried to melt some aluminum
>> swarf the neighboring metalshop discarded, at least what we thought was
>> aluminum (no confirmation, but it wasn't steel). We got almost no metal,
>> just inches of dross.  No inert gas or flux there, in fact the metal was
>> exposed to combustion gases.
>>
>>
> It may have been "pot metal", which is what carburetors and
> such parts used to be made of.  Mostly zinc.
>
> Some guys here have various melting furnaces, and general
> chunks of aluminum melt well.  You do get some dross, but
> not INCHES of it, for sure.  One guy has an old ladle
> furnace, which looks like a stainless wok with 3 gas burners
> below it.  So, you light the burners, and then toss various
> chunks of aluminum castings (pistons, sawed-up transmission
> casings, etc.) in the bowl, and they slowly disappear.  When
> all melted, you skim the dross and ladle out the aluminum
> into your molds.  Even with the pool of molten aluminum
> exposed to the air, you don't get much dross.
>
> Jon
>
> --
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


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

Re: [Emc-users] Anybody have an idea how...

[Danny Miller]

Some of our folks built a metal kiln and we tried to melt some aluminum 
swarf the neighboring metalshop discarded, at least what we thought was 
aluminum (no confirmation, but it wasn't steel). We got almost no metal, 
just inches of dross.  No inert gas or flux there, in fact the metal was 
exposed to combustion gases.

Danny


On 9/25/2016 1:05 PM, Stuart Stevenson wrote:
> Industrial recyclers also discount aluminum swarf unless it is in compacted
> puck form. Much of the uncompacted swarf oxidizes (disappears) when melted.
> The pucks melt the same as solids.
>
> On Sep 25, 2016 12:50 PM, "Todd Zuercher"  wrote:
>
> A number of recycles don't want anything to do with the stuff and only take
> it at a deeply discounted rate.
>
> - Original Message -
> From: "Gene Heskett" 
> To: emc-users@lists.sourceforge.net
> Sent: Sunday, September 25, 2016 1:29:29 PM
> Subject: Re: [Emc-users] Anybody have an idea how...
>
> On Sunday 25 September 2016 12:39:38 Jon Elson wrote:
>
>> On 09/24/2016 01:58 PM, Gene Heskett wrote:
>>> 43/25.4 to get inches = 1.6929etc * pi *rpms, so its 3.63"
>>> per rev, and 800 revs is then 2900 SFM. So I should be
>>> turning at least 1200-1300 to dup that. Right?
>> Well, 800 RPM is not insanely slow for that diameter.  I
>> didn't realize this was lathe work.  But, you definitely can
>> spin faster, and nothing wrong with 4000 SFPM.
>>
>>> The question then is what the heck do I do with all that
>>> curly swarf? Making the pulley hub, and boring the pulley,
>>> (its all together and looks like it will work, well
>>> aligned now) made at least a bushel of gnarly swarf I
>>> carried to the bin in sometimes soccer ball gobs. And the
>>> lathe is still a mess. I'll have to sweep before its clean
>>> enough not to clog up the vacuum.
>> When I make a lot of aluminum chips, I can turn them in for
>> recycling.
> I tried that but these local people on the monthly recycle pickup seem to
> equate pop & beer cans with alu and didn't take a bag of alu swarf.
>
>> I used to run them through an antique meat
>> grinder, but that was too much work.  If the chips won't
>> mash down, you can snip them a little with tin snips to make
>> them compact better.  A proper chip breaker on the cutter
>> may  make them crumble into shorter pieces.
> I don't think 6061 is hard enough to break.  Most of my inserts have a
> nice little chip breaker bump, curls them up fairly tight.  Makes sand
> out of cast iron and 7075-T6 or better though.
>
>>> Interesting swarf story to tell this morning. The wire
>>> joints for the x motor, located where the wire comes out
>>> of the cable/chain, so its basically on top of the motor,
>>> and the cable is connected to the motor wires with little
>>> grey wire nuts. A lot of the swarf went off the back of
>>> the toolbox, crossing the area where those nuts are
>>> sticking up. Your are guessing by now where this is going.
>>> At one point it must have shorted, and about 4" of that
>>> swarf lit up white hot and exploded like a fuse, nice
>>> attention getting pop. The 2M542 driver protected itself
>>> well, the motor never missed a step! I either need to put
>>> that in a small box, or at least fill the back of the nuts
>>> with Go-2. Cheers, Gene Heskett
>> Well, I had swarf get in the motor once or twice on my old
>> lathe, but very lucky that didn't blow up your driver.
>>
> They are advertised as being self protective. I don't know if once proves
> the point.  It did make quite a flash though. I had a thumb poised over
> the esc key just in case, but when it came time to move it again, it
> moved as per told.  And I took a breath. :)
>
>> Jon
>>
>> --
>>  ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>
> Cheers, Gene Heskett
> --
> "There are four boxes to be used in defense of liberty:
>   soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author)
> Genes Web page 
>
> 
> --
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
> 
> --
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
> --
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>

Re: [Emc-users] Trouble with TWOPASS and Mesa loadrt

[Danny Miller]

On 9/25/2016 10:05 AM, andy pugh wrote:

> On 25 September 2016 at 08:36,  wrote:
>
>> I am trying to add a run timer as per the time man page.
>>
>> LinuxCNC fails to load.  Initially I think the offending line was "not:
>> already exists" in the postgui hal.  Yes it uses loadrt not, and if I
>> comment all the "not" stuff out, linuxCNC loads fine.  loadrt is used in
>> another hal.
>>
> To use TWOPASS the comonents loaded by more than one "loadrt" need to use
> the names= syntax.
> You can, if you want, use names=not.0,not.1,not.2 so that there is no need
> to change the HAL lines that use the not components.

Huh, I tried that last night and it didn't work, but I may have screwed 
something up.
What I saw, though, was the duplication of "loadrt not" was causing the 
problem, and TWOPASS broke not because of the "not" naming scheme but 
the HOSTMOT2((DRIVER) syntax error.  Things might not be as they seem, 
though.  The debug dump is long and confusing.
>
> twopass: Error in file ./7i92_AM882_hack.hal:
>>  can't read "::HOSTMOT2((DRIVER)": no such element in array
>> twopass: Error in file ./xhc-hb04.hal:
>>  can't read "::HOSTMOT2((BOARD)": no such element in array
>>
> There is absolutely no reason to have a "loadrt hostmot2"  in more than one
> file. It won't work.
I didn't load it twice, though!

7i92_AM882_hack.hal has the loadrt line
xhc-hb04.hal does NOT have the loadrt line.  However, it references hostmot2 
pins.  All that worked fine.
The xhc-hb04.hal and 7i92_AM882_hack.hal loaded fine and linuxCNC works great 
before including TWOPASS.

My theory is that TWOPASS is breaking on the unusual syntax with these 
brackets.  Note the errors "no such element in array" came up in both 
7i92_AM882_hack.hal which DOES have the appropriate loadrt line, and 
xhc-hb04.hal which does NOT have its own loadrt line.  i.e. after the first 
pass it made its own malformed loadrt line which broke because of the brackets 
and now neither file can find [HOSTMOT2](DRIVER) because there's no functional 
loadrt line.  That's my initial theory, anyhow.

Does the nodelete option leave around a temporary file which would help 
diagnose this?  Where does that file end up?


>
> I don't know if the brackets have any effect. I see them in the sample
> configs, but only on the hostmot2 items.
>


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

Re: [Emc-users] "Probe triggered during jog"

[Danny Miller]

Well, how do I modify it?  Like I say, pressing Probe Z momentarily 
starts an ngc.

My best guess is the probe input needs to AND with "isAuto", does that 
sound right?  I don't think I saw the probe trip during g-code running 
(which would be weird if it's to protect the probe).

Danny


On 9/7/2016 7:52 AM, John Kasunich wrote:
>
> On Wed, Sep 7, 2016, at 02:43 AM, Danny Miller wrote:
>> I have a wireless toolsetting probe here.  It's nice, but it's naturally
>> sensitive.  Even when you're not using it, very easy to trip, and
>> LinuxCNC generates a "Probe triggering during jog" error.  Just bump the
>> table.  The probe doesn't even have a physical hard switch, just the plate.
>>
>> Hmm.  I don't know why it  was coded to do this, but it's a problem.
>> Basically it should be ignoring Probe-in if it's not probing.
> You wouldn't think that if you ever ran the business end of a $500
> probe into the side of a vise while jogging
>
> The normal behavior is to protect sensitive and expensive equipment.
> If you don't like that, you can modify it using HAL logic quite easily.
>


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

[Emc-users] "Probe triggered during jog"

[Danny Miller]

I have a wireless toolsetting probe here.  It's nice, but it's naturally 
sensitive.  Even when you're not using it, very easy to trip, and 
LinuxCNC generates a "Probe triggering during jog" error.  Just bump the 
table.  The probe doesn't even have a physical hard switch, just the plate.

Hmm.  I don't know why it  was coded to do this, but it's a problem. 
Basically it should be ignoring Probe-in if it's not probing.

How can I stop this?  I should probably AND the signal with one that 
says it's actually in a probing cycle.  Thus if it's not running the 
probe ngc, the AND output isn't true and LinuxCNC won't see a probe trip 
even if you genuinely trigger the probe.

So the probing cycle button (momentary) triggers an MDI command which 
starts an NGC.  As best I understand it, AND requires a HAL net.

And HAL nets can't be assigned in g-code, can they?

Can an MDI line assign a HAL net?

Can an MDI line do things in addition to calling the ngc, or is it 
single-op?  What would the syntax be?

The ProbeStart button is  only a momentary.  I believe it can set a HAL 
net and trigger the MDI on the same line, but is there any way to signal 
back to the HAL at the end to set this net to disable the Probe-in signal?

This seems similar to the question I had before about how you might set 
up ProbeStart to operate as a deadman button, but I haven't yet tried to 
go through this.

Danny


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

Re: [Emc-users] Need unlocked copy of a pdf for a breakout board I just bought two of.

[Danny Miller]

On 9/7/2016 12:53 AM, Gene Heskett wrote:
> On Wednesday 07 September 2016 01:41:24 Dave Caroline wrote:
>
>> That looks like a board I would not recommend look close to the white
>> sockets near the axis outputs, it has series resistors for the driver
>> optos, these of course are in the driver as well, this means half
>> current through the opto and failure to work.
>>
>> I was working with a free issue board where I met that problem before
> So have I, Dave, and I wound up bypassing that particular opto in the
> board I used to drive my G0704, but this one doesn't have the opto's
> that I know of.  Because they are slow, I've found them far more trouble
> than they are worth as a safety measure.  I'd much druther trust a good
> grounding system, star topology of course.
>
> Thanks Dave.

I similarly dislike optos.  I have doubts that any properly made drive 
can even produce dangerous "spikes".   They seem pointless.

I work with a 7i92 card which is ethernet, and thus already offers 
galvanic isolation through the ethernet.  If there were a huge surge, it 
wouldn't propagate back to the PC.  The PC ain't a high-dollar item anyways.
The AM882 drives all have differential opto inputs/outputs themselves, 
there's no case for even hypothetical "spikes".

Danny
>> Dave Caroline
>>
>> --
>>  ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>
> Cheers, Gene Heskett


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

Re: [Emc-users] Need unlocked copy of a pdf for a breakout board I just bought two of.

[Danny Miller]

On 9/6/2016 6:03 PM, Gene Heskett wrote:
> On Tuesday 06 September 2016 18:21:14 dan...@austin.rr.com wrote:
>
> Back on the list Danny.
>
>> Mesaflash allows you to change the IO configuration, but only to a
>> number of premade configs, not arbitrary.  The GPIO has few limits but
>> the location of special-purpose STEPx/DIRx pairs are fixed. But those
>> locations change based on the mesaflash file you use.
>>
>> Danny
> I understand that, but without the docs on these boards, I'd be playing a 
> game of 25! to find the one correct config.
>
> We'll see what response I get from the vendor about the password in the next 
> 24 hours.
>
> In the meantime I have figured out how to put a pretty good swarf shield over 
> the X screw, and may duplicate it on the bottom of the saddle so even the air 
> hose will have a hard time moving junk into the screw.  And since this 
> barrier is on the bottom of the cross-slider, but leaves a catch trench in 
> the top, so I'll duplicate it on top so there is a smooth top on the 
> cross-feed. I'll have to machine a shallow trench on the top so the top 
> shield sits flush.

We can't open that ebay order URL, BTW.
I made a 7i92 breakout board that holds the card, provides 5v and 24v 
power, set up all my AM882H drives on 10-pin ribbon IDCs  for greatly 
simplified wiring, optos in the problematic inductive prox sensors, and 
made a DB25 connection to be a harness for all the accessory wiring 
(homing, estop, etc).  Screwed a few things, had to do some rework, but 
totally worth it.

Danny

>>  Gene Heskett  wrote:
>>> Greetings all;
>>>
>>> Those breakout boards I bought 2 of arrived today, and are
>>> apparently pin assigned according to what a mach3 system might use,
>>> including the male db25 on the pcb.  It apparently has pre-assigned
>>> output pin groups that are about as incompatible with a 5i25 as
>>> could be. I never did find where the X axis for my lathe comes out,
>>> but the Z is apparently in the B group of 4.  I ran the spindle,
>>> expecting to see the analog of the pwm someplace, but if its there,
>>> it is routed to the db15 on the other end of the card.  So I go to
>>> look at the pdf of the board thats on the mini cd-r, and found it,
>>> and a .doc file, both of which are password protected. Its not that
>>> much money, and it would cost thru the nose to put a stop payment on
>>> it, but I am tempted anyway.  What a maroon!
>>>
>>> The board is this one.
>>> >> d07f20cc1c0c4c0dc182854920714f1f1b6b4303135b734363275f603a92bce078a86
>>> 06039949992940667a6a5eaaa1a5b149716a727e1150b418a4ab24233537bf34af243
>>> 133af38bb12285aa66465540b00>
>>>
>>> Anybody recognize it, or has used it?  Better yet, knows how to
>>> unlock the pdf?  Nothing I have here, including libreoffice, can
>>> touch the pdf or the .doc files for the card.  All sorts of adv
>>> stuff for his driver boxes, you can look at them, but nothing on the
>>> card itself.
>>>
>>> I'd love to be able to buy about a 6 pack of cnc4pc's old C1G BoB's,
>>> as it has an led for state on every pin, a great troubleshooting
>>> tool.  But its been disco'd. Even at 80 plus bucks it was a good
>>> card.
>>>
>>> Thanks all, for any help on this.
>>>
>>> Cheers, Gene Heskett
>>> --
>>> "There are four boxes to be used in defense of liberty:
>>>   soap, ballot, jury, and ammo. Please use in that order."
>>> -Ed Howdershelt (Author)
>>> Genes Web page 
>>>
>>> 
>>> -- ___
>>> Emc-users mailing list
>>> Emc-users@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/emc-users
>
> Cheers, Gene Heskett


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

Re: [Emc-users] Can't figure out homing on gantry

[Danny Miller]

I'm all for JA but we don't have time to install and experiment with the 
pre-release right now.


Danny


On 9/5/2016 6:10 PM, John Thornton wrote:
> Oh I forgot to say I gave up on the gantry component once the JA branch
> could home a gantry correctly and now it is in master.
>
> JT
>
> On 9/5/2016 4:14 PM, Danny Miller wrote:
>> On 9/5/2016 9:24 AM, John Thornton wrote:
>>> Sounds like your home switches for the gantry are not connected
>>> correctly in hal. You do have one switch for each joint right?
>>>
>>> JT
>> Yep my HAL's right there.  And I've watched switches-x1 and switches-x2
>> on the HAL monitor while I triggered them.  They respond.
>>
>> At one point I switched the connection for gantry.0.joint.00.home and
>> gantry.0.joint.01.home, just in case I was using the switch for X1 on
>> what was actually the drive for X2.  It was a good theory, but it didn't
>> change anything.
>>
>> Danny
>>
>>>
>>> On 9/5/2016 7:34 AM, Danny Miller wrote:
>>>> On 9/2/2016 9:45 AM, Charles Steinkuehler wrote:
>>>>> On 9/1/2016 9:28 PM, dan...@austin.rr.com wrote:
>>>>>> Well, wait- just rechecked the gantry man page: "When the system is
>>>>>> homing and a joint home switch activates, the command value sent to
>>>>>> that joint is "frozen" and the joint offset value is updated
>>>>>> instead"
>>>>>>
>>>>>> It unambiguously DOES say it's per-axis homing, but I saw it stop
>>>>>> both when X1's limit tripped and X2 never went into seek, and if X2
>>>>>> was in front of X1, went over the homing switch with no effect
>>>>>> until X1 tripped.
>>>>>>
>>>>>> Here's what's in my HAL that should be relevant, did I screw
>>>>>> something up?
>>>>>>
>>>>>> loadrt gantry count=1 personality=2
>>>>>> net switches-x1   <=  hm2_[HOSTMOT2](BOARD).0.gpio.005.in_not
>>>>>> net switches-x2   <=  hm2_[HOSTMOT2](BOARD).0.gpio.003.in_not
>>>>>> net switches-x1   => gantry.0.joint.00.home
>>>>>> net switches-x2   => gantry.0.joint.01.home
>>>>>> net home-x <= gantry.0.home
>>>>>> net home-x  => axis.0.home-sw-in
>>>>> That looks OK, but it's not enough to verify your HAL file is correct.
>>>>>
>>>>> The behavior you describe could happen if the search-vel input is
>>>>> incorrect, if you're using the limit output instead of the home output
>>>>> to feed to motion, or if your home switch signals have the wrong
>>>>> polarity (they should be high if the switch is "tripped").
>>>> There is ONLY a homing switch.  No limit.
>>>>
>>>> Polarity is correct.
>>>>
>>>> I didn't have an [Axis3] section at all in my ini, since nothing seemed
>>>> to use it, all references for axis3 functions go back to axis0.  I did
>>>> copy Axis0 section into an Axis3 though, so search-vel does exist now
>>>> for axis3.
>>>>
>>>> Didn't help.
>>>>
>>>> I did see this: when it homes, if X1 is ahead of X2, it stops both on
>>>> X1.  However if X2 is ahead of X1, it ignores BOTH homing switches.  It
>>>> just drives both  into the stops and grinds forever. The sensors are far
>>>> enough from the stops now that there's room to go far enough to pass the
>>>> switches which does create an awkward case where it may start already
>>>> past it and  grind on the endstop without ever seeing the sensor.
>>>>
>>>> Danny
>>>>
>>>> --
>>>> ___
>>>> Emc-users mailing list
>>>> Emc-users@lists.sourceforge.net
>>>> https://lists.sourceforge.net/lists/listinfo/emc-users
>>> --
>>> ___
>>> Emc-users mailing list
>>> Emc-users@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/emc-users
>>>
>> --
>> ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>
> --
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


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

[Emc-users] Spurious edge on startup

[Danny Miller]

I have a HAL which is detecting a false edge when loading LinuxCNC.

I have AM882 drives which have an Alarm signal output, and an Enable 
signal input.  Alarm is active-low, Enable is active by default unless 
LinuxCNC deasserts it, all goes through a 7i92 card.  The Alarm signal 
is tied to its automatic stall detection. If tripped, it disables the 
STEP input, and only cycling Enable pin resets the drive.  The Enable 
pin also disables the motor current output entirely, which is 
undesirable, the Z will drop. Initially I had Enable go inactive during 
estop but it caused the Z to fall for no good reason.  Arguably less 
safe.  You have an emergency, bit is spinning, you hit e-stop, the Z 
falls with the motor still going.

I just simplified it in HAL as "if the estop-assert AND Alarm-assert 
have a posedge, trigger a one-shot to deassert the Enable signal for 
0.25 sec".

It mostly does what I expect!  However, when I boot up LinuxCNC, Enable 
deasserts for 0.25 sec and the Z falls a bit, I definitely don't want it 
to do that without Alarm being asserted.  At the time, the drivers were 
already physically powered and Alarm is NOT asserted (low) for sure.

I presume the first servo-thread cycle has a glitchy value on startup, 
probably the 7i92 itself.  I'm not sure what actually does that, the 
one-shot has to have a posedge to start the one-shot, not a level.  If 
the HAL saw Alarm pin as low (asserted) during booting and the first 
cycle of servo-thread saw its correct value (high, deasserted), that 
should NOT trigger the one-shot. Wrong edge.

Any idea?

Danny



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

Re: [Emc-users] Can't figure out homing on gantry

[Danny Miller]

On 9/5/2016 9:24 AM, John Thornton wrote:
> Sounds like your home switches for the gantry are not connected
> correctly in hal. You do have one switch for each joint right?
>
> JT
Yep my HAL's right there.  And I've watched switches-x1 and switches-x2 
on the HAL monitor while I triggered them.  They respond.

At one point I switched the connection for gantry.0.joint.00.home and 
gantry.0.joint.01.home, just in case I was using the switch for X1 on 
what was actually the drive for X2.  It was a good theory, but it didn't 
change anything.

Danny

>
>
>
> On 9/5/2016 7:34 AM, Danny Miller wrote:
>> On 9/2/2016 9:45 AM, Charles Steinkuehler wrote:
>>> On 9/1/2016 9:28 PM, dan...@austin.rr.com wrote:
>>>> Well, wait- just rechecked the gantry man page: "When the system is
>>>> homing and a joint home switch activates, the command value sent to
>>>> that joint is "frozen" and the joint offset value is updated
>>>> instead"
>>>>
>>>> It unambiguously DOES say it's per-axis homing, but I saw it stop
>>>> both when X1's limit tripped and X2 never went into seek, and if X2
>>>> was in front of X1, went over the homing switch with no effect
>>>> until X1 tripped.
>>>>
>>>> Here's what's in my HAL that should be relevant, did I screw
>>>> something up?
>>>>
>>>> loadrt gantry count=1 personality=2
>>>> net switches-x1   <=  hm2_[HOSTMOT2](BOARD).0.gpio.005.in_not
>>>> net switches-x2   <=  hm2_[HOSTMOT2](BOARD).0.gpio.003.in_not
>>>> net switches-x1   => gantry.0.joint.00.home
>>>> net switches-x2   => gantry.0.joint.01.home
>>>> net home-x <= gantry.0.home
>>>> net home-x  => axis.0.home-sw-in
>>> That looks OK, but it's not enough to verify your HAL file is correct.
>>>
>>> The behavior you describe could happen if the search-vel input is
>>> incorrect, if you're using the limit output instead of the home output
>>> to feed to motion, or if your home switch signals have the wrong
>>> polarity (they should be high if the switch is "tripped").
>> There is ONLY a homing switch.  No limit.
>>
>> Polarity is correct.
>>
>> I didn't have an [Axis3] section at all in my ini, since nothing seemed
>> to use it, all references for axis3 functions go back to axis0.  I did
>> copy Axis0 section into an Axis3 though, so search-vel does exist now
>> for axis3.
>>
>> Didn't help.
>>
>> I did see this: when it homes, if X1 is ahead of X2, it stops both on
>> X1.  However if X2 is ahead of X1, it ignores BOTH homing switches.  It
>> just drives both  into the stops and grinds forever. The sensors are far
>> enough from the stops now that there's room to go far enough to pass the
>> switches which does create an awkward case where it may start already
>> past it and  grind on the endstop without ever seeing the sensor.
>>
>> Danny
>>
>> --
>> ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>
> --
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


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

Re: [Emc-users] Can't figure out homing on gantry

[Danny Miller]

On 9/2/2016 9:45 AM, Charles Steinkuehler wrote:
> On 9/1/2016 9:28 PM, dan...@austin.rr.com wrote:
>> Well, wait- just rechecked the gantry man page: "When the system is
>> homing and a joint home switch activates, the command value sent to
>> that joint is "frozen" and the joint offset value is updated
>> instead"
>>
>> It unambiguously DOES say it's per-axis homing, but I saw it stop
>> both when X1's limit tripped and X2 never went into seek, and if X2
>> was in front of X1, went over the homing switch with no effect
>> until X1 tripped.
>>
>> Here's what's in my HAL that should be relevant, did I screw
>> something up?
>>
>> loadrt gantry count=1 personality=2
>> net switches-x1   <=  hm2_[HOSTMOT2](BOARD).0.gpio.005.in_not
>> net switches-x2   <=  hm2_[HOSTMOT2](BOARD).0.gpio.003.in_not
>> net switches-x1   => gantry.0.joint.00.home
>> net switches-x2   => gantry.0.joint.01.home
>> net home-x <= gantry.0.home
>> net home-x  => axis.0.home-sw-in
> That looks OK, but it's not enough to verify your HAL file is correct.
>
> The behavior you describe could happen if the search-vel input is
> incorrect, if you're using the limit output instead of the home output
> to feed to motion, or if your home switch signals have the wrong
> polarity (they should be high if the switch is "tripped").
There is ONLY a homing switch.  No limit.

Polarity is correct.

I didn't have an [Axis3] section at all in my ini, since nothing seemed 
to use it, all references for axis3 functions go back to axis0.  I did 
copy Axis0 section into an Axis3 though, so search-vel does exist now 
for axis3.

Didn't help.

I did see this: when it homes, if X1 is ahead of X2, it stops both on 
X1.  However if X2 is ahead of X1, it ignores BOTH homing switches.  It 
just drives both  into the stops and grinds forever. The sensors are far 
enough from the stops now that there's room to go far enough to pass the 
switches which does create an awkward case where it may start already 
past it and  grind on the endstop without ever seeing the sensor.

Danny

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

Re: [Emc-users] How do stepper drivers work?

[Danny Miller]

Stepper motors' drive is AC-like if moving, but DC when stopped. There 
is a common misconception that you target average current, but that's 
not correct.

There are two windings in bipolar steppers.  If you have a motor rated 
at 2 amps with 2 ohm DC winding,  that rating allows 2A, 8W total, 
because it can stop on a winding.  The motor is rated for 8W 
dissipation, presuming it is free-air with a reasonable ambient.  Also 
the construction guarantees you won't demagnetize the magnets or cause 
core saturation at 2A on a winding.

Alternately, same bipolar, it may stop between phases in a microstep 
situtation.  That's 45 deg, so with a 2A peak, a microstep driver 
delivers 1.414A^2 *2ohm * 2 windings =8W total spread between two 
windings.  It's spread out so the hottest winding is cooler because it's 
only doing 4W, but that's not where the rating comes from.  Still 2A per 
winding.

There is no case where a driver set for a 2A peak will put 2A on EACH 
winding. THAT is the mysterious, confusing part of the spec if you 
overthink it.  It says "2A/phase", 2 phases, doesn't that add up to 4 
amps?  Not really, that number is based on the assumption that you drive 
only one phase or a phase A=sin(angle)*2A + phase B=cos(angle)*2A.  
Industry std assumption.

There's ONE catch.  In the case of UNIPOLAR, you break each phase into 2 
windings.  With the same above motor tapped like that, each winding ould 
be 1ohm and still rated for 2A.  You could drive the two half-windings 
in series, making a 2ohm 2amp winding and that's the exact same motor 
the bipolar case above already is.  Or, unipolar is only pulling down 
one half-winding at a time (the only reason to do this is it's a simpler 
driver, and mostly irrelevant now because driver costs are quite low for 
bipolar drives).

In that unipolar case, you can still ONLY apply 2A/winding, even though 
you're using half-windings and only 1 of 4 windings is being used when 
it's on-pole.  Only 4W total motor heat, but only 1/(2^0.5) the torque 
is possible.  You might say "well the motor can handle 8W, so I'll give 
it 2.83 amps and get 100% of the bipolar torque again".  But this will 
exceed the winding's rating, even if it's only a half-winding.  This is 
why unipolar doesn't make much sense, a larger motor to meet a torque 
requirement is expensive, whereas even the bipolar driver is pretty cheap.

Danny

On 9/3/2016 3:02 PM, Kirk Wallace wrote:
> On 09/03/2016 11:32 AM, Andy Pugh wrote:
>>
>>> On 3 Sep 2016, at 21:12, Danny Miller <dan...@austin.rr.com>
>>> wrote:
>>>
>>> l approximation of current targets whose PEAK value is equal to the
>>> motor current rating.
>> Are you sure? I don't have any information at all but would expect it
>> to be average current.
> My understanding is that one should check the motor and driver specs for
> voltage limit, and supply with the lower of the two voltage limits.
> Then start with a low driver current limit, let the motor come up to
> temp, test the temperature, then turn the current up and repeat until
> the motor runs at the desired temperature (usually barely touchable). I
> believe that the motor sitting at idle will produce the highest temp.
>
> In addition to the Jones link, this should also have some good
> information: http://www.geckodrive.com/support.html
>
>


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

Re: [Emc-users] How do stepper drivers work?

[Danny Miller]

 From the user side, you provide 5v step/dir signals and need to pay 
attention to which active edge the driver uses, and know the microstep 
multiplier inside the drive.

Inside, it's a constant-current PWM.  If it's microstepping, it advances 
through a sinusoidal approximation of current targets whose PEAK value 
is equal to the motor current rating.

Steppers run on a constant current regardless of load.  However a 
stopped or unloaded stepper has minimal back EMF and draws little power 
(voltage * current).  Under heavy load, back EMF and the power 
increases, this is observable as greater power draw from the supply.  
Motor heat increases only a small amount with load.

Stepper drivers vary a LOT in effectiveness.  Short answer is 
AM882/DM542 DSP drives are "the best".   MX4660 is also a DSP but lacks 
the fine-tuning control interfaces of the other 2x. Geckdrive G540 was 
one of the first quality drives, still good, but not as great as the 
other ones here.

Danny


On 9/3/2016 12:58 PM, Chris Albertson wrote:
> I want to understand stepper motor controllers.   Yes I know I can just
> read the specs and buy one but I want to understand what's inside.
> Preferably someone has a link
>
> My use of stepper motors has been, I guess primitive.  I can write software
> to toggle bits on an output port and then I connect these logic level bits
> to MOSFETs that switch a power supply to the coils in the motor.  I can do
> full and half steps this way.
>
> But now I see something like the Geko driver that takes 40 volts input.  I
> know that if it simply switched that 40V into the coils it would burn out
> the motor in short order.  As the motor is typical rated for 7 volts DC or
> close to that.  OK I can understand that if it is switching an inductive
> load the voltage raise is slower and maybe not get past 7V before the coil
> is switched off.   But what if the motor is running slow?   So my final
> guess is that these kinds of drivers are supplying a constant current and
> applying whatever volts is required up to the supply voltage.
>
> Is it as simple as that? A constant current power supply and some MOSFET
> H-bridge switches?
>
>
>
>


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

Re: [Emc-users] Using 7i76 I/O as serial Rx/Tx?

[Danny Miller]

I'd asked them about using a 7i92 to drive an RS485 transceiver, which 
would require 4 signals.   They said the functionality was planned but 
no specific date.

I asked if it would just appear as a Com port, he said no it'd be a 
specific LinuxCNC driver.

There's a "UART" listed on the 7i92's features.  Actually since I have a 
hardware RS232/RS485 bridge I could still use that, if the config for it 
has been made.  But I don't know if it has.

Danny

On 7/7/2016 7:17 PM, Axel Zöllich wrote:
> I'ld like to control a Hungyang VFD via RS485 and the linuxcnc hy_vfd module.
> As I#ve got the machine running with the well known 5i25/7i76 combo, i'ld like
> to use one 7i76 input and one 7i76 output line as TTL serial and only would
> need an RS485 driver to get a working composition.
>
> Is it possible to compile/configure the 5i25/7i76s firmware to provide an
> /dev/ttySx to the linux kernel?
>
>
> Axel
>
> --
> Attend Shape: An AT Tech Expo July 15-16. Meet us at AT Park in San
> Francisco, CA to explore cutting-edge tech and listen to tech luminaries
> present their vision of the future. This family event has something for
> everyone, including kids. Get more information and register today.
> http://sdm.link/attshape
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Attend Shape: An AT Tech Expo July 15-16. Meet us at AT Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Spindle control panel breaks LinuxCNC?

[Danny Miller]

Well, I recreated it and confirmed cycling VFD power without rebooting 
LinuxCNC makes the spindle run again.

I looked into the HAL after the "Spindle CW" button breaks everything:
enable TRUE
is_alarm FALSE
is_at_speed FALSE
is_ready TRUE
is_running TRUE
reverse FALSE
run TRUE
watchdog_out TRUE

compared with actually having it running with M3:
is_at_speed becomes TRUE (duh)
watchdog_out is FALSE

Well, that's odd.  IIRC the watchdog is not getting data back from the 
VFD in a timely fashion.  It is somehow locked up so that does seem 
consistent with the situation.  But no idea how it's getting locked up.  
The x200 VFD code sends a speed, dir, and run command and reads back the 
coils.



Danny

On 6/16/2016 11:04 AM, dan...@austin.rr.com wrote:
> It's modbus, sorry forgot to mention that.
>
> What baffles me is the Spindle buttons on the panel, AFAIK, just map to the 
> same spindle-run in the HAL that M3 goes to.  I remember looking at the HAL 
> monitor on this weeks ago and IIRC it was "spindle run true, but 
> spindle-isrunning false".  The spindle isn't moving at all.  I'll recheck the 
> HAL and take proper notes.
>
> And like I say, the VFD won't run again even if you reboot LinuxCNC.  It will 
> run if I cycle power on the VFD.  So, it's like it sent a toxic command to 
> the VFD that changed a reg to something unusable.  I did make that x200_vfd.c 
> from the WJ200_vfd.c VFD code, but it wasn't a major change, and it's simple, 
> there's a run command, cw/ccw command (I made it so CCW just turns into CW), 
> rpm command, and reads is-running-at-speed and an error bit.
>
> It don't see where it could deliver something to de-configure the VFD, nor 
> why the Spindle panel buttons would do something different through the HAL 
> than you'd get by MDI M3.
>
> Danny
>
>
>  andy pugh <bodge...@gmail.com> wrote:
>> On 16 June 2016 at 07:57, Danny Miller <dan...@austin.rr.com> wrote:
>>   My MPG's "spindle" button works.
>>
>> On the default panel, ...  Clicking any of that doesn't make the spindle go,
>>
>> It's not JUST that.  Once you click on any of that, the VFD *will never
>> run again* until power is cycled.
> How fascinating..
> Can we see your HAL files?
>
> Is the spindle controlled by Modbus or DC voltage? What hardware?
>


--
Attend Shape: An AT Tech Expo July 15-16. Meet us at AT Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] G0 vs G1-specific functions?

[Danny Miller]

Two-parter here:

I was looking over another carving job on the preview window and 
realized there's really no way to tell where the work is.  That is, we 
go up to Z=2" for the start/end coords, but I need to know where the 
carving itself starts.  It's shown in a different color because it's all 
G1 (or G1 arcs)  vs G0.  But the limits are shown only for the limits of 
G1 AND G1 travel.  Can the preview show the limits of G1 only?

In a similar thread, I'm doing 3D carving with some aggressive 
acceleration profiles.  The XHC mpg is SUPER rough when stepping around, 
it's supposed to allow you to use different accelerations for mpg 
inputs, but none of that section has ever worked.  The mpg works but the 
section's options don't.

So, I should try again to figure out why it won't take the XHC options, 
but I started to think differently- we have different speeds for G0 and 
G1.  Wouldn't it make sense to have different accelerations too?  See, 
that 3D carving, I'm tuning my machine for like 150 ipm right now, it 
may not make sense to carve faster than that, but I do need to maximize 
acceleration.

Thing is, if you tune for max acceleration the motors can do without a 
stall at 150 ipm, then the 600 ipm rapids WILL stall it because it uses 
the same accelerations.  I can just limit rapids to 150 ipm so they 
agree with the G1 acceleration limit chosen, but then it has no "rapids" 
per se, and setting the machine up is a chore.

So, is there a way to have RAPID_ACCELERATION  (for G0 alone) as a 
separate thing from MAX_ACCELERATION for the axes, used for G1 motion?

Danny


--
Attend Shape: An AT Tech Expo July 15-16. Meet us at AT Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Major lags when zeroing axes

[Danny Miller]

I work a lot with 3D carving.  The files are often quite huge.

When I go to zero the work coordinates, EACH axis results in 
recalculating the entire file, which can take minutes.  Really it 
shouldn't require any recalculation (Mach3 doesn't) since it's just 
offsetting the coordinates, but I can imagine how the coding could try 
to do that.

Is there any way around it?

Also, minor note, but I don't see any way to UNLOAD the file so I can 
zero the work coordinates without interference.  All I could do was open 
a dummy file instead.

Danny


--
Attend Shape: An AT Tech Expo July 15-16. Meet us at AT Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Moving axis while paused

[Danny Miller]

Recently discovered this:

When paused, the axes can't be moved by the user.

While that sounds like a safe protection at first, that's actually a 
critical problem.  It is not uncommon to have to "fix" something 
mid-run.  In a few cases I've seen the CNC router pop freed pieces out 
and they land on the target of a plunge.  When that happens, the bit 
plunged into it and you have an unbalanced chunk of wood free-spinning 
on the bit.


The proper answer to that is to pause, stop the spindle, lift the Z 
enough to get the debris off the spindle, return to where the bit left 
off, restart spindle, and resume.  But, like I say, manual control of 
the axes is locked out and I can't fix it.


Danny


--
Attend Shape: An AT Tech Expo July 15-16. Meet us at AT Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Spindle control panel breaks LinuxCNC?

[Danny Miller]

I have an X200 VFD.  M3,M5,S18000 work fine, from MDI or code. My MPG's 
"spindle" button works.

On the default panel, there's a "Spindle" button flanked with a CW/CCW 
indicator.  Clicking any of that doesn't make the spindle go, even with 
"S" set to a valid RPM.


It's not JUST that.  Once you click on any of that, the VFD *will never 
run again* until power is cycled.


Any idea why this is?  It's a pretty serious problem because others are 
using this machine.


Danny


--
What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic
patterns at an interface-level. Reveals which users, apps, and protocols are 
consuming the most bandwidth. Provides multi-vendor support for NetFlow, 
J-Flow, sFlow and other flows. Make informed decisions using capacity planning
reports. http://pubads.g.doubleclick.net/gampad/clk?id=1444514421=/41014381
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Vectric Aspire 8.0 generates bad code for LinuxCNC

[Danny Miller]

Max vel 15, max accel 50.

That cuts like 1" radius arcs.  But the docs are  saying that's 
basically what G64 with no P-parameter does, basically totally 
disregarding accuracy to keep the feedrate up.

Danny

On 6/4/2016 4:16 PM, andy pugh wrote:
> On 4 June 2016 at 20:20, Danny Miller <dan...@austin.rr.com> wrote:
>> G64
>>
>> No "P0.001" parameter.  Well that's pretty damn toxic, it'll create huge
>> arcs instead of what it's supposed to do.
> If "best-speed" gives huge arcs I would suspect a miss-match between
> the cutting speed and the machine acceleration limits.
> What are your accel limits based on?
>


--
What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic
patterns at an interface-level. Reveals which users, apps, and protocols are 
consuming the most bandwidth. Provides multi-vendor support for NetFlow, 
J-Flow, sFlow and other flows. Make informed decisions using capacity 
planning reports. https://ad.doubleclick.net/ddm/clk/305295220;132659582;e
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Vectric Aspire 8.0 generates bad code for LinuxCNC

[Danny Miller]

I've seen this several times...

So Vectric Aspire 8.0 is a great program.  Most of the time it works 
ok.  However, we've found Gcode generation sometimes yield just this:

G64

No "P0.001" parameter.  Well that's pretty damn toxic, it'll create huge 
arcs instead of what it's supposed to do.  It'd doomed, but you won't 
know it's going to fail until you start running it.  The preview window 
does NOT reflect the dire problems with this code. You'd have to 
manually inspect all the files.

Well, for one, I'm a bit confused why LinuxCNC G64 with no parameter 
defaults to such a probably-harmful mode inside LinuxCNC to begin with.  
It's an interesting concept but unlikely to be of use and radically 
different than the G64 Px mode.  That is, the radically different mode 
that happens with G64 with no "P" really should be a G64.1, IMHO.

But, anyhow, any idea what's up with Aspire?  I'm not entirely sure 
what's triggering this no-P generation.  When working with someone else, 
I recall we concluded there was a "LinuxCNC arcs-inch" which generated a 
P, and a "LinuxCNC no arcs-inch" which did NOT generate a P and was 
doomed to fail.  But I'm trying it on another machine and I can't find 
the failing option.

Danny

--
What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic
patterns at an interface-level. Reveals which users, apps, and protocols are 
consuming the most bandwidth. Provides multi-vendor support for NetFlow, 
J-Flow, sFlow and other flows. Make informed decisions using capacity 
planning reports. https://ad.doubleclick.net/ddm/clk/305295220;132659582;e
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Copying an installation

[Danny Miller]

 grep ' ifname ' /var/log/dmesg to see
>>>> what it did call it, you should get something that resembles this:
>>>>
>>>> gene@coyote:~$ grep ' ifname ' /var/log/dmesg
>>>> [1.401462] forcedeth :00:08.0: ifname eth0, PHY OUI 0x5043
>>>> @ 1, addr 00:1f:c6:62:fc:bb
>>>> [1.929064] forcedeth 0000:00:09.0: ifname eth1, PHY OUI 0x5043
>>>> @ 1, addr 00:1f:c6:63:07:97
>>>> (word wrapped, darn it, what you want is the string after the
>>>> first ifname. in this example eth0)
>>>>
>>>> then use an editor as root to look at the
>>>> /etc/networking/interfaces file, and rename the stanza for eth0 to
>>>> whatever the system found and named it to in the /var/log/dmesg
>>>> file.
>>>>
>>>> You should at that point be able to do a "sudu service restart
>>>> networking" and have the ability to "ping -C2 yahoo.com" and get a
>>>> 2 normal ping responses from yahoo.com which indicates that
>>>> networking is now working.
>>>>
>>>>> It is an AMD64 though, and the installation was an i686.
>>>> A non-issue AFAIK.
>>>>
>>>>> Danny
>>>>>
>>>>>  "Peter C. Wallace" <p...@mesanet.com> wrote:
>>>>>> On Fri, 27 May 2016, Danny Miller wrote:
>>>>>>> Date: Fri, 27 May 2016 00:11:08 -0500
>>>>>>> From: Danny Miller <dan...@austin.rr.com>
>>>>>>> Reply-To: "Enhanced Machine Controller (EMC)"
>>>>>>>  <emc-users@lists.sourceforge.net>
>>>>>>> To: emc-users@lists.sourceforge.net
>>>>>>> Subject: Re: [Emc-users] Copying an installation
>>>>>>>
>>>>>>> I do recall we went through much more than expected just
>>>>>>> getting all that installed.  And I don't have a complete
>>>>>>> list of all that was done.
>>>>>>>
>>>>>>> I did poke around again on this machine.
>>>>>>>
>>>>>>> Mesaflash says the card's there at 10.10.10.10.
>>>>>>>
>>>>>>> After launching LinuxCNC, the VFD does respond to commands
>>>>>>> just fine.
>>>>>>>
>>>>>>> I experimented with the FERROR value- it'll allow the
>>>>>>> coordinates to change significantly before throwing an
>>>>>>> error, but the axes will never move regardless.  The 7i92
>>>>>>> won't put out steps at all. I don't have any enable line on
>>>>>>> it.
>>>>>>>
>>>>>>> Danny
>>>>>> Did you try swapping hard drives as someone suggested, in case
>>>>>> something was forgotten when moving?
>>>>>>
>>>>>> (when linux using generic kernels its much easier to just swap
>>>>>> hard drives than moving a setup to a new machine)
>>>>>>
>>>>>>> On 5/22/2016 6:27 PM, andy pugh wrote:
>>>>>>>> On 22 May 2016 at 19:51, Danny Miller
>>>>>>>> <dan...@austin.rr.com>
>>>> wrote:
>>>>>>>>> Any advice, folks?  I've gotta move off that Dell machine
>>>>>>>>> ASAP and really want to avoid a whole reinstall.
>>>>>>>> I would suggest a complete reinstall of the OS and
>>>>>>>> LinuxCNC, but keep the same config files. The LinuxCNC
>>>>>>>> config files should be entirely portable.
>>>>>>> 
>>>>>>> --  What NetFlow Analyzer can do for you?
>>>>>>> Monitors network bandwidth and traffic patterns at an
>>>>>>> interface-level. Reveals which users, apps, and protocols
>>>>>>> are consuming the most bandwidth. Provides multi-vendor
>>>>>>> support for NetFlow, J-Flow, sFlow and other flows. Make
>>>>>>> informed decisions using capacity planning reports.
>>>>>>> https://ad.doubleclick.net/ddm/clk/305295220;132659582;e
>>>>>>> ___
>>>>>>> Emc-users mailing list
>>>>>>> Emc-users@lists.sourceforge.net
>>>>>>> https://lis

Re: [Emc-users] Copying an installation

[Danny Miller]

I do recall we went through much more than expected just getting all 
that installed.  And I don't have a complete list of all that was done.

I did poke around again on this machine.

Mesaflash says the card's there at 10.10.10.10.

After launching LinuxCNC, the VFD does respond to commands just fine.

I experimented with the FERROR value- it'll allow the coordinates to 
change significantly before throwing an error, but the axes will never 
move regardless.  The 7i92 won't put out steps at all. I don't have any 
enable line on it.

Danny

On 5/22/2016 6:27 PM, andy pugh wrote:
> On 22 May 2016 at 19:51, Danny Miller <dan...@austin.rr.com> wrote:
>> Any advice, folks?  I've gotta move off that Dell machine ASAP and
>> really want to avoid a whole reinstall.
> I would suggest a complete reinstall of the OS and LinuxCNC, but keep
> the same config files. The LinuxCNC config files should be entirely
> portable.
>


--
What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic
patterns at an interface-level. Reveals which users, apps, and protocols are 
consuming the most bandwidth. Provides multi-vendor support for NetFlow, 
J-Flow, sFlow and other flows. Make informed decisions using capacity 
planning reports. https://ad.doubleclick.net/ddm/clk/305295220;132659582;e
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Copying an installation

[Danny Miller]

Any advice, folks?  I've gotta move off that Dell machine ASAP and 
really want to avoid a whole reinstall.

No idea why it works up to the point where I try to move axes and get 
joint following error just right there.

Danny

On 5/21/2016 3:07 AM, Danny Miller wrote:
> OK, we cloned the drive to an SSD off the Intel-based machine, now it's
> on an AMD64.
>
> It found the ethernet 7i92 card no problem.  However, any attempt to
> move an axis throws a Joint Following Error immediately.  So, it don't work.
>
>   >uname -a
> Linux localhost 3.2.0-4-rt-686-pae #1 SMP PREEMPT RT Debian 3.2.78-1
> i686 GNU/Linux
>
> Ran a latency test, found the number was surprisingly high:
> servo 1ms interval 1056515 jitter  58515
> baseinterval  82973 jitter 57973
>
> What concerns me is the i686 installation on an AMD64 machine. Does that
> need to be changed?
>
> Danny
>
> On 5/8/2016 6:37 AM, John Thornton wrote:
>> While dd is touted as easy and convenient it is indeed a linux guru tool
>> only with many gotchas and very difficult to use. Here is an example dd
>> comand line for your viewing pleasure.
>>
>> sudo dd if=/dev/sdb1 of=/dev/sdc1 bs=4K count=2654720
>>
>> and how the numbers are arrived at lol
>>
>> fired up a python REPL as a calculator and checked partition and
>> filesystem sizes
>> for /dev/sdb1 I calculated (end-start)*sector size
>> from tune2fs you multiply block count by block size
>> divide two times by 1024 to get MB
>> divide the partition size by 4k to get a proper block count for dd
>>
>> I use Clonezilla almost everyday and if you clone a drive the receiving
>> drive must be the same exact size or larger. If they are not the same
>> exact drives then you must use the expert mode and tell Clonezilla to
>> not check the target drive size.
>>
>> The easy way is to copy the home directory then install from a live cd
>> then copy the linuxcnc directory to the new drive.
>>
>> JT
>>
>>
>> On 5/7/2016 10:34 AM, Danny Miller wrote:
>>> So as per previous thread, I may be moving this to another PC.  And
>>> probably change to an SSHD.
>>>
>>> I spent a ridiculous amount of time getting the OS and LinuxCNC RT
>>> installed and configured.  And I didn't do all of it myself so it's not
>>> a straight shot to repeat.  I'm just saying if I start from scratch
>>> again, this will not be a quick operation.
>>>
>>> How reliable is it to just copy the entire installation to a new drive
>>> on a new (different) machine?
>>>
>>> On Windows, I've just moved my old drive to a new machine a number of
>>> times.  But I was foiled when I replaced my AMD-based motherboard with
>>> an Intel-based MB, no boot.   My LinuxCNC machine is currently on an
>>> Intel Core Duo, and I might change to an AMD-based machine. Will it even
>>> be able to boot?  Will it be easy to adapt to an AMD without a full
>>> reinstall?  What's it take to adapt it?
>>>
>>> I was told there's a convenient tool or method for creating an image of
>>> the whole installation (OS and LinuxCNC) and packing it up.  Is that
>>> true?  'Cause I sure wish one of those had been on Bittorrent, that
>>> would have saved my a crazy amount of time.  If it can be done maybe
>>> I'll do it and put it out there for others.
>>>
>>> Danny
>>>
>>>
>>>
>>>
>>> --
>>> Find and fix application performance issues faster with Applications Manager
>>> Applications Manager provides deep performance insights into multiple tiers 
>>> of
>>> your business applications. It resolves application problems quickly and
>>> reduces your MTTR. Get your free trial!
>>> https://ad.doubleclick.net/ddm/clk/302982198;130105516;z
>>> ___
>>> Emc-users mailing list
>>> Emc-users@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/emc-users
>> --
>> Find and fix application performance issues faster with Applications Manager
>> Applications Manager provides deep performance insights into multiple tiers 
>> of
>> your business applications. It resolves application problems quickly and
>> reduces your MTTR. Get your free trial!
>> https://ad.doubleclick.net/ddm/clk/302982198;130105516;z
>> ___
>> Emc-

Re: [Emc-users] Copying an installation

[Danny Miller]

OK, we cloned the drive to an SSD off the Intel-based machine, now it's 
on an AMD64.

It found the ethernet 7i92 card no problem.  However, any attempt to 
move an axis throws a Joint Following Error immediately.  So, it don't work.

 >uname -a
Linux localhost 3.2.0-4-rt-686-pae #1 SMP PREEMPT RT Debian 3.2.78-1 
i686 GNU/Linux

Ran a latency test, found the number was surprisingly high:
servo 1ms interval 1056515 jitter  58515
baseinterval  82973 jitter 57973

What concerns me is the i686 installation on an AMD64 machine. Does that 
need to be changed?

Danny

On 5/8/2016 6:37 AM, John Thornton wrote:
> While dd is touted as easy and convenient it is indeed a linux guru tool
> only with many gotchas and very difficult to use. Here is an example dd
> comand line for your viewing pleasure.
>
> sudo dd if=/dev/sdb1 of=/dev/sdc1 bs=4K count=2654720
>
> and how the numbers are arrived at lol
>
> fired up a python REPL as a calculator and checked partition and
> filesystem sizes
> for /dev/sdb1 I calculated (end-start)*sector size
> from tune2fs you multiply block count by block size
> divide two times by 1024 to get MB
> divide the partition size by 4k to get a proper block count for dd
>
> I use Clonezilla almost everyday and if you clone a drive the receiving
> drive must be the same exact size or larger. If they are not the same
> exact drives then you must use the expert mode and tell Clonezilla to
> not check the target drive size.
>
> The easy way is to copy the home directory then install from a live cd
> then copy the linuxcnc directory to the new drive.
>
> JT
>
>
> On 5/7/2016 10:34 AM, Danny Miller wrote:
>> So as per previous thread, I may be moving this to another PC.  And
>> probably change to an SSHD.
>>
>> I spent a ridiculous amount of time getting the OS and LinuxCNC RT
>> installed and configured.  And I didn't do all of it myself so it's not
>> a straight shot to repeat.  I'm just saying if I start from scratch
>> again, this will not be a quick operation.
>>
>> How reliable is it to just copy the entire installation to a new drive
>> on a new (different) machine?
>>
>> On Windows, I've just moved my old drive to a new machine a number of
>> times.  But I was foiled when I replaced my AMD-based motherboard with
>> an Intel-based MB, no boot.   My LinuxCNC machine is currently on an
>> Intel Core Duo, and I might change to an AMD-based machine. Will it even
>> be able to boot?  Will it be easy to adapt to an AMD without a full
>> reinstall?  What's it take to adapt it?
>>
>> I was told there's a convenient tool or method for creating an image of
>> the whole installation (OS and LinuxCNC) and packing it up.  Is that
>> true?  'Cause I sure wish one of those had been on Bittorrent, that
>> would have saved my a crazy amount of time.  If it can be done maybe
>> I'll do it and put it out there for others.
>>
>> Danny
>>
>>
>>
>>
>> --
>> Find and fix application performance issues faster with Applications Manager
>> Applications Manager provides deep performance insights into multiple tiers 
>> of
>> your business applications. It resolves application problems quickly and
>> reduces your MTTR. Get your free trial!
>> https://ad.doubleclick.net/ddm/clk/302982198;130105516;z
>> ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>
> --
> Find and fix application performance issues faster with Applications Manager
> Applications Manager provides deep performance insights into multiple tiers of
> your business applications. It resolves application problems quickly and
> reduces your MTTR. Get your free trial!
> https://ad.doubleclick.net/ddm/clk/302982198;130105516;z
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


--
Mobile security can be enabling, not merely restricting. Employees who
bring their own devices (BYOD) to work are irked by the imposition of MDM
restrictions. Mobile Device Manager Plus allows you to control only the
apps on BYO-devices by containerizing them, leaving personal data untouched!
https://ad.doubleclick.net/ddm/clk/304595813;131938128;j
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Linuxcnc and Arduino

[Danny Miller]

I'm not sure what context you're saying you have for the XHC.

I have he XHC-HB04 on LinuxCNC  RT on the Linux RT kernel.

It's a VERY smooth response, but partly because my machine is very fast.

There is, however, a case for time delay if you select jogmode=normal 
and not =vnormal on a slower machine, or with greater scale on the 
handwheel.

Depends on the jogwheel speed vs the machine's acceleration and max 
velocity.  IF you set the wheel to its highest speed and spin it fast, 
you might command a router to move 50" in one quick twist of the hand 
that will take the machine 5 sec to execute.

This rarely comes up with me.  My machine is very fast and you have to 
spin the wheel unnaturally aggressively at top speed to get ahead of it, 
and it's not possible to get ahead by more than like 1 sec.

It cannot always keep the commanded position AND stop immediately when 
the handwheel stops.  That's logically impossible.

If you don't like that, you can select jogmode=vnormal and it will 
always stop when the handwheel stops.  However, you lose sync with the 
wheel.  e.g. quickly spin the wheel +2 turns forward and it may only 
register 1.5 turns before the wheel being stopped means the machine 
stops moving.  Then turn back -2 turns slowly and the machine is -0.5 
turns worth of motion from where it started, even though the handwheel 
is where it started.

I don't like this functionality myself.  And even that vnormal mode only 
means the gantry begins decelerating when you stop turning the 
handwheel.  If the machine takes 0.75 sec to come to stop from jog 
speed, it will not come to a stop until 0.75 sec after you stop the 
handwheel.

I recommend you reduce scale if you don't like it.  e.g. if you would 
never turn more than 2 turns/sec on that handwheel, and your machine's 
got a quick acceleration and 5 inches/sec jog speed, the scale should 
be  ~2.5 in/turn so you never outrun it.  With acceleration being 
significant, there's no definitive answer.

Danny


On 5/17/2016 8:22 AM, Forum Deswysen wrote:
> I have a XHC - HB04 Wireless MPG during LinuxCNC HAL module.
> If I turn the crank too fast I have a time delay ???
>
> It's dangerous
>
> 2016-05-17 11:21 GMT+02:00 Forum Deswysen :
>
>> Hi Andy,
>>
>> With a python script and the serial port :
>>
>> https://www.youtube.com/watch?v=flajZMff28U
>>
>> https://ckcnc.wordpress.com/basic-arduino-emc-howto/python-emc-module/
>>
>> With arduino, today I 'm testing my order for tool change turret.
>> It works fine
>>
>>
>> @+
>>
>> Pierre
>>
>> 2016-05-16 23:48 GMT+02:00 andy pugh :
>>
>>> On 16 May 2016 at 15:03, Forum Deswysen  wrote:
 The reaction speed is the same as a button on a Parport entry?
>>> That rather depends on how the Arduino connects to LinuxCNC, but the
>>> answer is likely to be that it won't be anywhere near as fast.
>>>
>>>
>>> --
>>> atp
>>> "A motorcycle is a bicycle with a pandemonium attachment and is
>>> designed for the especial use of mechanical geniuses, daredevils and
>>> lunatics."
>>> — George Fitch, Atlanta Constitution Newspaper, 1916
>>>
>>>
>>> --
>>> Mobile security can be enabling, not merely restricting. Employees who
>>> bring their own devices (BYOD) to work are irked by the imposition of MDM
>>> restrictions. Mobile Device Manager Plus allows you to control only the
>>> apps on BYO-devices by containerizing them, leaving personal data
>>> untouched!
>>> https://ad.doubleclick.net/ddm/clk/304595813;131938128;j
>>> ___
>>> Emc-users mailing list
>>> Emc-users@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/emc-users
>>>
>>
> --
> Mobile security can be enabling, not merely restricting. Employees who
> bring their own devices (BYOD) to work are irked by the imposition of MDM
> restrictions. Mobile Device Manager Plus allows you to control only the
> apps on BYO-devices by containerizing them, leaving personal data untouched!
> https://ad.doubleclick.net/ddm/clk/304595813;131938128;j
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users


--
Mobile security can be enabling, not merely restricting. Employees who
bring their own devices (BYOD) to work are irked by the imposition of MDM
restrictions. Mobile Device Manager Plus allows you to control only the
apps on BYO-devices by containerizing them, leaving personal data untouched!
https://ad.doubleclick.net/ddm/clk/304595813;131938128;j
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Copying an installation

[Danny Miller]

So as per previous thread, I may be moving this to another PC.  And 
probably change to an SSHD.

I spent a ridiculous amount of time getting the OS and LinuxCNC RT 
installed and configured.  And I didn't do all of it myself so it's not 
a straight shot to repeat.  I'm just saying if I start from scratch 
again, this will not be a quick operation.

How reliable is it to just copy the entire installation to a new drive 
on a new (different) machine?

On Windows, I've just moved my old drive to a new machine a number of 
times.  But I was foiled when I replaced my AMD-based motherboard with 
an Intel-based MB, no boot.   My LinuxCNC machine is currently on an 
Intel Core Duo, and I might change to an AMD-based machine. Will it even 
be able to boot?  Will it be easy to adapt to an AMD without a full 
reinstall?  What's it take to adapt it?

I was told there's a convenient tool or method for creating an image of 
the whole installation (OS and LinuxCNC) and packing it up.  Is that 
true?  'Cause I sure wish one of those had been on Bittorrent, that 
would have saved my a crazy amount of time.  If it can be done maybe 
I'll do it and put it out there for others.

Danny




--
Find and fix application performance issues faster with Applications Manager
Applications Manager provides deep performance insights into multiple tiers of
your business applications. It resolves application problems quickly and
reduces your MTTR. Get your free trial!
https://ad.doubleclick.net/ddm/clk/302982198;130105516;z
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users