Re: [Emc-users] Beam Stiffening?

[Chris Albertson]

> On May 20, 2024, at 6:26 AM, Todd Zuercher via Emc-users 
>  wrote:
> 
> I did to exactly that.  With the dial indicator on the ends vs the center, it 
> moved 10 times more in the center of the gantry than it did on the ends.  I 
> will be the 1st to agree that the servo tuning could probably be better.  But 
> this thing really is(was) a wet noodle.  At 12ft long, with the ends 
> disconnected from the ball screws, +/-1 inch differences between the ends of 
> the gantry wasn't difficult to achieve.  The two servos driving them really 
> interact very little and behave independently of each other with one not 
> really affecting the tune of the other.  More than likely this has been the 
> true root cause of the majority of my servo tuning/carving issues with this 
> machine since day 1. (We bought it about 15yrs ago.)  Originally the 
> machine's two ends were driven by rack and pinions with a 12ft long torque 
> tube connected between the pinion gears.  I had been blaming most of the 
> wobble problems on torsional twisting of tube, belt squirm on the 7":1/2" 
> sprockets on the belt reduction pullies for the servo, and a huge inertia 
> imbalance for the servo, being used with a gear ratio more appropriate for a 
> stepper motor.
> 
> While I kind of liked the idea of using cable trussing, my colleges did not.  
> So we decided to go ahead with the steel tubing we had on hand.

Good.  The cable idea would not have worked.  You would have done better with a 
steel rod the same diameter as the proposed cable.   The rod is stiffer than 
the cable.  But either way, the cross sectional area of the cable or rod is 
tiny compared to the beam. It would have a small effect because of the small 
amount of steel added.

What you did with that steel tube is you made the beam much wider.  Stiffness 
goes up with the cube of the width.  If the beam is now twice as wide, it is 8 
times more stiff.

> 
> That said, I a cut a piece of the 2x4x1/8" wall steel tubing, drilled holes 
> in it spaced 1ft apart and mounted it on edge on the back of the beam.  It 
> has made a huge difference, and now the gantry no longer wobbles when homing 
> and hitting the center behaves approximately the same as hitting it on the 
> ends.


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

[Emc-users] A great example of CNC work and motion control.

[Chris Albertson]

Just by chance is anyone here interested in advanced machine control and CNC’d 
metal parts?



Here is a real product you can buy today for $16K that defines the above. It is 
just like our 3-axis milling machines but 43-axis. It just went on sale.  API 
is on GitHub.
See video —>  
https://www.unitree.com/images/Unitree%20G1%20EN%201080p%20%281%29.mp4

No connections to LCNC except to show what a few guys in China are making with 
their CNC mills.   

Already on another list, people are organizing some sort of group to buy one 
for shared access.  This is not at all like the Boston Dynamics “Atlas” which 
cost millions of dollars and was never intended to be sold.  This machine is 
currently being mass produced and is for sale today.

For the technically minded.   The robot has up to 43 motors, half of which are 
in the hands.  Each motor accepts commands for position, velocity, and torque 
and has two rotary encoders, one on the motor and one on the joint after the 
geared reduction drive.   

There is a network protocol to send the motor commands.  They need to go over 
WiFi unless you pay extra for a second computer that fits inside and then 
commands can go over Ethernet.

The other control mode works in cartesian coordinates and you can give 
high-level commands like “stand and balance” or “walk” or “move hand to X,Y,Z”.

The internal battery claims a 2-hour run time based on some average motion.

Mechanically, it is just some BLDC motors and sun-planet gear reductions of 
maybe about 6:1 or 9:1, lower than some might expect.  All of the rotation 
shafts are hollow to allow cables to be routed through the moving joints.  What 
is impressive is the machine work.  As seen in the video, It is light enough to 
be lifted by one of the engineers at Unitree.

No more science fiction or “some day you will be able to buy a robot…”. If 
you've got $16K it ships directly from China.  The company has a decent 
reputation for building good quality hardware.

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

Re: [Emc-users] Beam Stiffening?

[Chris Albertson]

I think I fell for the “beam is bending” idea too.   It might be.   The way to 
find out is to measure the beam center with the dial indicator and then measure 
the ends of the beam.   It might be that the entire beam is moving

Is the gantry belt driven?   Long belts can act like springs.  The solution is 
wider belts.  It is easy to see that a belt that is twice as wide is twice as 
stiff.

I’d measure movement at the ends before any more thinking about the beam.

As for modifying the beam, you have to model it.  Guessing and “eyeball 
engineering” generally does not work well.   Any fix is going to be very 
expensive.  It is best to know it will work.


Everything that you add to that beam also adds mass.   Mass is what you want to 
get rid of.





> On May 14, 2024, at 1:56 PM, gene heskett  wrote:
> 
> On 5/14/24 14:34, Eric Keller wrote:
>> Do something cheap because I'm not convinced it's the beam.  I've done
>> troubleshooting on things like this, and sometimes it's stiffness and
>> sometimes it's not stiffness. But it really doesn't make sense that it
>> would sit there and ring after a move, so you also may have some
>> tuning to do.  Possibly a notch filter?
>> Eric Keller
>> Boalsburg, Pennsylvania
>> On Mon, May 13, 2024 at 4:50 PM Todd Zuercher via Emc-users
>>  wrote:
>>> 
>>> Anyone have any brilliant ideas to stiffen a woefully inadequate cross beam 
>>> on a gantry router without adding too much mass?  What is there now is a 4" 
>>> x 8" rectangular 3/8" walled extrusion that is 145" long.
>>> 
>>> Under normal jogging commands the two servos control the ends of this 
>>> gantry reasonably well, but while the axis is homing the thing shakes and 
>>> wobbles terribly bad.
> 
> This, on 3rd or 4th read, sounds as if the two servo's are not in tune with 
> each other.  Tuning servo's is not my strong suit, (and the only servo I had 
> was destroyed by the new autotune pid in linuxcnc, it found settings that mde 
> it ocillate and fried a $125 motor in around a minute. But this would be a 
> lot easier to synchronize if stepper/servo's were used. Rigged with a home 
> switch, maybe a prox switch since its non contact, with logic rigged so they 
> can back away from home and move in sync the rest of the day, getting sync is 
> running toward home until the switch trips on that end of the beam, run 
> toward home until both ends have tripped, call that home. From then until 
> powerdown, both motors getting the same step/dir signals will be in sync till 
> the powerdown. No fighting because the two servo's are not in an identical 
> state of tune. Hanpose has nema 34 and 42 motors of 12 NM, probably with more 
> torque and speeds than your servo's. The best description is that they just 
> work. And they use much less power than regular steppers to get the job done. 
> A diff you can see in the power bill if replacing burn your hand regular 
> steppers.
> 
> How fast and how strong are the servo's you are using now? Gear ratio's too.
> 
>  Also If I put a dial indicator in the center of the bridge and hit the 
> bridge forward or backward it will flex and wobble enough to displace the 
> dial indicator +/-0.03 and it takes nearly a dozen wobbles to dampen it.  But 
> on the ends the servo's only have a few thousandths of give.
>>> 
>>> I'm less concerned about the actual stiffness and more worried about 
>>> dampening the wobble.
>>> 
>>> Todd Zuercher
>>> P. Graham Dunn Inc.
>>> 630 Henry Street
>>> Dalton, Ohio 44618
>>> Phone:  (330)828-2105ext. 2031
>>> 
>>> 
>>> ___
>>> 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, CET.
> -- 
> "There are four boxes to be used in defense of liberty:
> soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author, 1940)
> If we desire respect for the law, we must first make the law respectable.
> - Louis D. Brandeis
> 
> 
> 
> ___
> 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] Beam Stiffening?

[Chris Albertson]

This kind of design “works” only if you make the overall dimension MUCH larger. 
   It is an overall more efficient design but bmovimng material does not add 
streght of stiffness.  You would need to do something like scale the beam up to 
maybe twice its size then cut away half the metal.

The first step should be to look for a material with the best stiffness to 
weight ratio you can afford.  Then if you change the shape, you woiuld have to 
go outside of the current 4x8 dimensions.

“Strength” is not just the tensile strength of the material but, that tiimes 
the cross sectional area,  and then you multiply by the distance from a kind of 
“center line”.   

The problem is cost.  The machine has the aluminum extrusion likely because it 
was the loest cost reasonable solution.   They could have used a high grade of 
tool steel machine into a truss frame but that might cost more than some cars.

My suggestion of carbon fiber has because the materials are not super expensive 
but what you pay for is the huge amount of labor

Whatever you do it will cost a bunch more then that extrusion.  So be sure to 
have it modeled using finite element analysis.  You hate to spend $10K or more 
only to find iot made thins worse.



> On May 14, 2024, at 3:21 AM, Gregg Eshelman via Emc-users 
>  wrote:
> 
> A triangular tube with an isogrid pattern cut into it to reduce mass without 
> sacrificing stiffness. Could have it laser cut with slots on the fold lines 
> to make it easy for a sheet metal break to fold accurately. The design could 
> have tabs and slots to interlock on the joining edge. Then TIG weld along the 
> bend slots and joining edge. Weld it like they do top fuel dragster frames, a 
> little bit here, a little bit there - to eliminate warping.
> 
> Or it could be possible to design three panels to bolt together and to the 
> gantry using tabs and Rivnuts.
> 
> The round holes at the vertexes of the triangles wouldn't need to be cut, 
> except in places where you'd want Rivnuts to mount things.
> 
> For isogrid design there's the 1973 book 
> https://femci.gsfc.nasa.gov/isogrid/index.html Page 42 of the PDF has the 
> dimensions for the panels used for walls and floors in Skylab. The photos in 
> it are mostly useless since the PDF was apparently produced from a microfiche 
> of a FAXed (or early non-greyscale photostat) copy of an original printed 
> copy of the book.
> 
> In some dusty, forgotten file cabinet there must be an original printed copy 
> of
> Isogrid Design Handbook - NASA CR-124075, Rev. A, Feb. 1973
> 
> A triangular tube is more twist and bend resistant than a square, 
> rectangular, or round tube, and it is lower mass than a square or rectangular 
> tube. Even less mass with all the bits removed to cut an isogrid.
> 
> On Monday, May 13, 2024 at 02:49:51 PM MDT, Todd Zuercher via Emc-users 
>  wrote: 
> 
> Anyone have any brilliant ideas to stiffen a woefully inadequate cross beam 
> on a gantry router without adding too much mass?  What is there now is a 4" x 
> 8" rectangular 3/8" walled extrusion that is 145" long.
> 
> Under normal jogging commands the two servos control the ends of this gantry 
> reasonably well, but while the axis is homing the thing shakes and wobbles 
> terribly bad.  Also If I put a dial indicator in the center of the bridge and 
> hit the bridge forward or backward it will flex and wobble enough to displace 
> the dial indicator +/-0.03 and it takes nearly a dozen wobbles to dampen it.  
> But on the ends the servo's only have a few thousandths of give.
> 
> I'm less concerned about the actual stiffness and more worried about 
> dampening the wobble.
> 
> 
> ___
> 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] Beam Stiffening?

[Chris Albertson]

Adding anything inside is the worst place to add material.   Add it outside.   
Stiffness is the cube of the beam thickness, so you really want to make it 
bigger.

Then secondary to making it bigger is to improve the shape to remove those 
parallel sides.  

So it you are just going to epoxy something on, try adding something like a 
channel section to the top or side of the beam.




> On May 13, 2024, at 2:11 PM, andy pugh  wrote:
> 
> On Mon, 13 May 2024 at 21:51, Todd Zuercher via Emc-users <
> emc-users@lists.sourceforge.net> wrote:
> 
>> 
>> I'm less concerned about the actual stiffness and more worried about
>> dampening the wobble.
>> 
> 
> Maybe you could epoxy a smaller (aluminium?) extrusion or box inside the
> existing one? The epoxy interface should add some damping.
> 
> 
> -- 
> 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, 1912
> 
> ___
> 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] Beam Stiffening?

[Chris Albertson]

> On May 13, 2024, at 1:45 PM, Todd Zuercher via Emc-users 
>  wrote:
> 
> Anyone have any brilliant ideas to stiffen a woefully inadequate cross beam 
> on a gantry router without adding too much mass?  What is there now is a 4" x 
> 8" rectangular 3/8" walled extrusion that is 145" long.
> 
> Under normal jogging commands the two servos control the ends of this gantry 
> reasonably well, but while the axis is homing the thing shakes and wobbles 
> terribly bad.  Also If I put a dial indicator in the center of the bridge and 
> hit the bridge forward or backward it will flex and wobble enough to displace 
> the dial indicator +/-0.03 and it takes nearly a dozen wobbles to dampen it.  
> But on the ends the servo's only have a few thousandths of give.
> 
> I'm less concerned about the actual stiffness and more worried about 
> dampening the wobble.


What is the extrusion made of, I assume it is some kind of aluminum alloy.
The simplest but expensive option is to replace it with a stronger/stiffer 
material with the same dimensions.   Of course Titanium comes to mind but that 
is maybe not in the budget.Carbon fiber could work and it is possible to 
DIY carbon fiber beams with just hand tools.   I have made 4 meter long racing 
kayaks with carbon, using just a paint brush and scissors in one weekend.   

The first class way is to make a female mold and polish it well so the part 
looks nice.   The cheap way is to make one like they make surfboards.  You 
start with a foam block, shape it then wrap it in fiber and resin.

The neat thing about carbon composite is that you are not limited to the 
extrusion shape.   I would make the entire beam a compound curve with no flat 
or straight or cylindrical sections,  Maybe like a very elongated American 
football but with ovil cross section.

I like to use the car hood story.  A flat sheet of sheet steel is bendable by 
hand.  But after they stamp it into the shape of a car hood it becomes rigid.   
So rather then a square tube, way not oval but with a larger diameter in the 
center where all the bending force is?

The way you make it is to first make a full-size model out of wood and bondo. 
Do a test-fit and give it an automotice grade paint finsh and then paste wax.  
Make a fiber glass mold, then from that your part.   Yes that is a lot of work. 
 This is why you have an aluminum extrusion there now, because that was easy 
and cheap.






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

[Chris Albertson]

> On May 10, 2024, at 9:49 AM, gene heskett  wrote:
> 
> On 5/10/24 11:25, Chris Albertson wrote:
>> Rsync will copy data at the file system level.I think the OP wants to 
>> copy the partition tables and boot sector.But rsync can copy across a 
>> network and is a decent way to make a backup of your data.
>> Clonzilla loks like it can work.  I’ve always used “dd” because it is a two 
>> letter command and very easy to rember the exact spelling, and it just 
>> works.Clonzilla might be better for people who find it hard to type in 
>> long complex commands like “dd” and prefer a menu-based system.
> 
> All good advice. Choose your favorite.  The one I miss is the one that can 
> take an arm64 install, with all the additions to do a job, and back it up 
> over the network to a file that only the total used on that arm64 system. one 
> that can then be copied to a fresh u-sd card of much greater capacity, and 
> which will then on first boot, expand the partitions to use all of the new 
> u-sd cards capacity. I've been using 64G cards and had had no losses. 16G 
> cards are big enough but have a lifetime of around a year. I have some 128G 
> I'll use for the next install.

Raspberry Pi5 can use PCIe storage, no SD card is needed.   Eventually, we will 
all stop using SD cards for system drives.

I keep wanting to reconfigure my Pi4 to boot off the network with no SD card 
installed.   The Pi3’s networking was too slow for this and the Pi5 does not 
need it.  I might never get around to it.

So what I do today is a compromise.   I install Linux from an image file.  I 
never store my data on the SD card.   data is NFS mounted from the network 
server.   This also means the data is always available on my other computers.  
I never have to move it. and if an SD card dies, nothing is lost.

I never have to backup an SD card.


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

Re: [Emc-users] cloning hard drive

[Chris Albertson]

Rsync will copy data at the file system level.I think the OP wants to copy 
the partition tables and boot sector.But rsync can copy across a network 
and is a decent way to make a backup of your data.

Clonzilla loks like it can work.  I’ve always used “dd” because it is a two 
letter command and very easy to rember the exact spelling, and it just works.   
 Clonzilla might be better for people who find it hard to type in long complex 
commands like “dd” and prefer a menu-based system.



> On May 10, 2024, at 4:03 AM, gene heskett  wrote:
> 
> On 5/10/24 06:11, andrew beck wrote:
>> hey everyone
>> a bit off topic here
>> i have my main laptop that i want to clone the hard drive on it for a
>> identical laptop for a backup
>> this is used for running the linuxcnc machines and programming drawing etc
>> anyway just want to know what software people prefer for disk cloning i
>> have never done it before and i'm sure someone on here is a expert on it
>> cheers
>> andrew
> rsync can do that but please read the man page carefully. It can bite you 
> just as easily as it can help you.  I use it, but will not call myself an 
> expert.
> 
> Cheers, Gene Heskett, CET.
> -- 
> "There are four boxes to be used in defense of liberty:
> soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author, 1940)
> If we desire respect for the law, we must first make the law respectable.
> - Louis D. Brandeis
> 
> 
> 
> ___
> 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] cloning hard drive

[Chris Albertson]

I assume the laptop runs Linux.If so then you’d use  “dd” to clone the disk 
drive. “dd” is installed on every Linux system I’ve ever seem.  So you already 
have it.  This assumes you are copying it to an identical drive.

The trouble is that you can not have booted from the drive you are copying from 
because the OS will be writing to that drive as it is being copied.   So what 
you do is boot off the USB drive or if it is an older computer off the CDROM.   
 This way my assumption that you are running Linux is true because you would 
use a Ubuntu install image on the USB stick

Both drives need to be connected to the computer, the second one can be placed 
in a USB enclosure

Get to the command line and type dd if=/dev/sda of=/dev/sdb

SDA and SDB are place holders for the real names of the two drives, likely your 
drives have different names.  Do not swap the names of the drive or you will 
copy the empty drive over you data.  make a backup of you data first, then 
veriify you can read the backup, then clone the disk.  Rea the dd man page.  
Remember that “if" is “input file” or “of” is “output file"
For more info read the dd man page or Google “clone drive using dd”

All that said “cloning” the drive is not very usfull.  Betther to just backup 
the data and when the disk dies replace it with a better one (bigger, faster) 
reinstall the OS and then restore your data.  The clone will qickly become 
obsolite as you make changed and update software, best to just do a regular, 
noroaml backup.My computers mostly have a continous backup runnig that 
copies changes in real time so I loose at most an hour’s work




> On May 10, 2024, at 2:10 AM, andrew beck  wrote:
> 
> hey everyone
> 
> a bit off topic here
> 
> i have my main laptop that i want to clone the hard drive on it for a
> identical laptop for a backup
> 
> this is used for running the linuxcnc machines and programming drawing etc
> 
> anyway just want to know what software people prefer for disk cloning i
> have never done it before and i'm sure someone on here is a expert on it
> 
> cheers
> 
> andrew
> 
> ___
> 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] BeagleY-AI

[Chris Albertson]

> On May 1, 2024, at 8:55 AM, andy pugh  wrote:
> 
> 
> I think that the attraction of the Beagleboard was that the PRU could be
> used for step generation and encoder counting. 

This gets into the difference between a Linux PC and a microcontroller.The 
microcontroller has lots and lots of peripheral hardware, things like timers 
and PM generators, and hardware quadrature decoding and D/A converts and 
whatnot.   Whereas the typical PC only has “ports” that move data. The 
Beagleboard is kind of in the middle.   It is actually poor at doing either job 
but the fact that it had real-time outpit pins made it really attractive.

There are some really exelent hardware platforms that could be used for LCNC 
but who wants to get them to work when there are off-the-shelf solutions.  
Saving $300 on the hardware is not attractive if you have to write the software 
to make it work.


Look at this board.  It sells for $90 and could run LCNC.  It has a Raspberry 
Pi equivalent that comes with Debian Linux installed, has 2 GB RAM and fast 
32GB storage, and has 4 stepper moter drivers that are good for about 2 amps.  
Really, this is a Pi4, with breakout and drivers for $90 and has a warranty and 
some minimal level of customer support.The problem is that the LCNC 
ecosystem is not yet large enough to attract enough software developers to 
write firmware for every “every chip in the world”. So we are kind of stuck 
with what’s available now.

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

Re: [Emc-users] BeagleY-AI

[Chris Albertson]

The magoroty of people who do not want to make CNC tools their hoby and just 
want to cut metal, wood or plastic, they would be best sewrved by one of the 
“out of the box” solutions, like Acorn.  No one, except Torch sells an LCNC 
solution that come rewady to run with hardware and pre-istalled software.  Most 
users will want that.

Then if you are selling a turn-key setup, LCNC requires more expensive 
hardware.  90% of the market is going to be runnng a very simple 3-axis CNC 
router.  Yoiu can buy PCBs that have a poerfull microcontroller and some 
stepper moter drivers for $40 retail or have one made in china for 1/3rd that 
price. You then but $50 worth of electronics in a $10 enclosure and sell it for 
$300.

The end user will never in a million years complain that he can not edit the 
config file to run industrial servos over Ethercat.

On the other hand many people do seem to make CNC a hobby even if they really 
don’t need to make a ‘billion CNC’d parts.  For them LCNC is prefect and a 
machine that “just works” would be usless because they’d have nothing to do.

> 
> What I am having good luck with running all the fancy stuff for 3d printers 
> on them is the currently $65 bananapi-m5. All 4 usb ports are usb3 so speeds 
> are not a problem. I am just now bringing up an old Ender 5 Plus that died a 
> couple years ago, able to run at 30mm/second max because it comes with a puny 
> Y motor, but now has 2 more bigger higher voltage power supply's, the X 
> motors are now stepper/servo's, I belted the z motors together and unplugged 
> one so bed tilt is locked, lots of heavy flying weight is now CF tubing, much 
> lighter to throw around, and its loafing at 200mm/sec speeds.  What it 
> formerly took 3 days to make is now done in 17 hours. That bpi-m5 is talking 
> to the $52 control card that runs the printer with a single usb-C cable.
> 
> I have not tried linuxcnc on one of them but it runs fine on an rpi4b with 2 
> gigs of ram,

Gene, if that Pi4 is running Klipper, of course it is not loaded up.  Kipper 
pushes 100% of the real-time work onto the MCU.  The Pi only has to read the 
g-code file and do the motion planning and run the web server based GUI.  
Screen rendering and mouse tracking and all the low-level GUI stuff is done on 
the user’s web browser.LCNC is just the opposite, so you can’t compare.

Klipper does not care at all which OS you run.  It runs very well in a virtual 
machine or even on a Pi-zero.  I’m using a Pi3, 1GB and see only 8% CPU usage   
In fact you can run multiple copies of Klipper on one computer and drive 
multiple printers at the same time.

> running my 11x54 Sheldon lathe. Install the build-essential & the latest 
> python 3, hook it to a breakout board fed by a usb-3 cable, and build lcnc 
> from master's src. It ought to just work. Might have to build a rt kernel, 
> but my 3d printers don't seem to mind the current jammy offering.


But let’s say you were a sign maker and wanted to cut out plastic letters with 
a CO2 laser.  You bill your customers at $125 per machine-hour.  How many hours 
do you want to spend learning to build real-time kernels,  it costs you $125 
for every one of those non-billable hours.  And on top of that maybe marketing 
and sales are not getting done while you work on a DIY CNC project.   

Same with plumbing, It is not technically hard to repair pipes, snake drains, 
or replace a water heater.But what if you own a restaurant?   The owner is 
more than happy to pay the $250/hr rate to have two plumbers and a well-stocked 
truck show up and fix the problem because he has to close the business until 
the problem is fixed.   

So in an industrial setting DIY could very easily be much more expensive than 
buying a turn-key system or even hiring the work done for you.

But on the other hand, many of us are not in a situation where “toime is money” 
and LCNC is a good fit.  But it is not a good fit at all for everyone.








> It ID's itself as:
> 
> Linux bpi51e5p 6.6.16-current-meson64 #1 SMP PREEMPT.
> 
> Arm64 stuff is encroaching on our territory, its power miserly AND stable as 
> the Rock of Gibralter. Uptimes are from power outage to power outage.
>> Just curious if this is finally the step into the single small box with
>> LinuxCNC or MachineKit Turnkey CNC system for the older MACH or other users.
>> Or has the market now matured enough with other solutions like the Centroid
>> ACORN
>> https://www.centroidcnc.com/centroid_diy/acorn_cnc_controller.html
>> and it's pointless to even bother with LinuxCNC?
>>  John
>> ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>> .
> 
> Cheers, Gene Heskett, CET.
> -- 
> "There are four boxes to be used in defense of liberty:
> soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author, 1940)
> If we desire respect for the law, we must 

Re: [Emc-users] Big Tree Tech Re: Mesa Card Stepgens?

[Chris Albertson]

No it does not fail, we can always send new commands in real time to be 
executed “right away” (at the current time) and make the queued execution 
system look and act as if there were no queue.   A command taged with the 
current time will jump to the front of the queue.

The system also allows for very delayed execution of commands, like “in five 
minutes”.   This works well because you can queue a macro to do something like 
poll the voltage of a battery and then place itself back in the queue for five 
minutes later.  But as said with a current time or a special tag that says 
“now” it reverts back to the way LCNC works.

Remember that a time-tagged queue is not executed in order.  It is not  FIFO, 
so a command that is placed in the queue, even with 100 commands waiting does 
not need to wait for the others.


> On Apr 17, 2024, at 5:05 PM, Ray Henry  wrote:
> 
> On Wed, 2024-04-17 at 09:13 -0700, Peter Wallace wrote:
>> 
>>> LCNC really should be doing this.  If the Measa cards would
>>> maintain a queue 
>>> and a synchronized clock we would not care at all about latency. 
>>> Klipper 
>>> proves the idea works.
>> 
>> This is basically what Mach 3/4 do
>> 
>> This fails as soon as you need the control to repond to
>> feedback in real time (Spindle synchronized motion, plasma THC
>> etc etc) Its possible of course to build this feedback into the
>> interface hardware, but there is a big advantage to having
>> this feedback in LinuxCNC where is works with any interface
>> hardware, is extensible and open, and where the feedback
>> control math/logic has unlimited resources.
>> 
>> 
>> 
>> Peter Wallace
> 
> 
> The message on my 25+ year old coffee cup from the NIST Intelligent
> Systems Division reminds me at least three times a week to "SENSE ->
> MODEL -> ACT." in that order.  Thanks. 
> Peter.
> 
> Ray
> 
> 
> 
> 
> ___
> 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] Big Tree Tech Re: Mesa Card Stepgens?

[Chris Albertson]

For those on this LCNC list who have not seen Klipper, Klipper’s architecture 
is very much like Linux CNC’s.   Like LCNC, the g-code interpreter, kinematics, 
and motion planning live in a “real computer” like a Raspberry Pi.  Or I was 
running it on a virtual machine running Linux on my Mac mini,  Any Linux 
computer or virtual computer will do.  There is no need for a real-time kernel. 
 Klipper is written in Python and is pretty easy to extend with “plug-ins”.
All the real-time critical timing is done on a microcontroller.   The micro can 
be even an 8-bit Arduino, but more likely it is an STM32, or RP2040.   The 
micro(s) connect to the Pi with a serial interface, typically USB.   Lantancy 
in the interface is unimportant.

The trick is that the Linux PC sends commands to the microcontroller and each 
command is time-tagged with the time it should be executed.  the 
microcontroller maintains a queue and does the command at the right time.  The 
Linux PC and the controllers also use a protocol to synchronize their clocks.   
   The time sync works. well enough that you could in theory put each axis on a 
different microcontroller.   The Pi only needs to be fast enough and “real-time 
enough” to keep the queues full.

LCNC really should be doing this.  If the Measa cards would maintain a queue 
and a synchronized clock we would not care at all about latency.  Klipper 
proves the idea works.   


> On Apr 16, 2024, at 8:46 PM, gene heskett  wrote:
> 
> On 4/16/24 20:46, Chris Albertson wrote:
>>> On Apr 15, 2024, at 7:20 PM, gene heskett  wrote:
>>> 
>>> On 4/15/24 21:16, Gregg Eshelman via Emc-users wrote:
>>>> I have a Hictop printer with some issues so I'm gathering parts to update 
>>>> it. Among them is a Big Tree Tech SKR 1.4 board (not the Turbo, which is 
>>>> identical except runs the CPU 10Mhz faster).
>>> 
>> I assume you would be running Kiper on it.  If so, then what isn’t it 
>> compatible it?  It rebuillding a printer and you get to choose a display, 
>> then by all means use an HDMI touch screen.  You can get them for $60 now.  
>> But any SPI or USB screen should also work.Mostly the screen attaches to 
>> the Pi.
>> BTW, the BTT Raspbery Pi3 clone’s price just “fell throught rhe floor”.  
>> They are now under $10 in Aliexpress.   This means you have to to buy two of 
>> them to qualify for free shipping. (iIt is kind of a no-brainers to buy one 
>> %7 Pi3 and pay $$6shipping or to buy two for $7 each and get free shipping.)
>> The BTT Pi3 clone runs Klipper on my printer at about 7% CPU load and 35% 
>> RAM usage.   Spending even $35 for a Pi4 is not going to make the printer 
>> run faster but it would be fun to see the load go from 7% to 4%.  The BTT 
>> clone has a 12 to 24 volt power inpuit using screw terminals, but the word 
>> is “no higher than 12 volts if you want the Pi to live a long life.”
>> I have a BTT Pi3 clone and SKR combo powering my Voron V0.2 “printer for 
>> ants” and I’m making very nice prints in ABS at up to 340 mm/second.   At 
>> 260c nozzel and 105C bed.   The combo is not powerfull enough to drive my 80 
>> LED neopixel strip with crazy animations (80 RGBW LEDs updated at 24 frames 
>> per second)  while printing.  I have to tame it down while printing.   That 
>> is the only limitation I’ve found with the SKR.   But do you really need a 
>> moving rainbow on the printer while printing?   I turned on the effect after 
>> the print finished to show it is done.
> The advantage of fluidd/moonraker/klipper is that all data is presented on a 
> web page driven by ngnix, I can run the printer, controlling any detail, from 
> any machine on my home net. I have no idea what the load on the bananapi-m5 
> is but its plenty fast enough for at least 10x an Ender5+'s original speed of 
> around 30mm/sec. I expect my speed limit will be the hot end, its a trinity 
> labs with a 50 watt cylindrical heater and one piece nozzle/heat break.
> 
>> ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
> 
> Cheers, Gene Heskett, CET.
> -- 
> "There are four boxes to be used in defense of liberty:
> soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author, 1940)
> If we desire respect for the law, we must first make the law respectable.
> - Louis D. Brandeis
> 
> 
> 
> ___
> 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] Big Tree Tech Re: Mesa Card Stepgens?

[Chris Albertson]

> On Apr 15, 2024, at 7:20 PM, gene heskett  wrote:
> 
> On 4/15/24 21:16, Gregg Eshelman via Emc-users wrote:
>> I have a Hictop printer with some issues so I'm gathering parts to update 
>> it. Among them is a Big Tree Tech SKR 1.4 board (not the Turbo, which is 
>> identical except runs the CPU 10Mhz faster).
> 

I assume you would be running Kiper on it.  If so, then what isn’t it 
compatible it?  It rebuillding a printer and you get to choose a display, then 
by all means use an HDMI touch screen.  You can get them for $60 now.  But any 
SPI or USB screen should also work.Mostly the screen attaches to the Pi.

BTW, the BTT Raspbery Pi3 clone’s price just “fell throught rhe floor”.  They 
are now under $10 in Aliexpress.   This means you have to to buy two of them to 
qualify for free shipping. (iIt is kind of a no-brainers to buy one %7 Pi3 and 
pay $$6shipping or to buy two for $7 each and get free shipping.)

The BTT Pi3 clone runs Klipper on my printer at about 7% CPU load and 35% RAM 
usage.   Spending even $35 for a Pi4 is not going to make the printer run 
faster but it would be fun to see the load go from 7% to 4%.  The BTT clone has 
a 12 to 24 volt power inpuit using screw terminals, but the word is “no higher 
than 12 volts if you want the Pi to live a long life.”

I have a BTT Pi3 clone and SKR combo powering my Voron V0.2 “printer for ants” 
and I’m making very nice prints in ABS at up to 340 mm/second.   At 260c nozzel 
and 105C bed.   The combo is not powerfull enough to drive my 80 LED neopixel 
strip with crazy animations (80 RGBW LEDs updated at 24 frames per second)  
while printing.  I have to tame it down while printing.   That is the only 
limitation I’ve found with the SKR.   But do you really need a moving rainbow 
on the printer while printing?   I turned on the effect after the print 
finished to show it is done.



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

Re: [Emc-users] Carving a spiral

[Chris Albertson]

> All very true and well for someone equipt with the income and mental gear to 
> use that chain of tools profitably. But I'm an old Iowa farm kid, we made 
> what we needed.  The "store" was 15 miles of horse drawn wagon over a mud 
> road the county graded about 2x a year and all of a days ride in a wagon 
> away. So we grew it, or made it from the woodyard, whatever. 2 miles to the 1 
> room school, I rode an old gentle mare the first mile but had to walk the 2nd 
> mile because there wasn't a barn for the mare during the day any closer to 
> the school when the weather was bad. Grandpa across the road had electricity, 
> a 32 volt delco wet glass batteries, charged by a zenith windcharger. The 
> prop broke, so mother who was the only girl in the 1929 class on aviation 
> technology at Des Moines Tech Hi School, proceeded to teach her father how to 
> carve the wing chord in a new prop. Worked well in less wind than the one we 
> could get from Chicago.  That led to grandpa having the first electric 
> washing machine in Madison County Ia when the Maytag hit & miss tried to 
> start backwards, broke the starter gears and grandma's ankle. A wagon load of 
> shelled corn went to town, and was replaced by an electric motor and enough 
> heavy wire to convert the Maytag. I still wear scars on one hand from getting 
> it caught in the wringer when I was 5. We did not want for anything, we "made 
> do"  That is a hard habit to outgrow.

But today you own a computer, lots of CNC equipment, a 3D printer and education 
is free and just a mouse click away.  None of the stuff I wrote about costs 
even one dollar.   I’m the old ririred guy now.  Fusion360 is free to use.  I 
can print ther prats and then if. Needed sand the same design to CNC machine or 
to an injection molder 

I think you are right about relativity, Einstein very much admired James Clerk 
Maxwell.  Someone said Einstein ”stood on the shoulders of Newton”.   Einstein 
corrected him and said “I stood on Maxwell’s shoulders”.   

Thanks for the story.   I always like to hear those “when I was a kid…” 
stories.   My four grandparents were born in 1902 through 1911 they could talk 
about the days before radio broadcasting and one-room schoolhouses.  One 
grandfather was a professional boxer in the 1920s and traveled a lot.   But 
even more interesting to me, my wife’s parents and uncles were born in pre-war 
Japan.   I think they lived through more change than any living American.  
Sadly the last of them is in very poor health.  My wife is visting her mom in 
Tokyo right now.   

Maybe when I am older I will talk about the days of manually driven gas cars.  
> 
>>  
>>> 
>>> My electronics education is 100% self taught. My mother gave me an IQ good 
>>> enough to pass the CET test w/o cracking a text to study it. I understand 
>>> the physics of it including Relativity. Electronics and Relativity go hand 
>>> in hand, cannot be separated.
>>> 
>>> On Apr 10, 2024, at 12:09 AM, gene heskett  wrote:
>>> 
>>> On 4/10/24 01:57, John Dammeyer wrote:
 A friend and I have been discussing exactly how to write the G-Code to
 create a spiral scroll.
 His rotary table 90:1 reduction with a 1600 micro-step motor could be 
 set up
 to move N steps for each step of the X axis to create the spiral.  But 
 that
 approach seems clumsy.
 Say I wanted to cut a scroll with a 6mm pitch using a 3mm cutter.
  Without using G2 or G3 it's really just a triangle isn't it?  Move 
 rotary
 table distance A and move X axis distance A'.  Do it in small enough
 increments and you get a spiral.  But I feel like I'm missing something
 really simple.
>> Do you need a rotary table to cut a spiral?   It is just a series of 
>> locations in (x,y).  OK, if you wanted to use only (say) the X and A 
>> axis then you should use polar coordinates, not cartesian. The equation 
>> of a spiral on polar coordinates is very simple.  Then you evaluate itat 
>> many thousands of points and at each point write gcode to “cut to” that 
>> point.You would not need the rotary table.
>> Also why think in micro-steps and worm gear rates, you are using LCNC to 
>> do the kinematics, Use millimeters.
>> I think this problem shows that in some cases you really can not write 
>> the gcode by hand.  FOr continous curves in (x,y) there might be 100,000 
>> or more lines of code in the file, especially if you don’t do the cut in 
>> one pass.  You would nee towrite software to generate the g-code.   Or 
>> use existing software, a lot of CAD systems will do this for you
>>> First, a 90/1 is quite high. I have two rotary's, both consisting of a 
>>> 3NM 3phase stepper/servo I made by combining the 3NM motor with a 5/1 
>>> worm. Using a screw in the worms output hub as a single prox sensor 
>>> index pulse generator. To 

Re: [Emc-users] Carving a spiral

[Chris Albertson]

> On Apr 11, 2024, at 11:49 AM, John Dammeyer  wrote:
> 
> Where I'm still having trouble is understanding the compromises.
> 
> Using G02 I,J motion it's possible to specify a feed rate and spindle RPM 
> that remains constant regardless of the spiral diameter.  This approach 
> generated by the CAM software does a number of segments.  But since a spiral 
> is continuously changing the radius doesn't this approach still create a 
> blocky segmented series of lines?  Shouldn't there really be more like a new 
> line for every say one tenth of a degree of rotation?

What you should care about is just one thing: that the resulting spiral is 
within tolerance.   If it is done with a series of straight lines or a series 
of arcs should not matter.

I think the example g-code below is not real.  What I mean is it is not what 
you would use because it does the cut all at once

> %
> N1 G17 G21 G40 G90
> (2 1/2 Axis Profiling)
> N2 T1 M06
> N3 S2000 M3
> N4 G0 Z11.35
> N5 X-6.795 Y-2.3642
> N6 G1 Z5. F745.1
> N7 X-4.971 Y3.7182 F372.5
> N8 G17
> N9 G02X2.9354Y7.9767I6.0824J-1.824 F279.4
> N10 X0.0086Y-11.1633I-3.0663J-9.3249 F372.5
> N11 X-4.8878Y12.9255I-0.3348J12.4739
> N12 X10.8896Y-12.3617I5.6431J-14.0447
> N13 X-16.8958Y8.9702I-11.9852J13.1505
> N14 X21.6077Y-2.7715I18.1973J-9.3286
> N15 X-23.7802Y-5.6409I-22.9524J2.6529
> N16 X22.4641Y15.1503I24.9984J6.2225
> N17 X-18.2168Y-23.5731I-23.4829J-16.0603
> N18 X12.0427Y30.2156I18.8718J24.7892
> N19 X-3.5415Y-35.0813I-12.341J-31.5628
> N20 X-6.7258Y37.3924I3.4567J36.4587
> N21 X17.9557Y-36.5564I7.1836J-38.6957
> N22 X-29.1519Y32.2355I-18.7497J37.6857
> N23 X39.2113Y-24.4201I30.2257J-33.1012
> N24 X-47.7931Y-8.2457I-40.4241J24.6445
> N25 G1 X-49.1271 Y-2.0374 F745.1
> N26 G0 Z11.35
> N27 M30
> %
> 
> On the other hand with the rotary table the motion is continuous for both 
> axis.
> G00 G49 G40.1 G17 G80 G50 G90
> G20
> (2 1/2 Axis Profiling)
> M6 T2
> M03 S2000
> G1 F100 A3600 X60
> G00 Z0.2362
> M5 M9
> M30
> 
> However the F100 is a joke because my A axis can't run more than 7 ipm or so. 
>  I have faint memories of that from a previous discussion on how to set up 
> the A axis.  My INI file says
> DEFAULT_ANGULAR_VELOCITY = 90.00
> MIN_ANGULAR_VELOCITY = 0
> MAX_ANGULAR_VELOCITY = 180.00
> 
> I think that's in degrees per second because at a G01 F6 A3600 it takes 2 
> seconds to do one revolution.  Now that's still plenty fast if we are 
> slotting with a 3mm or 0.118" end mill.  However the problem of SFM is now in 
> the picture.   Again, a custom G-Code solution would likely be to do 
> something like the following.
> 
> G01 F6 A360 X10
> F5.5 A720 X20
> F5  A1080 X30
> 
> And so on...  Picking both X resolution and A distance to create a closer to 
> ideal SFM.  I don't know if there is a G-Code that states the F rate for the 
> A axis is in degrees per minute so it's some sort of conversion from IPM.
> 
> I only have the 4th Axis version of MECSOFT so I can't generate a fifth axis 
> program cutting a spiral with the mill using the rotary table.
> 
> But I think I agree the best way to do this is probably still by hand or with 
> a support program that generate the G-code based on parameter input like 
> min/max SFM.
> 
> John
> 
> 
> 
> ___
> 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] Carving a spiral

[Chris Albertson]

> On Apr 11, 2024, at 9:05 AM, gene heskett  wrote:
> 
> On 4/11/24 11:38, Chris Albertson wrote:
>>> On Apr 10, 2024, at 1:14 PM, gene heskett  wrote:
>>> 
>>> On 4/10/24 13:30, Chris Albertson wrote:
>>>> Do you even need a rotory table to cut a spiral?
>>> 
>>> Now that you ask, I think the answer is probably no. Take a look at the 
>>> nurbs command G5.2 where a group of points describes the curve. Then rotate 
>>> the points about one end in polar/rectangular space.  I haven't played with 
>>> that myself, perhaps Andy has a better way?
>> If you know how to find the 3 or 5 or 10 points that define the curve, then 
>> you can use the same method to find 10,000 point on that curve and just do a 
>> straight line between them.
>> Finally to TEST your idea, you need a way to calculate the error.  You find 
>> the exact solution for any arbitary point and then to compare that to what 
>> you are doing.   If you know how to calulate the exact solution, then why 
>> bother with approxomations like nurbs?  The nuirbs thing is for filling in 
>> points you have no way of knowing.
>> What nurbs is good for is when you want to mill a copy of some object that 
>> you don’t have an exact solution for.  The perfect example is that you were 
>> given an STL and all they have is triangle vertici.   You can’t know where 
>> the STL got its shape, is the curve a circle, elipse or an artest’s pen 
>> stroke.   But the nurbs-like or spline-like fitting funtion will generate a 
>> smooth curve through the list of points.
> All of the above is my weak point. I only got to a couple months of my 
> freshman year in high school, and the math teacher fancied himself as a 
> standup commedian, far more interested in what was in the girls panties that 
> in teaching math.


The best way to catch up on math is Kahn Academy.  The instructor who makes the 
videos is a very good teacher.  He covers from kindergarten (literally) to 
lower division university level, although the more advanced math is taught is a 
slightly less rigorous level than at most universities.  what is a Good Thing 
if you are an engineer or scientist who cares more about practice than theory

I did study the full program of undergrad math but that was. in the 1970s and 
80s.   I had forgotten most Linear Algebra and quite a lot of Calculus.   Then 
I got interested in robots and for that, you pretty much need what I had 
forgotten.   Kahn Academy is ideal for as a refresh class.   But you have to 
invest some time.  He covers a full semester of material in each course.

Electronics is so very much dependent on math that many would-be electrical 
engineers switch majors to civil or environmental engineering because the math 
is easier

Back to g-codes.It should not be a surprise that the g-code for complex 
shapes is either volumes or very complex.   Just look at the 3D printer.  No 
one would attempt to hand code, even a simple printed part can have 200,000 
lines of code.  So we use software to generate the code and it takes usually 
less than two seconds.

Cutting a scroll on a CNC mill is a really hard problem if you care to get it 
correct.   That means (1) the result is within tolerance, (2)  cutting at 
optimal feed rate, (3) roughing and finish passes with tool changes.   The way 
I’d do this is to model it in Fusion360 and then send the file to my 3D printer 
to make a working prototype.   After seeing that the plastic part is correct 
and fits the rest of the machine I’m making then I’d send the same Fusion360 
file to CAM software to create g-code for milling.   Becuse the same design 
file feeds both the printer and the CNC mill, I can have some confidence that 
the plastic part will match the metal part.



 
> 
> My electronics education is 100% self taught. My mother gave me an IQ good 
> enough to pass the CET test w/o cracking a text to study it. I understand the 
> physics of it including Relativity. Electronics and Relativity go hand in 
> hand, cannot be separated.
> 
>>>>> On Apr 10, 2024, at 12:09 AM, gene heskett  wrote:
>>>>> 
>>>>> On 4/10/24 01:57, John Dammeyer wrote:
>>>>>> A friend and I have been discussing exactly how to write the G-Code to
>>>>>> create a spiral scroll.
>>>>>> His rotary table 90:1 reduction with a 1600 micro-step motor could be 
>>>>>> set up
>>>>>> to move N steps for each step of the X axis to create the spiral.  But 
>>>>>> that
>>>>>> approach seems clumsy.
>>>>>> Say I wanted to cut a scroll with a 6mm pitch using a 3mm cutter.
>>>>>>  Without using G2 or G3 it's really just a triangle isn't 

Re: [Emc-users] Carving a spiral

[Chris Albertson]

> On Apr 10, 2024, at 12:47 PM, gene heskett  wrote:
> 
> On 4/10/24 13:03, John Dammeyer wrote:
>> Hi Gene,
>> Just like I can't get my head around Fusion360 or similar CAD.


I find that many people’s problem is that they are still thinking that CAD is 
used to “draw” a part and that somehow the final product is a “‘drawing”.   
This is completely contrary to how it works.

What you do is make the part, just like you would in the shop.  There is no 
“drawing”.   In the shop you might start with an extrusion then cut it and mill 
a pocket and then make some holes and then what ever until it was the right 
shape.   With CAD now, the first step is usually drawing the profile for a 
custom extrusion.   

If you try and hand code gcode or use OpenSCAD you are limited to very simple 
shapes and can never exploit the full capability of the milling machine.   Just 
TRY and hand code a typical hand-held power tool or even a triangular pocket on 
the side of the sphere in hand-written gcode. 

So, if you want to model a part in CAD, just think how you would make it in a 
very well equipped shop.

>> AlibreCAD has gone downhill since they trashed their relationship with
>> MecSoft which correspondingly trashed AlibreCAM.
>> I'll take a closer look at your approach later today.
> 
> My approaches subroutine will take a loop as the primary slicer because each 
> piece of the scroll will need to start at the root of that slice but the 
> curved pie slice is wider at the outside.

A spiral is a continuous curve, already we jhave assen the curve’s equation.   
You just marchthrough the point the equation producse and it is literally 
“perfect”, no approximation.   The only problems left are
1) The method is “too perfect” and will generate to make points and be slow
2) doing it all in one pass is a poor machining technique.




> The target is to carve each slice from its apex point but widen the rim so 
> that your 3mm wide fin is all that's left, then advance to the apex of the 
> next pie slice. So the subroutine will start at 6 common points around the 
> central axis, and you wind up with 6 3mm wide fins in the scroll shape.
> 
> That won't be the most efficient shape but will get you started. To improve 
> the efficiency, that central subroutine will need to be converted to use G4 
> nurbs to describe the spiral path. But that's for next week.  And I expect 
> Andy might have some better ideas on the nurbs points.  I have a good idea 
> what they can do but zip about the math involved. Flow thru a venturi etc 
> type math.
> 
>> Thanks
>> John
>>> -Original Message-
>>> From: gene heskett [mailto:ghesk...@shentel.net]
>>> Sent: April 10, 2024 3:42 AM
>>> To: emc-users@lists.sourceforge.net
>>> Subject: Re: [Emc-users] Carving a spiral
>>> 
>>> On 4/10/24 03:37, John Dammeyer wrote:
 Hi Marcus,
 
 Here's the problem.
 
 My Alibre CAD/Cam can produce a spiral slot in a disk using X and Y
>> motion.
>>> 
>>> I looked at alibre, could not get my Iowa farm kid head around it. Far
>>> easier for me to write my own gcode.
>>> 
>>> In linuxcnc, you can trade the axis names around to fit your hdwe.
>>> 
>>> How far can you tilt your Z? Mine can do a full 90, aka lay the spindle
>>> horizontal.  Either way IIRC. I can then use X as X, A as A, and a
>>> single straight line move to carve the spiral using Z while A is turning
>>> N degrees to carve the spiral. So the gcode then becomes a subroutine to
>>> do that, and a 2nd loop routine to handle the start of the spiral and
>>> possibly a master outer loop to do any incremental cuts to get to the
>>> depth needed. Maybe 80 LOC total.
>>> 
>>> If your head cannot tilt that far, then you''l have to cobble up a C,
>>> facing up which I CAN do but its a 90/1 drive and will restrict the
>>> speed as It can't turn fast enough. Also has a std stepper motor, push
>>> its speed and it stalls. Someday I'll put a good motor on it.
>>> 
>>> A Warning though, most of the combo gizmos they sell for $300 or so on
>>> ebay are belt drive and no-where near strong enough for this. I did use
>>> my 90/1 as A when making my tap hats. Used it to drill & tap the 4 grub
>>> screw holes. I setup workstations on the length of the go704's table,
>>> put a piece of brass rod in the spindle, drilled the hole for the tap,
>>> move the brass to a clamp, drilled and tapped for a locking to an r8
>>> collet screw hole, moved the brass to the A chuck and drilled and tapped
>>> all 4 grub screws. All in the same gcode file with pauses and automatic
>>> tlo offset corrections as the drills were different lengths.  Made a
>>> regular production line out of it, took longer moving the brass around
>>> than the total run time for the machine.
>>> 
>>> 
 
 
 If I tell it to use my 4th axis it's like the video you posted.
>> Designed
 for creating a spiral on something horizontal to say the X axis.
 
 I think I'd have to buy the 5th axis capability in order to be 

Re: [Emc-users] Carving a spiral

[Chris Albertson]

> On Apr 10, 2024, at 1:14 PM, gene heskett  wrote:
> 
> On 4/10/24 13:30, Chris Albertson wrote:
>> Do you even need a rotory table to cut a spiral?
> 
> Now that you ask, I think the answer is probably no. Take a look at the nurbs 
> command G5.2 where a group of points describes the curve. Then rotate the 
> points about one end in polar/rectangular space.  I haven't played with that 
> myself, perhaps Andy has a better way?

If you know how to find the 3 or 5 or 10 points that define the curve, then you 
can use the same method to find 10,000 point on that curve and just do a 
straight line between them.

Finally to TEST your idea, you need a way to calculate the error.  You find the 
exact solution for any arbitary point and then to compare that to what you are 
doing.   If you know how to calulate the exact solution, then why bother with 
approxomations like nurbs?  The nuirbs thing is for filling in points you have 
no way of knowing.

What nurbs is good for is when you want to mill a copy of some object that you 
don’t have an exact solution for.  The perfect example is that you were given 
an STL and all they have is triangle vertici.   You can’t know where the STL 
got its shape, is the curve a circle, elipse or an artest’s pen stroke.   But 
the nurbs-like or spline-like fitting funtion will generate a smooth curve 
through the list of points. 


>>> On Apr 10, 2024, at 12:09 AM, gene heskett  wrote:
>>> 
>>> On 4/10/24 01:57, John Dammeyer wrote:
>>>> A friend and I have been discussing exactly how to write the G-Code to
>>>> create a spiral scroll.
>>>> His rotary table 90:1 reduction with a 1600 micro-step motor could be set 
>>>> up
>>>> to move N steps for each step of the X axis to create the spiral.  But that
>>>> approach seems clumsy.
>>>> Say I wanted to cut a scroll with a 6mm pitch using a 3mm cutter.
>>>>  Without using G2 or G3 it's really just a triangle isn't it?  Move rotary
>>>> table distance A and move X axis distance A'.  Do it in small enough
>>>> increments and you get a spiral.  But I feel like I'm missing something
>>>> really simple.
>> Do you need a rotary table to cut a spiral?   It is just a series of 
>> locations in (x,y).  OK, if you wanted to use only (say) the X and A axis 
>> then you should use polar coordinates, not cartesian. The equation of a 
>> spiral on polar coordinates is very simple.  Then you evaluate itat many 
>> thousands of points and at each point write gcode to “cut to” that point.
>> You would not need the rotary table.
>> Also why think in micro-steps and worm gear rates, you are using LCNC to do 
>> the kinematics, Use millimeters.
>> I think this problem shows that in some cases you really can not write the 
>> gcode by hand.  FOr continous curves in (x,y) there might be 100,000 or more 
>> lines of code in the file, especially if you don’t do the cut in one pass.  
>> You would nee towrite software to generate the g-code.   Or use existing 
>> software, a lot of CAD systems will do this for you
>>> First, a 90/1 is quite high. I have two rotary's, both consisting of a 3NM 
>>> 3phase stepper/servo I made by combining the 3NM motor with a 5/1 worm. 
>>> Using a screw in the worms output hub as a single prox sensor index pulse 
>>> generator. To calibrate a complete rev, I measure the steps by starting the 
>>> count on the 3rd turn ans stopping the count on the 103rd turn, which gives 
>>> me a scale*100.  Shift the decimal point 2 places left this becomes the 
>>> scale for the axis in the .ini file.  All this math in linuxcnc is floating 
>>> point so I can ask it for 33.333 degrees and it will run to what it thinks 
>>> is 33.333 degrees. This stepscale:
>>> STEPSCALE   = 22.222 = 1 degree
>>> So one count is about 1/22.222 degrees, probably less than the 
>>> backlash in the rvs39 worm, a pretty cheap worm.
>>> 
>>> Currently to make one of my maple vise screws, starting at 0 degrees its 
>>> around 60,000 degrees it turns for around 400 mm of screw that y travels. 
>>> Then I lift the tool, turn it another 180 degrees, re lower the tool and 
>>> bring y back to zero and b=180. Makes a perfect two start buttress thread. 
>>> The B is turning, in perfect sync with the Y motion, at something in the 
>>> 300 to 400 rpm range. That 3NM motor is heating but not dangerously so.
>>> 
>>> There is no reason you couldn't lay it down to make a C drive, and 
>>> simultaneously drive X Z & C to carve an impeller in a quite serviceable 
>

Re: [Emc-users] Carving a spiral

[Chris Albertson]

> On Apr 10, 2024, at 10:44 AM, Ralph Stirling via Emc-users 
>  wrote:
> 
> Use the parametric equation of a spiral and compute it *inside* your
> g-code program.
> 
> x(t) = c1 * t * cos(t) + x0
> y(t) = c2 * t * sin(t) + y0


COnceptually that is it.  But do you really want to cut afull depth 3mm slot 
with a 3mm tool in one pass?  

There is another problem.  If you make equal increments of “t” the step size 
increases over time.  The steps are farr too small near the center so you waste 
time cutting too slowly.   You want to move in steps such that the desired arc 
and the straight line approximation is within some tolerance.   This is a hard 
problem.   I tried to solve it myself years ago.  My boss suggested a 
sollution:  Advance ’t’ is very tiny steps and co,put the error, if the error 
is small advance it advance it again, if not drop a point.  If you do this each 
point is just inside the tolerance.   It was slow.   I decided to make bigger 
jumps and if i overstepped to backup.   

A good solution is not something you can program in g-cose and every get it 
right.

Today,  any reasonable engineer would simply draw the spiral in CAD and then 
click “generate path” and the CAD system would output the gcode and it would 
handle the roughing and finishing cuts and the tool changes and handle the 
spindle speed and cutting rates.   No math or thinking is required.

But as a challenge or a student exercise, writing an OPTIMAL solution in pure 
g-code that cuts the metal at the best rate while keeping tolerances is an 
interesting problem, not simple at all.  Most people would let their CAM system 
figure this out for them.



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

Re: [Emc-users] Carving a spiral

[Chris Albertson]

Do you even need a rotory table to cut a spiral?

> On Apr 10, 2024, at 12:09 AM, gene heskett  wrote:
> 
> On 4/10/24 01:57, John Dammeyer wrote:
>> A friend and I have been discussing exactly how to write the G-Code to
>> create a spiral scroll.
>> His rotary table 90:1 reduction with a 1600 micro-step motor could be set up
>> to move N steps for each step of the X axis to create the spiral.  But that
>> approach seems clumsy.
>> Say I wanted to cut a scroll with a 6mm pitch using a 3mm cutter.
>>  Without using G2 or G3 it's really just a triangle isn't it?  Move rotary
>> table distance A and move X axis distance A'.  Do it in small enough
>> increments and you get a spiral.  But I feel like I'm missing something
>> really simple.

Do you need a rotary table to cut a spiral?   It is just a series of locations 
in (x,y).  OK, if you wanted to use only (say) the X and A axis then you should 
use polar coordinates, not cartesian. The equation of a spiral on polar 
coordinates is very simple.  Then you evaluate itat many thousands of points 
and at each point write gcode to “cut to” that point.You would not need the 
rotary table.

Also why think in micro-steps and worm gear rates, you are using LCNC to do the 
kinematics, Use millimeters.

I think this problem shows that in some cases you really can not write the 
gcode by hand.  FOr continous curves in (x,y) there might be 100,000 or more 
lines of code in the file, especially if you don’t do the cut in one pass.  You 
would nee towrite software to generate the g-code.   Or use existing software, 
a lot of CAD systems will do this for you


> First, a 90/1 is quite high. I have two rotary's, both consisting of a 3NM 
> 3phase stepper/servo I made by combining the 3NM motor with a 5/1 worm. Using 
> a screw in the worms output hub as a single prox sensor index pulse 
> generator. To calibrate a complete rev, I measure the steps by starting the 
> count on the 3rd turn ans stopping the count on the 103rd turn, which gives 
> me a scale*100.  Shift the decimal point 2 places left this becomes the scale 
> for the axis in the .ini file.  All this math in linuxcnc is floating point 
> so I can ask it for 33.333 degrees and it will run to what it thinks is 
> 33.333 degrees. This stepscale:
> STEPSCALE   = 22.222 = 1 degree
> So one count is about 1/22.222 degrees, probably less than the 
> backlash in the rvs39 worm, a pretty cheap worm.
> 
> Currently to make one of my maple vise screws, starting at 0 degrees its 
> around 60,000 degrees it turns for around 400 mm of screw that y travels. 
> Then I lift the tool, turn it another 180 degrees, re lower the tool and 
> bring y back to zero and b=180. Makes a perfect two start buttress thread. 
> The B is turning, in perfect sync with the Y motion, at something in the 300 
> to 400 rpm range. That 3NM motor is heating but not dangerously so.
> 
> There is no reason you couldn't lay it down to make a C drive, and 
> simultaneously drive X Z & C to carve an impeller in a quite serviceable 
> scroll.
> 
> The versatility of the closed loop stepper/servo, which does EXACTLY what the 
> TP tells it to do, without a PID in the path, is amazing. I have them rigged 
> to e-stop linuxcnc in about a millisecond if they make an error, like losing 
> a step. Tested till the cows come home, has yet to happen working a job. I 
> haven't hobbed any gears, but it certainly seems accurate enough to do it.
> 
>> Suggestions?
>> Thanks
>> John
>>  ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>> .
> 
> Cheers, Gene Heskett, CET.
> -- 
> "There are four boxes to be used in defense of liberty:
> soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author, 1940)
> If we desire respect for the law, we must first make the law respectable.
> - Louis D. Brandeis
> 
> 
> 
> ___
> 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] Raspberry Pi5

[Chris Albertson]

The Pi5 has the advantage over a Pi4 of not only a faster CPU but the real step 
up over the Pi4 is that you can install the file system on a M2 SSD and not 
have to run off the SD card.

But if you look at the price, you get more for the same money if you buy a 
low-end Intel PC from a place like Newegg.  They offer a refurbished Intel I5 
with 4gb RAM and SSD for $95.This might seem like it costs $10 more than 
the Pi5 but the PC has a power supply, case, and cooling fan and the SSD is 
included too.So it is a bit cheaper than the Pi5.

I think the Raspberry Pi is very good for my robotics projects that have to run 
off battery power and have very limited internal space for the control 
electronics.   But if you are building a machine that runs off AC mains power 
and is not portable, why bother with a Raspberry Pi?

The Pi4 or even Pi3 might be OK for driving the milling machine while it is 
cutting metal but you do other things like software updates, compiles, and 
visualizations. 




> On Apr 9, 2024, at 9:45 AM, John Dammeyer  wrote:
> 
> Hi all,
> 
> Will the Pi5  be any better for LinuxCNC than the Pi4 with a MESA Ethernet
> Controller or is the Pi4 already more than what LinuxCNC requires?
> 
> John
> 
> 
> ___
> 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] black list aliexpress

[Chris Albertson]

I remember once working at a company where we bought a fairly high-end PC.   We 
found we could not export it without a license.  But guess what?  It was made 
in China.   Some rules are pointless.


> On Mar 18, 2024, at 7:33 PM, gene heskett  wrote:
> 
> On 3/18/24 18:50, Jon Elson wrote:
>> On 3/18/24 07:53, Todd Zuercher via Emc-users wrote:
>>> Gene,
>>> 
>>> In all honesty I wouldn't mind knowing more about your VFD source you say 
>>> you want black listed.  I'm so sick of trying to track down high speed VFDs 
>>> that I'm almost tempted to pay the machine manufacturer's ridiculous 3X 
>>> normal retail list price markup the next time we need a new one.   We 
>>> occasionally have needed a replacement VFD capable of running about a 15kw 
>>> 4pole router spindle at up to 24krpm (that is 800hz output, most drives 
>>> seem to max out at or are limited to 500-600hz output.)  The last time we 
>>> tried to order a new drive with th

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

Re: [Emc-users] looking for advice --> trick to sell an extra toolsetter

[Chris Albertson]

Silicon carbide is classified as a “semi-conductor”.It is used to make some 
heavy duty rectifier diodes and maybe other stuff

Oxide-coated tools likely would conduct because the oxide wears off really 
fast, 

OK, I just tried a black oxide tool and I see ZERO conductivity.  But then I 
test the cutting sharp edge and it conducts where the black stuff has been worn 
off.





> On Mar 16, 2024, at 2:23 AM, gene heskett  wrote:
> 
> On 3/15/24 14:05, Nicklas SB Karlsson wrote:
>>> ...
>>> Is the coating on the bit electrically conductive?  I know some oxides 
>>> don’t conduct electricity.
>> Then sales person first sell you a tool setter that work only condictive 
>> mills and drills. Second time he come back and
>> sell you one these non coductive bits. Then of course you need to buy a new 
>> tool setter and he could sell this to.
>> Nicklas Karlsson
> Interesting that no one has named a non-conducting tool. That is something 
> I've not yet seen.
> 
> Does this boogieman even exist? Produce it by actual name if you have such. 
> I'd like to obtain and test it.
> 
> Take care & stay well Nicklas K.
> 
> Cheers, Gene Heskett, CET.
> -- 
> "There are four boxes to be used in defense of liberty:
> soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author, 1940)
> If we desire respect for the law, we must first make the law respectable.
> - Louis D. Brandeis
> 
> 
> 
> ___
> 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] looking for advice

[Chris Albertson]

If you are using it as a probe, why not put the drill in the chuck backward, 
then you have a precision rod, not a drill.

Spinning seems like an interesting idea, moving fraction is always less than 
static friction.  You would likely need to spin a 3mm rod fast enough so that 
the tangential speed is comparable to the up and down speed.   Let's see,… 
spinning a 3mm drill at 60 RPM, one rev per second you have 3 * 3.14 mm.  per 
second tangent speed.  That is something close to 9mm/sec.  that might be too 
fast, I don’t know

Even if used “point down”.  My guess is is that you want the tangent speed 
comparable to the Z-speed.   


I use those very cheap 2, 3 and 4mm drillig for 3D printed plastic to ream the 
holes if they print undersized.  The cheapest Aliexpress drills I can find are 
good enough.  Good that you have found another use for them.   The diameters 
are spot-on, or as good as I can measure.

Is the coating on the bit electrically conductive?  I know some oxides don’t 
conduct electricity. 



> On Mar 15, 2024, at 7:52 AM, gene heskett  wrote:
> 
> I have obtained a 10 pack of uxcel 3mm mills. Look more like a rasp than a 
> normal spiral mill.  2 edges on the plunge face though.
> 
> They will be used as both the probe signal generator to find the 3mm holes 
> for bolts in a 9mm wide linear bearing rail.  That means they will have maybe 
> .2mm total clearance when inserted into the 3mm hole in the rail to reach the 
> bottom of the hole. When functioning as the probe contact I'm spinning them 
> in reverse, so if they do touch, tripping the probe, they should skip w/o 
> cutting because they are turning backwards. I intend to do the usual 4 corner 
> search when locating the hole, but once found, I'll g38.2 Z deep enough to 
> reach the cf tube, reverse the direction to fwd, and run it down several mm 
> thru the cf and stroke it up and down to clear as much of the cf from the 
> hole as possible.
> 
> What I need is a good idea of an rpm and feed rate that will not wear the 
> edge from too slow.
> 
> What revs and feed rate stands a chance of doing a decent job of making a 
> clean hole w/o excess tool wear? Max revs about 2750/minute.
> 
> Thanks all.
> 
> Cheers, Gene Heskett, CET.
> -- 
> "There are four boxes to be used in defense of liberty:
> soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author, 1940)
> If we desire respect for the law, we must first make the law respectable.
> - Louis D. Brandeis
> 
> 
> ___
> 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] How to treat rusted linear rails?

[Chris Albertson]

Are the rails a standard part, or are they custom?   If a standard part you 
could replace them, they install with just a screwdriver.  These used to be 
very expensive until they started making themin Chine.  Now pretty good rails 
are available at the 20-something dollar price point.   But only if they are 
the standard type.  This might not be the case on an older machine



> On Mar 11, 2024, at 5:11 AM, Roland Jollivet  
> wrote:
> 
> Personally, I would use an electrolytic process with carbon and a power
> source. Any other method would only further damage the rails.
> You're 'plating' onto the rails, so no metal is lost. The rust gets reduced
> back to Fe and can be brushed off.
> 
> Roland
> 
> 
> 
> On Mon, 11 Mar 2024 at 09:47, Viesturs Lācis 
> wrote:
> 
>> Hello!
>> 
>> Last year there was a chance to acquire Biesse machine for a a really
>> low price so I surprised myself with pretty large and heavy Christmas
>> present. Just put it in the workshop (where heating is not yet
>> present) and was waiting for a warmer weather to start retrofitting it
>> with LinuxCNC.
>> 
>> Since this machine has spent at least several years in a shed - roof
>> above it but no other protection against outside temperatures, I did
>> not worry about adding one more winter to it.
>> 
>> I have 2 questions:
>> 1) are there any recommendations for outside temperature when trying
>> to connect it to power? I want to see the status messages on
>> servodrives - hopefully they are good (I have done 4 similar retrofits
>> so I feel familiar with those old drives).
>> 
>> 2) my main problem is the rust on linear rails. what is recommended
>> procedure to treat this?
>> Here is a picture that show the extent of the issue:
>> https://pasteboard.co/C1EAn0w5t8KT.jpg
>> What I did is brushing it with a piece of steel wool moisted with oil.
>> I am not sure that it is sufficient so I would appreciate any tips on
>> how to treat them.
>> 
>> Viesturs
>> 
>> 
>> ___
>> 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] Acceleration question.

[Chris Albertson]

It is very easy to compute the rotational momentum of a complex-shape flywheel. 
  Let’s use a simplified case as an example:  You have a 100 mm diameter steel 
disk with an 80 mm hole.  This looks like a ring made of 10 mm thick metal.

First compute the momentum as if there was no hole, for a solid 100 mm disk,
Next compute the same for an 80 mm disk. Subtract the 80mm disk momentum 
from the 100 mm disk momentum.

If you want to account for the spokes, figure out their “average thickness” as 
if the spokes were replaced by a thin sheet of metal and add that back in. 

There is a more complex way to do this but it requires Calculus.   I think they 
showed us the hard way just so that they could come back and show us students 
that you could decompose any complex wheel into a set of simple disks and then 
add and subtract them.

+

Stepper motors CAN work.   But not if you use the simple Step/Direction 
interface most drivers offer.   The step/dir convention is NOT a function of 
the motor.  It is a function of the motor driver.  The motor itself as A+/A-, 
B+/B- leads and takes analog voltages.  It is nothing more than a two-phase 
BLDC motor with many pole pairs.   You can drive a stepper in “continuous and 
smooth non-stepping” mode if you like, if you get a smarter driver that can 
continuously vary the input voltages.  The more sophisticated controller can 
driver the motor in “torque mode” so that it supplies a specified torque.   





> On Feb 23, 2024, at 10:05 AM, gene heskett  wrote:
> 
> On 6/16/22 21:57, John Dammeyer wrote:
>> Hi Gene,
>> Quite right.  That link I posted used a table while in fact, as you pointed 
>> out, the mass is mostly on the outside of a flywheel with spokes.  I would 
>> imagine at there is some average where if it's a 300 lb disk that is 24" 
>> might be the same as a 36" disk that is 400 lbs.
>> Think of a fly press for example with a clutch that engages the tooling.  
>> Even if it does take 5 seconds to get up to speed, the clutch engages, the 
>> tool moves down and punches and moves up and the clutch releases.  Even if 
>> the speed slowed down by 20% when the clutch released then assuming linear 
>> acceleration now only 1 second is required to bring the speed back up.  At 
>> 50 RPM (0.83 seconds per rev) then you could do another punch stroke 1 
>> second later and so possibly run 30 strokes per minute.
>> That jpg chart I included suggests with 100% efficiency and no real friction 
>> that 45 oz-in are required.   Seems very low to me hence the questions.  
>> Even if I did use a stepper motor and went 16:1 to bring the RPM down to 800 
>> RPM the motor could easily be a size 23 300 oz-in.
>> Could that actually bring a flywheel up to that speed in 5 seconds?
> 
> The closest I could come, assuming no frictional losses, would still be just 
> a SWAG. But it sure seems to me a decimal point got moved or left out 
> someplace.
> 
> A stepper would be a poor choice of power unless the stepper drive also 
> started at zero. A stepper unable to stay synced with the incoming step rate 
> has next to zero torque. A vfd makes far more sense as you could set it for 2 
> or 3x the motors FLA and the vfd would then throttle the current, using 
> seriosly more drive currant immediately after a strike to get it back to 
> speed, but the average would still be only maintenance unless it was striking 
> with every revolution. EG 50 strikes a minute. Given the time to extract and 
> replace the next work piece is going to be at least a second, that would be 
> one heck of a busy machine. Much the same could be said of a hirez encoder 
> whose output was compared to the desired speed and a 1 horse treadmill motor 
> being run by one of Jon's pwm-servos. Both solutions would need far less 
> electrical power to get the job done than a steeper could do.
>>> -Original Message-
>>> From: gene heskett [mailto:ghesk...@shentel.net]
>>> Sent: June-16-22 6:34 PM
>>> To: emc-users@lists.sourceforge.net
>>> Subject: Re: [Emc-users] Acceleration question.
>>> 
>>> On 6/16/22 20:54, John Dammeyer wrote:
 OK.  I realize this will be a dumb question but please bear with me 
 especially since I've included the ability
>>> to accelerate in my Electronic Lead Screw project.
 
 A friend and I were discussing bringing a 300 pound flywheel up to speed.
 Vz=0 RPM, Vf=50 RPM.  Reduction drive to the flywheel shaft is 32:1 so 
 final speed of motor is 1600 RPM.
 
 Assume we're happy with 5 seconds to accelerate for Tz to Tf.  Motor 
 voltage is 12V.
 
 We have the mass, we have the velocity, we have the time and motor 
 voltage.  The question is what are
>>> the calculations to determine how much current the motor will require to 
>>> create this acceleration?
>>> Assuming of course the motor is 100% efficient.
 
 We're getting all confused with F=ma and 1/2*a*t^2 etc.
 
 What size motor is actually needed to 

Re: [Emc-users] Axis direction

[Chris Albertson]

It is pretty easy to see WHY the z-axis is set up the way it is on a lath and a 
mill and why it is different.   You need a well defined “zero”.  On a mill, the 
machine's “zero" is the table and one a mill it is the chuck.   A lathe has not 
other well define place on the machine, the tailstock moves.

The after defining the zero point you use the “natural” convention the number 
get bigger if you go to the right or up.

This is the normal way engineers thing about corodrnttes.  you ask thee 
questions IN ORDER
1) where is the origin?
2) which way does “Z” point
3) then apply the right hand rules for X and Y

It’s not just machine tools that do this but everyone from physic research to 
aircraft manufacturing.  And you do have to answer those question in order.

#2 is really arbitrary but from 8th grade algebra on, we are used to drawing 
graphs with the numbers getting bigger as you go up or right and with the zero 
point inthe lower left



> On Feb 6, 2024, at 6:06 AM, gene heskett  wrote:
> 
> On 2/6/24 07:54, Stuart Stevenson wrote:
>> Ray Henry's story is a Murphy's Law occurrence. I once had a car with a
>> starter problem. I changed the starter three or four times in quick
>> succession. I could not figure out why so I purchased two, put one in the
>> trunk along with tools. I never had to use it.
> Chuckle. BTDT...Still do at times. Works better than average. But the way Ray 
> worded it caught me off guard and I literally shook myself laughing for quite 
> a spell. I spent some time in Iron Mountain MI, about a long hour from Ray's 
> place, so I visited a couple times while modifying the tv station there for 
> digital.
> 
> If you ever come across any more of those teeny ball screws like you sold me 
> around 20 years back I've got a place to put them.
> 
> Take care & stay well, Stuart.
> 
> Cheers, Gene Heskett, CET.
> -- 
> "There are four boxes to be used in defense of liberty:
> soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author, 1940)
> If we desire respect for the law, we must first make the law respectable.
> - Louis D. Brandeis
> 
> 
> 
> ___
> 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] Ball Screw Driving Questions

[Chris Albertson]

When it comes to flexing and bending, think about the bed too.  If the servo 
motor and pullies are mounted at the tailstock end, then the bed will carry the 
reaction forces and see exactly the same tension and “twist” force as the 
screw.  But if the motor is mounted at the headstock end the bed sees no net 
forces.

The twist force on the bad will be the motor’s stall torque times the pulley 
reduction.  Would that be enough to warp a cast iron bad?  That depends on the 
detainees.

Also you. want the force of the motor applied to the fixed end near the 
headstock because none of the mounting points will move as force is applied.

In short, think about what bends as force is applied, the motor pushes on the 
screw one way and the motor mounts in an equal but opposite direction.


> On Dec 25, 2023, at 3:37 PM, Linden via Emc-users 
>  wrote:
> 
> Hello All,
> 
> In the early stages of converting a 13 x 24 inch manual Chinese lathe to 
> run with Linux CNC.
> 
> I have 2 questions regarding replacing the Z axis lead screw with a 3205 ball 
> screw:
> 
>  What I am thinking is mounting the fixed end in a pillow block at the head 
> stock end of the lathe and the floating end in a second pillow block at the 
> tail stock end of the bed. The question I have is there any reason I 
> shouldn't  drive the ball screw from the floating end? My logic for driving 
> at the floating (tali stock) end is  1 I have more room for belt reduction at 
> this end and 2  with the fixed end of the ball screw at the head stock end is 
> that the ball screw will be in tension when it is pulling the carriage toward 
> the head stock during cutting and less likely to flex or bend.
> 
> The second question I have is what would be a realistic cutting speed range 
> for the ball screw in RPM?  The servo motor I am using has a top speed of 
> 3000rpm and I am trying to figure out reduction ratio that is realistic.
> 
> Thanks in advance for any help or suggestions.
> 
> 
> Linden
> 
> ___
> 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] Suggested enhancment to linuxcnc

[Chris Albertson]

>> 
>> Depends on the mill or? In the case of my G0704, the accelerometer should go 
>> on the table for xy motion, not the relatively stationary chuck which 
>> generally only moves vertically. In the chuck on something like the 6040, a 
>> medium sized gantry I have.  The acceleromter s/b on the moving part IOW.

Yes, certainly, it’s the table that moves.  I was thinking of some kind of CNC 
router.But I think the idea is the same, yu’d see very quickly if there was 
any kind of error to be corrected out.  Even on the 3D printer, you would 
need an accelerometer on BOTH the print head and the table.

My guess is that massive tables don’t shake much.

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

Re: [Emc-users] Suggested enhancment to linuxcnc

[Chris Albertson]

Machine tools are a little different from 3D printers because
1) Printers are not built nearly are solidly are milling machines.  Milling 
machines do not flex very much
2) the printer head moves in air and there is no resistance to movement.  While 
an end mill has to have force applied to cut metal
3) if anything causes the mill’s spindle to vibrate it is the teeth of the 
rotating end mill, not “spring” in the machine's structure

That said, I had planned to use accelerometers in the feet on my next robot 
project.  Robot legs are very much like 3D printers or milling machines.  Each 
leg has three or more axes and needs to be precisely moved and while moving, 
the foot might unexpectedly collide with some object.  We can compute the 
expected acceleration on the foot and compare to the actual and find the 
difference.  The robot operates in the uncontrolled real-world and I expect 
many “surprises”.  The 3D printer is “bendy" and flexible so I would expect a 
difference between expected and actual.  But the milling machine is so sturdy, 
I’d expect little difference.

But still, it would be a very easy experiment.  Decent accelerometers cost less 
than $20 and you could simply place one in the jacobs chuck and then jog the 
mill around and see what the accelerometer says.   No software changes to LCNC 
would be needed, just put the little PCB in the chuck and look at some plots.  
Here is one that I have 
https://www.amazon.com/HiLetgo-Gyroscope-Acceleration-Accelerator-Magnetometer/dp/B01I1J0Z7Y/ref=sr_1_1_sspa
HiLetgo MPU9250 GY-9250 9-Axis 9 DOF 16 Bit Gyroscope Acceleration Magnetic 
Sensor 9-Axis Attitude +Gyro+Accelerator+Magnetometer Sensor Module IIC/SPI
amazon.com


> On Dec 17, 2023, at 8:52 AM, gene heskett  wrote:
> 
> One of the not so quiet revolutions taking place in the 3d printer landscape 
> over the last couple years is called input shaping, something probably done 
> best in the cards like Peter Wallace makes.
> 
> A simplified explanation is fitting an accelerometer to to tool head, the 
> exciting each axis it turn withe an audio sweep of a small amplitude, 
> sweeping from 10 Hz to about 250 Hz while measuring the resulting movement 
> with an adxl345 recording the data it outputs for each of the axises tested. 
> Since a 3d printer works in slices, the z axis is generally much slower on 
> the 3d printer, so is left out in some versions.
> 
> This data is then run thru a forrier or butterfly transform to develop a 
> compensating acceleration curve that does not effect the overall speed, but 
> does reduce the fine detail or boost it, to essentially cancel the machines 
> natural vibrations. On a 3d printer, given enough heat to supply the head 
> with continuous hot plastic, the use of this compensation has taken the 3d 
> printer from 50 or 60mm/second maximum speeds to 200 mm/second over the last 
> 2 years. I can now buy a printer with this built into its OS for under $1000.
> There is one obstacle, most 3d printers do not have an interpreter that knows 
> about G2/G3 and its ilk, so most slicer's converts those to tiny straight 
> line moves that look like a circle in plastic.
> 
> This is done in the default interpreter, Marlin, but is done better by 
> klipper in the better printers but can be reflashed into 99% of the 
> controller cards out there just like we can do with Peters cards.
> This interpreter runs on the controller cards, often stm32 based cards that 
> sell for, in the ones BTT makes for $59 for a low end octopus card, which can 
> drive up to 8 motors, 5 fans, 3 heaters and all our limit/home switches and 
> probing gizmo's.  Some of these even include the G2/G3 stuffs.  These cards 
> are nearly all designed to handle nema-17 motors at 24 vols and maybe 3 amp 
> max motors, but one line of the octopus family of cards has a separate motor 
> supply input that assumes 60 volt rated drivers so even those tiny motors can 
> be moved at amazing speeds.
> 
> Top that with signal stealing plugins that fit the driver socket of these 
> boards I'm rebuilding 2 bigger printers with nema-17 versions of the closed 
> loop servo/steppers with optical encoders that use drivers like we use with 
> linuxcnc, 2m542 sized stuff, but now rated for 90 volts and up so they can be 
> driven at unreal speeds w/o losing home. The PID is in the driver, linuxcnc 
> just tells them what to do and they do it. And if they can't do it, they will 
> tell linuxcnc, stopping it by linking that signal into the F2 of linuxcnc. 
> Doing the stop quick enough on my Sheldon the I can position a chuck jaw in 
> the way, and jog a carbide chipped tool into that jaw at 20mm a second, it 
> hits the jaw, the driver shuts off the motor drive at the same time it tells 
> linuxcnc to stop, the release of the motor drive lets it spring back away 
> from the jaw by 10 thou or so. The carbide chip is not damaged and the chuck 
> jaw is not marked.  Tested many times, but has never occurred wile 

Re: [Emc-users] Stepping through linuxcnc

[Chris Albertson]

Debuggers don’t work well with real-time code. and there is WAY to much code 
for that. 

Professionally I’ve had to come up to speed on large software projects several 
times.  The best way is to first become an expert at the user level and learn 
what it does.  This applies to radars, camera firmware, and everything else.  
Then you skim-read the code just to see what is there. and take notes on what 
the parts are.  Next study one section at a time.  You generally don’t need to 
know how everything works in great detail, just where to find it if you need it.

Another good way to learn is to write new regression tests, no software ever 
has enough of these.  Writing a test forces you to look in detail and the test 
result proves you understand.

Then, at last, you can make changes, then build and test.  But the first step 
is to be an expert-level user.



> On Dec 10, 2023, at 3:08 PM, alanmthomason--- via Emc-users 
>  wrote:
> 
> Hi There.I am looking for guidance to build linuxcnc in a way that I can
> step through it.  I can build currently, but in trying to understand the
> underlying code I was hoping to use a debugger to step through it.  
> 
> I could not find any flags in the configure -help that looked right.  
> 
> 
> 
> Thanks so much, Alan.
> 
> 
> ___
> 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] OT Fanuc ot-c cpu power supply A16B-1212-0950

[Chris Albertson]

I missed the beginning of this.  What kind a machine is this?

As for the power supply, they are very hard to debug if you don’t have a 
schematic.   But what about replacing it entirely?  

Good motors are always expensive.  But encoders have gotten cheap and easy to 
use.  The last one I bout was an AS5048.  It cost less than $20 and the specs 
are incredible. 

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

Re: [Emc-users] OT Fanuc ot-c cpu power supply A16B-1212-0950

[Chris Albertson]

If it is a simple linear power supply, yes they are very easy to debug and 
repair,  LIkey it is the caps.

But switching power supplies are much harder to debug, You can guess it might 
be the same issue but these have dozens of parts that can fail..  In any case, 
the cost to repair is small.

On the other hand, you might be able to replace the power supply with a PC 
power supply that you already have inside some older “junk” computer.

> On Dec 2, 2023, at 9:58 AM, John Dammeyer  wrote:
> 
> Replace the electrolytic capacitors in the power supply.If you can get
> one of these meters  you can verify this is a good solution by measuring the
> one of the old ones with a new one.
> https://vi.aliexpress.com/item/4001294233271.html
> 
> The 3 phase motor controller on our expensive Bosch Fridge failed after 10
> years.  Replacing the capacitors put the power supply voltages that ran the
> motor control part back to the correct value.  A new capacitor, IIRC, had an
> ESR of about 0.5 ohms.  The old one was 30 Ohms.There was no physical
> visual indication that there was a problem.  Fridge has been running
> perfectly now a year later.
> 
> John
> 
>> -Original Message-
>> From: Leonardo Marsaglia [mailto:ldmarsag...@gmail.com]
>> Sent: December 2, 2023 7:58 AM
>> To: Enhanced Machine Controller (EMC)
>> Subject: [Emc-users] OT Fanuc ot-c cpu power supply A16B-1212-0950
>> 
>> Hello everybody, I hope you're all doing well!
>> 
>> Sorry for the OT but I think this is a good place to ask for what I intend
>> to do.
>> 
>> One of our lathes has this Fanuc OT-C with this a16b-1212-0950 power
> supply
>> which is now failing. I can only use the lathe a couple of hours before
> the
>> control stops working and the PSU lights its alarm led.
>> 
>> The main problem is, these PSU are pretty expensive when purchased new,
>> and
>> the ones refurbished take almost three weeks to arrive in my country.
>> 
>> So, one of my ideas to keep the lathe running before I completely retrofit
>> it to LCNC is to install generic switching PSUs to match the voltages and
>> fool the CNC and it's control signals (which I don't know if they are too
>> complicated to fool but apparently there are only a couple of pins on the
>> PSU for that purpose, so It shouldn't be that complicated).
>> 
>> Is this nonsense? should I even bother? Because the next step really is to
>> adapt the machine to LCNC and forget about all these problems once and for
>> all. But if I can manage to fool the control this could be done in a
> couple
>> of hours.
>> 
>> Thanks as always for your help guys!
>> 
>> Leonardo
>> 
>> ___
>> 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] Run Fusion 360 on LInux

[Chris Albertson]

I ran Fusion360 for a long time on Linux.   It’s easy, just install a virtual 
machine and run Windows on the VM and Fusion on Windows.   I got pretty decent 
performance using a low-end Nvidia GPU card.

Now I run Fusion on an M2-Pro powered Mac.  Performance is better now that they 
have an Apple Silicone native version of Fusion.

I keep trying to move away from Fusion360 to FreeCAD but for the kinds of parts 
I like to make, Fusion360 works so much better.  Most 3D CAD works well for 
“geometric” mechanical-looking parts but I want “all compound organic curves” 
not only because of the way they look but because with 3D printing, organic 
shapes are far stronger for their weight.  Think about how strong a car hood is 
after they stamp it with compound curves and ridge lines, comparing that to a 
flat sheet of the same steel.  3D printed parts are thin sheets of solid 
plastic with low-density infill inside.   Structurally more like sheet metal 
than a machined metal part.   Yes, you can solid-fill the printed part but it 
massively adds weight. like 3X heavier and does not make the part even close to 
3X stronger.The best strength to weight is if you follow the example of car 
body designers.So can you design car bodies with FreeCAD?  I find it easier 
with Fusion360.  Still, I’d like to transition away from Autodesk.

I’m making robot parts.  These are very much like the case for a handheld 
battery powered drill.  Smooth and free-form curved on the outside but inside 
are gears and other machined mechanical parts and printed circuit cards.

I recently bought an iPad and will eventually learn to do CAD work on the iPad 
using the pencil interface.  It seems to have a learning curve.  The pencil 
seems to be a very good user interface tool for creative work of all kinds but 
it also seems to be a skill that will only come with practice.   My Vision (a 
dream) is to be able to sketch an object with a pencil, maybe it looks like a 
dog’s leg, then send the design to CAM and have it manufactured in a 
combination of metal and plastic.  

Next up is to try OnShape in a browser on the iPad.   




> On Dec 1, 2023, at 9:39 AM, Thaddeus Waldner  wrote:
> 
> Yeah, I saw this after I posted. I use an educational account.
> 
> Cheapest option that works appears to be “startup” at $150/3 years. 
> 
> 



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

Re: [Emc-users] OT: novel 3+ n axes

[Chris Albertson]

Yes, a magnet and a steel disk is the simplest possible design.  If you don’t 
need indexing.  Nothing would move.  The disk could slide over the magnet and 
then when energized the sliding would stop. 

The slop issue is normally solved by always rotating in one direction so maybe 
it could be controlled?

I can’t believe a 1/2 degree error is acceptable.  On a 100mm radius, a 0.5 
degree error amounts to mislocating the hole by almost 1mm.

My cheap made-in-India rotary table is way better than 0.5 degrees error.


> On Nov 23, 2023, at 12:11 PM, Ryan McIntosh  wrote:
> 
> If you are going to clamp a metal disc with anything, use an
> electromagnet...  But if it's a worm drive with that much slop, I guarantee
> there's some improvement to be found by controlling the thrust movement of
> the pinion gear.  Inserting a shim (could be punched out of an old set of
> feeler guages) under the gear where it shoulders up to the case can
> probably reduce the backlash by a significant amount.
> 
> *Ryan McIntosh*
> E4F Designs
> (204)770-3682
> r...@pebkac.ca
> 
> 
> On Thu, Nov 23, 2023 at 1:25 PM Chris Albertson 
> wrote:
> 
>> 
>> 
>>> On Nov 22, 2023, at 1:06 PM, gene heskett  wrote:
>>>> 
>>> I'd have to agree Andy.  I need to design a disk brake for my A axis as
>> there's half a degree of slop in the worm drive which if it gets to the
>> motor will be held, but the slop in the worm makes for sloppy looking
>> holes, but the activation method is still in my head. Even that is only
>> important when drilling a hole for the handle. I'll come up with something.
>> 
>> Would a simple disk brake work?  I think it might clamp the part any place
>> within that 1/2 degree of slop range.   Ideally, you would have an index
>> plate and some way to place a pin in a hole to lock the table.  But you
>> could only have a few holes and not enough holes for one at every place you
>> might want to stop.
>> 
>> What if the disk had “V” slots cut in the edge.  Like a very large US
>> Quarter Dollar, but with maybe 360 grooves.  Then you could wedge a shape
>> (like a tiny axe head?) into the slot and force the disk to stop on an even
>> degree.  You could use more than one wedge, space them around the disk.
>> 
>> A wedge would need a lot less clamp force than a brake disk pad because it
>> is a mechanical lock almost like an index pin
>> 
>> If you still want a “disk brake” why not use an actual disk brake from a
>> bicycle?  They are about the right size, have huge holding power, and don’t
>> cost all that much.   The ones you would want to use are hydraulic so the
>> electric actuator could be a couple feet from the rotating table.   They
>> still make cheaper ones that use cable but avoid those, they are too hard
>> to adjust.
>> 
>> Or maybe both,  Use the slot and wedge to aligh the table to the nearest
>> degree and then clamp it with a hydraulic disk brake.  Cost is under $100.
>> 
>> 
>> 
>> 
>> ___
>> 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] OT: novel 3+ n axes

[Chris Albertson]

> On Nov 22, 2023, at 1:06 PM, gene heskett  wrote:
>> 
> I'd have to agree Andy.  I need to design a disk brake for my A axis as 
> there's half a degree of slop in the worm drive which if it gets to the motor 
> will be held, but the slop in the worm makes for sloppy looking holes, but 
> the activation method is still in my head. Even that is only important when 
> drilling a hole for the handle. I'll come up with something.

Would a simple disk brake work?  I think it might clamp the part any place 
within that 1/2 degree of slop range.   Ideally, you would have an index plate 
and some way to place a pin in a hole to lock the table.  But you could only 
have a few holes and not enough holes for one at every place you might want to 
stop.

What if the disk had “V” slots cut in the edge.  Like a very large US Quarter 
Dollar, but with maybe 360 grooves.  Then you could wedge a shape (like a tiny 
axe head?) into the slot and force the disk to stop on an even degree.  You 
could use more than one wedge, space them around the disk.

A wedge would need a lot less clamp force than a brake disk pad because it is a 
mechanical lock almost like an index pin

If you still want a “disk brake” why not use an actual disk brake from a 
bicycle?  They are about the right size, have huge holding power, and don’t 
cost all that much.   The ones you would want to use are hydraulic so the 
electric actuator could be a couple feet from the rotating table.   They still 
make cheaper ones that use cable but avoid those, they are too hard to adjust.

Or maybe both,  Use the slot and wedge to aligh the table to the nearest degree 
and then clamp it with a hydraulic disk brake.  Cost is under $100.




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

Re: [Emc-users] OT: novel 3+ n axes

[Chris Albertson]

Wh

> On Nov 22, 2023, at 12:12 PM, andy pugh  wrote:
> 
> Quite a clever idea for machines with closed controllers and no easy
> way to provide motor power into the workspace.
> 
> With LinuxCNC it's probably always going to be better to be fully motorised.

What about cost?  If this device is cheaper than one with a motor then that is 
an advantage.   But I doubt this thing is cheap.




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

Re: [Emc-users] New Beaglebone

[Chris Albertson]

This is a VERY capable board.  The main thing it has is the FPGA and a very 
fast SYZYGY connector.   It could drive a multi-axis mill as well as any MESA 
board.  But it is not plug and play with LCNC.   You’d need to write code for 
the onboard FPGA and create a small PCB to connect SYZGY to a set of screw 
terminals using some high-performance buffers or optical isolators.

As for not having a connector for a monitor, that is no big deal as you could 
export the screen an iPad over WiFi.

I think this device’s best use case is robotics where very a high-speed control 
loop is needed.But my experience is that today, microcontrollers are good 
enough for this.




> On Nov 4, 2023, at 10:56 AM, andy pugh  wrote:
> 
> https://www.beagleboard.org/blog/2023-11-02-beaglev-fire-announcement
> 
> I don't know how usable it will be for LinuxCNC, as I don't know if we
> can compile for RISCV.
> 
> Also, I don't know if it has any video output?
> 
> -- 
> 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, 1912
> 
> 
> ___
> 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] Reverse Pulse Control on Dust Collectors

[Chris Albertson]

Today’s best practice is to put the code into version control in Git as soon as 
you have written the first few lines of code, long before you have even started 
testing it.

Github is not for distributing finished projects.  It is a public version 
control system. So if you have something and you want to ask “Why does this 
not work” placing it on GitHub is a reasonable thing to do.


> On Aug 22, 2023, at 6:47 PM, John Dammeyer  wrote:
> 
> I'm not ready with this code to distribute it as a project.  I haven't 
> touched it for almost 2 years.   To put it out as a project means writing a 
> user manual and far more comprehensive information on CANopen. 
> 
> My point of making the python and hal info available was to show that it's 
> pretty easy to integrate some sort of serial port control into linuxCNC.
> 
> John
> 
>> -----Original Message-
>> From: Chris Albertson [mailto:albertson.ch...@gmail.com]
>> Sent: August 22, 2023 4:50 PM
>> To: Enhanced Machine Controller (EMC)
>> Subject: Re: [Emc-users] Reverse Pulse Control on Dust Collectors
>> 
>> The best way to distribute stuff like Python source code is to place it on
>> GitHub and then every time you update it, send out a notice to tell people
>> what�s new and where it is.
>> 
>> GitHub is a free service and it offers more than just a place to store code. 
>>  It
>> also offers bug tracking and a forum and you can even build a simple web
>> page.
>> 
>> 
>> 
>>> On Aug 22, 2023, at 9:59 AM, John Dammeyer 
>> wrote:
>>> 
>>> I tried to send some python source code, an xml and text file as
>>> attachments, both as the files and in a zip and got the attachment
>>> blacklisted reply.  Is there another way so everyone on the list can get the
>>> attachment.  I can put it on my web site but eventually it will disappear
>>> rather than end up in archives.
>> 
>> 
>> 
>> ___
>> 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] Reverse Pulse Control on Dust Collectors

[Chris Albertson]

The best way to distribute stuff like Python source code is to place it on 
GitHub and then every time you update it, send out a notice to tell people 
what’s new and where it is.

GitHub is a free service and it offers more than just a place to store code.  
It also offers bug tracking and a forum and you can even build a simple web 
page.



> On Aug 22, 2023, at 9:59 AM, John Dammeyer  wrote:
> 
> I tried to send some python source code, an xml and text file as
> attachments, both as the files and in a zip and got the attachment
> blacklisted reply.  Is there another way so everyone on the list can get the
> attachment.  I can put it on my web site but eventually it will disappear
> rather than end up in archives.



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

Re: [Emc-users] Rigid tapping

[Chris Albertson]

I think with encoders that are not symmetric 50/50, you need to look only at 
the leading edges. (Or only at the trailing edges)

Yes,this cuts the resulution in half but removes the nose.I think for 
somethig with a lot of rotational inertia, like a spindle, you can use 
interpolation to get all the resolution you need.




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

Re: [Emc-users] Beaglebone black latency

[Chris Albertson]

As I said, this is what I did.  It is not a machine tool but a walking 
dog-robot.  The human interface can run on my iPhone.

Today we have things like $4 micro controllers (like the RP2020) that can run 
an RTOS.  It makes sense to use them as they are more capable than the IBM PC 
was when LCNC was new.

Would I ever do this for a machine tool?  No.  For me a machine tool is a tool. 
 I use it to make stuff.


> On Aug 5, 2023, at 5:26 PM, John Dammeyer  wrote:
> 
> Chris,
> I must admit I had a bit of a chuckle when I read your answer.  It's so easy 
> to just do this or use that or add these and then you are good to go.


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

Re: [Emc-users] Beaglebone black latency

[Chris Albertson]

Why start with a BBB/Pi4 and then add a screen and so on.

A better starting place would be a low-cost Android tablet.  Perhaps like this 
one.  For $120 you get the CPU (that runs Linux), a touch screen and WiFi.  
What is needed is a way to connect it to a machine tool.  
https://www.bestbuy.com/site/samsung-galaxy-tab-a7-lite-8-7-32gb-with-wi-fi-dark-gray/6464584.p?skuId=6464584
Samsung Galaxy Tab A7 Lite 8.7" 32GB with Wi-Fi Dark Gray SM-T220NZAAXAR
bestbuy.com

 I would use something like a small 32-bit micro controller to do the real-time 
pulses and switch monitoring.   If you would choose an ESP32 then the interface 
to the tablet is wireless.   But you could use USB as both devices have USB 
built-in.  USB has a built-in asocronus mode where data moves in strict 
real-time.  But this might not be needed if 100% of the RT loops are onthe 
microcontroller.

The ESP32 is a dual-CPU, 32-bit device with more then enough computer power to 
run a CNC controller.  I would 100%, for sure run an RTOS (FreeRTOS or Mbed) on 
the ESP32.  The tablet is just the user interface and “MPG”. 

I am currently using ESP32 to generate PWM control for a 12-axis robot and I’m 
getting sub microsecond pulse accuracy which is more then my application needs. 
 It is very fast and written 100% in Python.

If the user already had an Android tablet or wanted to buy a used tablet you 
really could under cut the cost of a $500 Chinese device.  

LCNC was written to take advantage of “cheap” community hardware.  This is when 
a basic 16-bit PC would sell for “only” about $3,500.   Today’s commodity 
hardware is called a “cell phone” and they are MUCH more powerful then the old 
PC was.





> On Aug 5, 2023, at 3:23 PM, John Dammeyer  wrote:
> 
> Me too.  A closer look at the fine print and selecting the correct button 
> shows the actual CNC controller is $513 or $541 for 3 or 4 axis respectively. 
>  The MPG is priced separately at $96.
> 
> And this is why a BBB with a cape that included a Break Out Board with a few 
> relays etc. for under $100 would be quite the bargain.  The question is then 
> what sort of display would be good.  And you still need some sort of MPG to 
> go with it for also about $100.
> 
> For $94 I can get, from amazon.ca  a 7inch HDMI LCD (H) 1024x600 Monitor IPS 
> Screen Capacitive Touch Display Supports Raspberry Pi Jetson Nano BeagleBone 
> Black Banana Pi & Microsoft XBOX360, Sony PS4,Nintendo Switch
> 
> And therein lies the problem.  Creating something similar to the $500 Chinese 
> CNC controllers with touch screen will ultimately cost in the same $500 
> range.  But unless the MachineKit or LinuxCNC version is stable with a target 
> life similar to the life of the machine tool no one will spend the energy and 
> time to make that.
> 
> The market isn't big enough to justify the work required.  Which is why so 
> many of the Beagle Capes are no longer available.  Next week there's an new 
> ESP32 with more bells and whistles.  Won't do what the Beagle does but ...
> 
> John
> 
> 
>> -Original Message-
>> From: Martin Dobbins [mailto:tu...@hotmail.com]
>> Sent: August 5, 2023 11:44 AM
>> To: Enhanced Machine Controller (EMC)
>> Subject: Re: [Emc-users] Beaglebone black latency
>> 
>> It makes me wonder which operating system they put on that Chinese
>> controller, and what motion control system.
>> 
>> Martin
>> 
>> 
>> From: John Dammeyer
>> Sent: Friday, August 4, 2023 7:41 PM
>> To: 'Enhanced Machine Controller (EMC)' > us...@lists.sourceforge.net>
>> Subject: Re: [Emc-users] Beaglebone black latency
>> 
>> My biggest issue with the Xylotex BBB Cape for CNC was that the logic for
>> ESTOP was inverted.  ESTOP was a N/O switch so if a wire was broken you
>> wouldn't find out until you tried to save a body part from being injured with
>> the ESTOP.
> 
> 
> 
> ___
> 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] Beaglebone black latency

[Chris Albertson]

I think the BBB was popular because of the twp PRUs.  These are programmable 
processors that are separate from the CPU.  So you can get very accurate 
real-time pulses out even with poor latency under Linux.

However, this does require programming the PRUs.Machinekit had a HAL unit 
for the PRUs.  I don’t know if LinuxCNC has this or not.




> On Aug 4, 2023, at 11:18 AM, andy pugh  wrote:
> 
> On Fri, 4 Aug 2023 at 01:48, alanmthomason--- via Emc-users
>  wrote:
>> 
>> Hi There.I'm trying to work through setting up linuxcnc on a beaglebone
>> black.  I've run the latency-test, with results that are not looking
>> good.has anyone else run this test with a Beaglebone black?  Is there any
>> way to improve results.
> 
> I think that when Machinekit were fond of the Beaglebone they tended
> to run it headless. I can't recall whether that was just poor graphics
> performance in general, or whether it was due to latency issues.
> 
> You could try installing the (old) Machinekit image to see if the
> latency with that kernel is better.
> 
> -- 
> 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, 1912
> 
> 
> ___
> 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] LinuxCNC build on BeagleBoneBlack

[Chris Albertson]

I have some Pi and BBB boards that I use for robotics projects.  I use them 
because they are physically small and can run on battery power.  But do people 
need that for a machine tool?   Are people building portable battery-powered 
CNC milling machines?If the machine is very heavy and requires AC mains 
power then why not make life easy for yourself and use a big old PC.  

Seriously, you can buy a good used PC that will run circles around a “BBB” for 
not much money.  Add a Measa FPGA card and you are good to go.
https://www.newegg.com/p/N82E16883461141?Item=9SIABX7JBU8332
Refurbished: DELL Desktop OptiPlex 3040 SFF Intel Core i5 6th Gen 6500 
(3.20GHz) 16 GB DDR4 256 GB SSD DVD Intel HD Graphics 530 Windows 10 Pro Grade A
newegg.com






> On Aug 4, 2023, at 9:44 AM, John Dammeyer  wrote:
> 
> I'm curious.  I have one of the original Xylotex boards.  The link to this 
> web site implies there's a newer version along with a link to 4GB BeagleBone 
> Image for MachineKit which I have read hasn't been supported for years.
> http://xylotex.netfirms.com/OSCommerce/catalog/product_info.php?cPath=40_id=169
> 
> My BBBs (3 of them) are all sitting in a box unused.  This includes things 
> like LCD displays, CAN bus capes etc.   I had thought about adding the BBB 
> with CNC to my Gingery Lathe which currently is controlled by my Electronic 
> Lead Screw product.
> 
> What is the status in general with the BBB and MachineKit or LinuxCNC.   
> Should I just put my stuff on EBAY?
> 
> I ask because I also have a 1GB Pi4 with a MESA 7i92H as a desktop play toy.  
>  I think the 1GB is a bit small for LinuxCNC but so far it has worked.  I'm 
> not sure whether support for LinuxCNC on the PI4 has also gone by the wayside?
> 
> So if someone had a BBB and a cape, installs a version of LinuxCNC 
> (MachineKit) do they run into what appears to happen often now with trying to 
> upgrade and unknown web links?
> John
> 
> 
>> -Original Message-
>> From: andy pugh [mailto:bodge...@gmail.com]
>> Sent: August 3, 2023 1:20 AM
>> To: Enhanced Machine Controller (EMC)
>> Subject: Re: [Emc-users] LinuxCNC build on BeagleBoneBlack
>> 
>> On Wed, 2 Aug 2023 at 17:51, Thomas J Powderly  wrote:
>> 
>>> Is there an explanation of gpiod?
>> 
>> It's a replacement for the sysfs interface for gpio (which hal_pi_gpio
>> uses, for example).
>> 
>> There is an article about it here:
>> https://lloydrochester.com/post/hardware/libgpiod-intro-rpi/
>> 
>> I have a working-but-slow demo here:
>> https://github.com/LinuxCNC/linuxcnc/commit/15138feb4a1a811a7b61a6605
>> 55d78b85bdcf86c
>> 
>> (later commits in that branch are not working at all at the moment)
>> 
>> to use the BBB output from gpioinfo above:
>> 
>> loadrr hal_gpio inputs=P8_25,P8_24  outputs=P8_10,P811
>> 
>> --
>> 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, 1912
>> 
>> 
>> ___
>> 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] CNC machining setup cards

[Chris Albertson]

> On Jul 11, 2023, at 5:42 PM, gene heskett  wrote:
> 
> ... I'll bet real money that OpenSCAD is easier to learn. 

Openscad is certainly easy for some simple things but impossible for others.  
Try making a cordless power tool in OpenScad.

But in Fusion a power tool is used as an intermediate-level tutorial project.  
The screenshot below is something a person could learn to make after some weeks 
of study.  I doubt anyone would even attempt it in OpenScad.   Then assuming 
you do make the OpenScad model, could you make changes in real-time as the 
client pointed out changed he would like

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

Re: [Emc-users] [Emc-developers] Trajectory planner shortcomings

[Chris Albertson]

> On Jul 6, 2023, at 9:22 AM, Roland Jollivet  wrote:
> 
> The small 4-axis CNC machine,'Pocket Nc' uses LinuxCNC as far as I can
> tell, and it seems to run just fine. Or does it?

Does this device move all four axes at the same time while cutting?  Or does it 
use the fourth axis only to reposition the work?   



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

Re: [Emc-users] CNC machining setup cards

[Chris Albertson]

It is not a public Beta.   First you get Fusion 360 running which means setting 
up an account.  They also need to see that you use Fusion regularly.

Then you apply for membership in the “Insider’s Program”.

More here 
https://feedback.autodesk.com/key/Fusion360Insider

There are other new fetures beibg tested, not just Apple Silicon.  One of them 
is a fastener library for nuts and bolts and they say there is a way to use yur 
company’s part number system for you fastener library.  Some other stuff about 
tool path generation.  I see something about version control when multiple 
users work on the same desgin files too. They add features frquently and 
then  few months later we see them in the public release.

What you can see and use in the Beta also depends if you have the free or paid 
version.  You can’t use paid beta features is you do not have a paid 
subscription


> On Jul 5, 2023, at 7:56 PM, Thaddeus Waldner  wrote:
> 
> Would you mind detailing how to get this beta?
> 
>> 
>> BTW, Fusion 360 has just come out with a Beta release that runs native on 
>> Apple Silicon.   The new Apple processors are astonishingly fast.   The 
>> above part is designed to be printed but it is fun to let Fusion great 
>> g-code files for milling and see what it comes up with.
>> 
> 
> 
> 
> ___
> 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] CNC machining setup cards

[Chris Albertson]

> On Jul 5, 2023, at 11:36 AM, gene heskett  wrote:
> 
> On 7/5/23 13:29, Thomas J Powderly wrote:
>> Since this thread has 'evolved' into free 3D modeling software...
>> Can I ask how to get Fusion 360 on a Linux box?
>> I remember several failed tries but I font recall why.
>> LinuxCNC community people seem to really like it,
> 
> That seems to be true, but it is not free anymore. So I'm making parts for 
> stuff on 3d printers using OpenSCAD as the design software.  You can be 
> productive quite rapidly in it.

What?   Fusion 360 is still free for personal use or even for commercial use as 
long as the bussines makes less than $50K.

If you want 3D CAD software that can run on Linux then you could try
1) FreeCAD,  Not unlike Fusion 360 but is mre limited and lacks sophisticated 
manufacturing ability
2) OnShape,  Online, runs in a web browser and nit somewhat like FreeCad but 
has ZERO manufacturing ability
3) OpenScad.  Good for very geometric designes that can be described in a 
modeling language.  Good for gears, bushing and such but VERY poor if you 
wanted to maybe make a car body part or a motorcycle engine cover

If the goal is to convert STL files to “real” CAD drawings that can be edited.  
 Fusion is your best bet and even then you have to do a lot of manual work


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

Re: [Emc-users] CNC machining setup cards

[Chris Albertson]

> On Jul 5, 2023, at 10:27 AM, Thomas J Powderly  wrote:
> 
> Since this thread has 'evolved' into free 3D modeling software...
> 
> Can I ask how to get Fusion 360 on a Linux box?

It is easy.  But you need Linux on Intel.  ARM-based Linux will not work

1) install a virtual machine system.  VMware works well and it is free
2) install Windows 11 in the virtual machine.  It is free also as long as you 
don’t care if it is “activated”.  The only problem with non-activated is you 
get nagged to activate it.
3) install Fusion 360 on the virtual Windows 11

It runs surprisingly fast if your Linux machine has enough cores and RAM to run 
both Linux and Windows at the same time.   In VMWare’s setting asign at least 
about 16GB and 4 cores to the Windows VM.   It can work with less but it gets 
slow.

I just moved to an Apple Mac M2 Pro.  This is a 10-core Apple Silicon system.  
A Beta release of Fusion 360 just came out that runs native on Apple Silicon.   
It is very fast now. and also very cost effective.  Apple is MUCH less cost 
than a top of the line Intel i9

A $2000 Intel i9 or Xeon with or 64 GB RAM and fast M.2 based SSD running 
Linux/VMware/Windows/Fusion is very usable for projects like robotics that have 
hundreds of complex 3D parts.   But such a PC is not cheap.   An Apple Mac mini 
configured to cost about $800 to $1,200 will give a better user experience.   I 
actually have both of these here now.   That said, I’d bet simply running 
Windows 11 directly on high-end hardware (with a good GPU) would be even better.

For many simple projects the above is over kill.  Use OpenScad or FreeCAD for 
many projects that have only a few parts and are very geometric.   Fusion 360 
(or the like) beciase needed if you project is something like a 
battery-operated drill that has internal gears, and complex shape housing with 
rubber over-molding.  Fusion is good for work of about the complexity of a 
Millwaulkee power tool.   The Mac or Intel i9 would be powerfull enough to 
Support the design of a typical power tool.For simpler parts less i 
required.

I have another data point at the low end.   An Intel “Core 2” with 8GB RAM and 
integrated graphics from 2014 is just barely able to run a 
VMware/Windows/Fusion stack.  I would not recommend it.   While my 16-core Xeon 
with 64 GB RAM runs it well as does the new Mac. (the Mac being a little faster)



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

Re: [Emc-users] [Emc-developers] Trajectory planner shortcomings

[Chris Albertson]

With no look-ahead, it might work well for those who use the fourth axis as a 
way to rotate the part only to expose a new side for more 3D milling.   In 
other words, you don’t need look ahead if you are not cutting will rotating.

But if all four axes are moving at the same time, looking ahead would be 
important.

Lookahead is easy to program and understand if the kinematics is trivial.  3D 
milling is what they call “trivial kinematics”. But as soon as the relation 
between movement in the part’s x,y,z space and the machines axis space is not 
direct the math gets harder by quite a lot.I usually ran away when the 
solution involves things like Jobobian matrixes and the like.

You can avoid all this if all four axis don't move at the same time.



> On Jul 4, 2023, at 3:23 PM, Sam Sokolik  wrote:
> 
>> 
>> The problem seems to be that our trajectory planner lack the ability to
>> look ahead for any rotary axis. This means that a coordinated move that
>> includes a rotary move will be dysfunctional at best.
> 
> 
> I read that as - poor look ahead - not 'doesn't work'
> 
> On Tue, Jul 4, 2023 at 5:21 PM Sam Sokolik  wrote:
> 
>> Lol - then I guess I read too far into it..  I have run 4th axis rotary
>> motion just fine..
>> 
>> On Tue, Jul 4, 2023 at 5:15 PM andy pugh  wrote:
>> 
>>> On Tue, 4 Jul 2023 at 22:50, Sam Sokolik  wrote:
 
 He is not saying it doesn't work...
>>> 
>>> Actually, he did say that.
>>> 
>>> --
>>> 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, 1912
>>> 
>>> 
>>> ___
>>> 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] CNC machining setup cards

[Chris Albertson]

The problem with transforming an STL file to a “real” CAD file is that the STL 
is only an approximation of the true shape.  The software would have to have 
some intelligence to see that the designer intended to (say) have a curved 
profile revolved around an axis to make a complex shape on a lathe.  A human 
can see this quickly, but software can’t (yet).  So you need a human in the 
loop to do it correctly.

Yes, you can convert.STL files into editable files and save them as .STEP.   
But not automatically.   Fusion 360 allows you to convert a triangle mesh (stl 
file) to native mesh and it can remove all the redundant points and do some 
curve-fitting and then you have an object that is a b-spline surface and is 
editable in a normal CAD system.   I’ve done this for designs as complex as a 
working quadruped robot. It takes some work but you can actually turn STL 
files into geometric objects that can be edited and parameterized.

Here is a part I made that started life in Blender, was saved as STL, Imported 
to Fusion turned into normal CAD and then greatly modified.  Two of these 
assemblies in each robot are used to connect the legs to the body.   You can 
see it in the on-line viewer linked below

https://a360.co/3ra8Dyh

BTW, Fusion 360 has just come out with a Beta release that runs native on Apple 
Silicon.   The new Apple processors are astonishingly fast.   The above part is 
designed to be printed but it is fun to let Fusion great g-code files for 
milling and see what it comes up with.


> On Jul 4, 2023, at 5:01 AM, Gregg Eshelman via Emc-users 
>  wrote:
> 
> Caligari trueSpace can import STL and it can save to a wide range of formats.
> 
> However, it hasn't had any official support since Microsoft bought Caligari 
> to position trueSpace as a competitor to Google's Sketchup. (I call it Messup 
> because I've seen some of the worst geometry ever made by people using 
> Sketchup.) Google had Sketchup for people to use to populate Google Earth 
> with 3D models of buildings. Microsoft whipped up Virtual Earth... and nobody 
> cared. People weren't stepping up to voluntarily use the free version of 
> Sketchup to make 3D stuff for Google Earth either.
> 
> Microsoft quickly swept trueSpace under the digital rug, left the site for it 
> unchanged for a while until they got around to deleting it. Lots of people 
> grabbed the free downloads and some patches and other things have been made. 
> There's a ton of tSx plugins available for free, including many formerly 
> commercial ones. Most tSx for version 6.x and some for 5.x will work but most 
> for older versions won't.
> 
> One glitch that nobody has fixed yet is when importing some 3D file formats 
> it scales the mesh down to make the largest axis, X, Y, or Z, exactly 8 units 
> of whichever is the currently selected unit in the workspace. With STL I load 
> up the model in a slicer to get the proper XYZ sizes then scale up to match 
> in trueSpace then save a copy in its native COB format.
> 
> I always save in COB (frequently! it appends an auto-incremented number to 
> the file names) and export to STL. Exports of the formats it supports are 
> fine. It's just importing it has the size issue with. Would have been so nice 
> if Microsoft hadn't killed it by buying the company. I've used every 
> trueSpace version since 2.2a so I'm used to its oddities it inherited by 
> originally being an Amiga program.
> 
> If you want to give it a whirl there are a few versions of 7.x here 
> http://truespace3d.free.fr/index.php/truespace-7-6/
> I use 6.6 since the Model Side of 7.x is essentially version 6.6 and I never 
> could get into the new 7.x Workspace. It's so different from 2.x through 6.6. 
> Hmmm, I don't remember if Workspace side has the size issue with importing.
> 
> On Tuesday, July 4, 2023 at 03:10:58 AM MDT, gene heskett 
>  wrote: 
> 
> On 7/4/23 02:01, andrew beck wrote:
>> I run a CNC machine shop full time.
>> 
>> We always want the step files.  Or SOLIDWORKS files etc.
>> 
>> And we make it own gcode from that.
>> 
>> It would be a nightmare to run someone else's gcode lol.
>> 
> I'll have to agree Andrew. Most of the stuff on thingiverse for 3d 
> printers is in .stl formats, cura can usually make something useful out 
> of them. but its maddening to see an .stl that needs fixed, and I've not 
> found anything that can convert an .stl back into something that can be 
> edited in openscad. So I wind up about 95% of the time, using the image 
> as a guide to compose something that looks like it well enough to work.
> 
> 
> ___
> 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] Drilling holes in the back of a stepper motor

[Chris Albertson]

I’m working on the same kind of thing. I want to add an encoder to a BLDC drone 
motor and we don’t have much shaft length at all.

The design “secret” is to realize that you do NOT need a plate or any material 
at all between the motor and the encoder.  You can support the encoder from the 
rear.

Make a 5-sided box housing that fits over the end of the motor and the encoder 
attaches to the inside of the box and also to the extended shaft.  he encoder 
does not even need to touch the motor housing.  Or maybe you only use locator 
pins in very shallow holes in the motor.

3D printers and CAD software make this made of complex parts easy.  It is easy 
to p3d print blind threaded holes to mount the encoder from the inside of the 
enclosure, then you screw the enclosure box with mounted encoder to the back of 
the motor.   Finally to tighten the set scre to the motor shaft. (Leave a hole 
for a hex wrench.  Then a pan head screw closes the access hole.

In my case I have the motor mounted inside two parts that assemble like a 
clamshell. Parts go into the shells then you assemble the shells.




> On Jun 26, 2023, at 12:28 AM, John Dammeyer  wrote:
> 
> Unfortunately access to the assembly is required to center things and then 
> tighten screws.  Nice as an enclosure might be it wouldn't solve the problem
> 
> What I have done, since I don't have the press in type screws, is with a 
> countersunk 4-40 flat head I used my little DB200 lathe to remove some of the 
> countersink taper.  I think gluing this into a 0.055" thick mounting plate 
> will prevent them from turning and also then fit the 4 mount locations along 
> with holding down the encoder.
> 
> Photo shows the screws and screen capture the plate I drew up.  I'll use 4x 
> 3mm screws to go into the 4 corners of the motor.  The two 4-40 threaded 
> studs (in effect) will come up through the other two holes.  If the corner 
> holes are large enough there should be enough play to be able to center the 
> disk in the encoder.
> 
> The down side is I've totally forgotten how to use my tool height setter and 
> touch probe to run the G-Code I've now created for profiling the plate.  
> Should have written myself better directions.
> 
> Sigh...
> John
> 
> 
> -Original Message-
> From: Gregg Eshelman via Emc-users [mailto:emc-users@lists.sourceforge.net] 
> Sent: June 25, 2023 11:22 PM
> To: Enhanced Machine Controller (EMC)
> Cc: Gregg Eshelman
> Subject: Re: [Emc-users] Drilling holes in the back of a stepper motor
> 
> Make covers that fit over and hold the encoder. Snap the encoder into the 
> inside of the cover, and the corners of the cover have holes to put screws 
> into the existing holes in the motor end cap. This would have the benefit of 
> protecting the encoder and the end of the motor shaft.
> 
> 
> On Sunday, June 25, 2023 at 03:33:26 AM MDT, John Dammeyer 
>  wrote: 
> 
> The problem is the rear shaft isn't that long.� 
> 
> I went through the process of considering an adaptor plate that fastened to 
> the 4 holes.� � There may be a fancy way of making a thick plate with a 
> thinned area just for the encoder mount.� But it would have to have a stud 
> protruding away from the back and then a nut.� The plate can't be thick 
> enough to hold both the encoder and threads for screws and still have the 
> encoder disk mount to the back shaft.
> 
> The alternative would be to press fit on a shaft extension and then use the 
> stepper motor to turn the extension to be symmetrical with the axis of the 
> motor shaft.� Then I could use a thicker mounting plate.
> 
> Or, given that the motor drives a 25:1 planetary gear and I'm really only 
> interested in tracking motor revolutions and detecting stall conditions 
> (hence quadrature) I could likely get away with a custom disk and some 
> slotted sensors too.� Also more complicated to build.� 
> 
> Trying to keep it simple and the easiest is to have the encoder screw 
> directly to the back of the motor.� However with StepperOnline motor by the 
> time it's here in Cdn $ it's over $100.� So I can take the risk and drill 
> holes in the back.� Or get creative with other approaches.
> John
> 
> 
> ___
> 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] Drilling holes in the back of a stepper motor

[Chris Albertson]

> Is there any reason I can't clamp the motor in a vice and pop in two holes
> and tap them with a bottom tap at the spacing of the mounting flange?

The metal end plate might be a lot thinner than it appears from the outside.  
Is it thick enough to support threading?  It might be a thin diarist part.  
Maybe look inside first?

But on the other hand, the load on the mounting screws would be nearly nothing. 
  

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

Re: [Emc-users] Sewing Machine Stiffness

[Chris Albertson]

> On May 20, 2023, at 2:12 PM, marcus.bow...@visible.eclipse.co.uk wrote:
> 
> 
> Interesting idea, which makes a lot of sense.
> So you would end up with a 'solid' epoxy/glass mix. What would its 
> performance be like compared to a heavy dense epoxy/stone mix?


Glass and stone are not much different.   Both make a hard composite.  But the 
shape of the chopped fiber is different from crushed stone.  Fiber is more than 
100:1 length-to-width ratio, even if chopped very short. 

People can argue that cast iron is stiffer.   But the iron is whatever shape 
and size it happens to be, the glass fiber can be custom-made to any size and 
shape up to making your own B2 bomber airplane.   So even if an equal size of 
iron is stronger, make the glass part 2X larger.

Even 3D-printed PLA plastic can be strong enough.  I have a 3D-printed CNC 
conversion of a harbor freight milling machine.  Normally they use metal parts 
for the motor mounts and such but PLA can be just as rigid if you make the 
parts thicker.

If you have access to a foundry then you can make your own custom case iron.

Metal is best if you need the part to fit in a small space but plastic can be 
as strong if you have the space for a larger part.   For example you might use 
module 0.5 steel gears but these are expensive to make. You could 3D print 
module 2.0 gears that are as strong but you will have a very different-looking 
machine

On the other hand, even a cheap sewing machine frame can be strong enough if 
you limit the cutting speed the keep the forces down.  



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

Re: [Emc-users] Sewing Machine Stiffness

[Chris Albertson]

If you are going to the trouble of filling a frame with epoxy then two things…

1) Why bother with the frame, just print a plastic mold and fill that. with 
epoxy.   Then it will be the exact size and shape you need.

2) Don’t use epoxy as it is not very stiff,  Mix the epoxy with glass fiber 
such that the fiber-to-epoxy ratio is as high as you can possibly make it.   
You don’t need expensive carbon fiber as you don’t care about weight.   You 
want as much glass in the mix as possible.  The usual way is to use woven 
glass, wet it with epoxy then use pressure from clamps or a vacuum pump to 
compress the part and push out as much epoxy as possible.  But you can also mix 
chopped fiber with the liquid resin and make a paste. 

This is actually very low-tech and can be done with simple hand tools like 
paintbrushes and scissors.

The problem is the same as with any new design, you have to make some 
prototypes and measure them and likey redesign them a few times.  This is why 
people usually copy existing proven designs.




> On May 20, 2023, at 11:28 AM, Gregg Eshelman via Emc-users 
>  wrote:
> 
> You'd need to get a frame casting, set up indicators, then do some pushing, 
> twisting, and leaning on it to see how much it moves.
> 
> What could stiffen it is filling all its internal space with epoxy. The upper 
> casting from one of these
> https://china-highly.en.made-in-china.com/product/gSHxMeWyblRA/China-Hl-246-Long-Arm-Compound-Feeding-Super-Heavy-Sewing-Machine.html
> could do the job since it bolts on from above it could be mounted over any XY 
> base you can build.
> 
> Just might be workable for a wood router for sign carving. Z travel would be 
> limited to whatever vertical slide you mount, and if you space the arm up 
> higher.
> 
> I'd expect that the really long ones with arms around 30 inches, would be 
> pretty well vibration damped in order to handle high speed sewing in heavy 
> materials like canvas. Sewing is mostly a short vertical motion. I'd assume 
> the main area of concern for adapting a long arm sewing machine to routing 
> would be resistance to bending sideways and twisting around the long axis of 
> the arm.
> 
> 
> 
> On Saturday, May 20, 2023 at 11:00:22 AM MDT, Thomas J Powderly 
>  wrote: 
> 
> I saw an overarm router recently
> 
> and wondered if a sewing machine frame was stiff.
> 
> Compared to a desktop gantry mill.
> 
> 
> I imagined a makita router mounted on the over arm
> 
> minimal Z travel ( 150mm at most)
> 
> 
> I can find castings for industrial machines pretty cheap in qty 1
> 
> What i see are C frames, single casting, with base plate as long as over 
> arm.
> 
> 
> Any thoughts?
> 
> ( I don't have a sewing machine to lean on ;-)
> 
> 
> ___
> 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] Sewing Machine Stiffness

[Chris Albertson]

I think there would be a huge variation.  A cheap $100 consumer sewingmachine 
would not be at strong as an industrial sewing machine that sells for 20 times 
as much.

One problem with many homemade gentry machines is that people us too-thin 
material for the base.If the gantry rails are screwed to a sheet of 
plywood, the entire machine will be as rigid as a sheet of plywood.

I’ve been working on some kitchens lately and visiting granite yards.  They 
typically have cut-off leftovers for not a lot of money.  A slab of stone would 
be much better than a plywood sheet and not expensive if it only has to be 
sewing machine sized.



> On May 20, 2023, at 9:28 AM, Thomas J Powderly  wrote:
> 
> 
> I saw an overarm router recently
> 
> and wondered if a sewing machine frame was stiff.
> 
> Compared to a desktop gantry mill.
> 
> 
> I imagined a makita router mounted on the over arm
> 
> minimal Z travel ( 150mm at most)
> 
> 
> I can find castings for industrial machines pretty cheap in qty 1
> 
> What i see are C frames, single casting, with base plate as long as over arm.
> 
> 
> Any thoughts?
> 
> ( I don't have a sewing machine to lean on ;-)
> 
> 
> TomP
> 
> 
> 
> ___
> 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] Medium voltage BLDC motor

[Chris Albertson]

iling 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.
> --
> "There are four boxes to be used in defense of liberty:
>   soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author, 1940)
> If we desire respect for the law, we must first make the law respectable.
>   - Louis D. Brandeis
> Genes Web page <http://geneslinuxbox.net:6309/>
>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Custom Cast Iron castings

[Chris Albertson]

If you can print the pattern and supply it to the foundry, the cost will be
low.  But you should talk to the foundry about how to make the pattern.
Some patterns will warp as they cool.  I think it has to do with the part's
thickness, or rather if there are thick and thin parts in the same
pattern.  Also, some patterns will shrink as they cool, and you need to
design the pattern differently from the final part. I know someone who
wanted to make some lighter-than-normal cast iron cookware, and the foundry
said the pattern was too thin in some places.

But as you made the pattern in CAD you can print it can give it to the
foundry and they can tell you, then fixing it is very easy, not nearly as
hard as making a new wood pattern.

Unless you have a lot of experience with making cast iron parts, you want
to use a LOCAL foundry where you can talk to the person there in person.
Then he can tell you how to redesign the pattern if that is required.   You
need to follow their design rules

On Mon, Apr 3, 2023 at 5:46 AM Todd Zuercher  wrote:

> Google up some names and make some phone calls.  There are literally
> thousands of small foundries across the US.  The trick is finding which one
> wants your money and is most convenient for you.
>
> Todd Zuercher
> P. Graham Dunn Inc.
> 630 Henry Street
> Dalton, Ohio 44618
> Phone:  (330)828-2105ext. 2031
>
> -Original Message-
> From: hubert 
> Sent: Sunday, April 02, 2023 1:11 PM
> To: emc-users@lists.sourceforge.net
> Subject: [Emc-users] Custom Cast Iron castings
>
> [EXTERNAL EMAIL] Be sure links are safe.
>
>   I have a new spindle that I would like to mount on my Mill.  It larger,
> faster and more powerful than my current spindle.  I have created a 3D
> model of the desired spindle holder box that I would like to get converted
> to cast iron.
>
> I would appreciate recommendations on sources of custom castings.
>
> I live in central Texas.
>
> Hubert
>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
>
> https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Femc-users=05%7C01%7Ctoddz%40pgrahamdunn.com%7Cb7c4ae84bd1546b4b75008db339f3b9e%7C5758544c573f47cebee96c3e0806fb43%7C0%7C0%7C638160531218900158%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C=rkhVNnOD5ghrL%2BUAxVXh0vwtzjq2NXxvWMPYm872ToA%3D=0
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] OT - Thoughts on making SheetCam open source

[Chris Albertson]

The problem with SheetCam is that I had to use Google to find out what it
is.Then I found the website and it is still not clear how I would use
it to convert my CAD files into g-code or if it would work for that.   You
might want to describe the use cases and sample workflows that go from an
idea to a finished part and where SheetCam fits into that.

Yes "Dumping it on Github" is required but not even close to sufficient to
make it into a successful project.   You will need to put all the above
information in the README.MD file.  The documentation has to be open source
too.

I think the easiest projects ones to get started are software tools because
all your users are software developers and have the skills to modify your
open source project.   But a product targeted to machinists means that the
vast majority of your users are not software developers and would lack the
skills required to contribute.   You have to solve that problem.

Finally, the platform matters a lot. Is it written in something
obscure?  An open-source CAM system written in Haskell is not going to take
off.But if written in Python you have 1,000 times more potential
contributors. What is the run-time platform?   Windows users almost
never contribute to Open Source projects.  It needs to be multi-platform.

So, ...  I idea would be...

I think the best way to transition this is to change just a little at a
time, first move the code github and drop the price to zero.   Then offer
to sell technical support contracts that include pre-compiled binary
files.   If you are lucky, dropping the price will dramatically increase
"sales" and with a much larger user base, you will be as busy as you want
to be doing support.   You can control the number of hours you work by
adjusting the price of the contract.   Others will see they can make money
by consulting and will compete with you, then you can drop out  or just
work less. Getting machinists to contribute will be hard, better I
think to give them some incentive by encouraging them to compete with your
tech support business.   But this depends on a MUCH larger user base.   You
get that by dropping the price to zero and improving the technical
documentation and quite a lot of self-promotion.

About self-promotion.  It should be easier with an open source product.
You will be asked to leave many forums if you are selling something but
giving something away is different.

In short:  Do a smaller transition.  Continue to "sell" it but drop the
price to zero, offer a new service for X dollars per year and yes, do put
it all in Github  Later you stop doing the sales and consulting and
leave it on GitHub.   The point is that there is no "hard" transition.

Also you might work on showing how SheetCam works with the CAD programs
people are using.   Perhaps FreeCad is a good starting place.

On Thu, Feb 23, 2023 at 5:27 AM Les Newell 
wrote:

> Hi all,
>
> I'm intending to retire some time this year and I am trying to decide
> what to do with SheetCam. One option that I'm seriously considering is
> making SheetCam open source (GPL). If this is too off topic for this
> list please tell me.
>
> I really want to see SheetCam continue to be developed and improved. My
> accountant thinks I'm nuts to even consider going open source but my
> biggest concern with selling it is that I don't want to see it get
> turned into a cash cow with very little development, pretty much the way
> Mach3 ended up.
>
> While I have contributed to a few open source projects including
> LinuxCNC I have never started one. To be successful an open source
> project needs a strong leader or team of leaders. I don't want to be a
> leader. I don't expect to just dump it on Github and walk away but I
> also don't want to end up doing just as much work as I'm doing now. My
> ideal scenario would be to recruit a team of core developers then
> gradually reduce my involvement. The question is, am I likely to be able
> to recruit those developers?
>
> Basically at the moment I'm looking for advice and opinions.
>
> Les
>
>
>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Parallel Port Current?

[Chris Albertson]

Damage from over-current is cumulative.  The bond wires that go from the
chip to the lead frame are VERY thin and will heat up and can only do this
for so many cycles.

What you might try, to protect the other pins, is to use a higher value
current limiting resistor on the LEDs



On Wed, Feb 22, 2023 at 8:30 AM Todd Zuercher  wrote:

> I have an older Linuxcnc installation that uses a parallel port.  It has 4
> TCRT5000 optical sensors, that I was powering the IR LED with an output
> from the parallel port (and reading the IR sensor with an input.)  It had
> been working fine for some time, but I only use this equipment once a
> year.  This year when I drug it out of storage I could not get one of the
> sensors to work.  After several hours of checking and rechecking every
> connection and trying multiple sensors, I figured out that the LED on the
> sensor that wasn't working simply could not draw enough current from the
> parallel port output to light bright enough, even though it had been
> working in the past.  I switched it from pin14 to pin7 and full
> functionality was restored.  Does this sound like I've partially ruined
> pin14 on the parallel port? (It still turns on and off.)  Do different
> parallel port output pins have different current sourcing capabilities?
>
> I'm a little surprised that this happed to this sensor rather than the
> other 3, because it is fed through about 30 more feet of wire than the
> other 3 that are still working happily (fed by output pins 2-4.)
>
> I guess I should probably set up an alternate power source for the LEDs on
> the sensors, but they had been working fine before and I didn't feel like
> doing the additional wiring.
>
> Todd Zuercher
> P. Graham Dunn Inc.<http://www.pgrahamdunn.com/index.php>
> 630 Henry Street
> Dalton, Ohio 44618
> Phone:  (330)828-2105ext. 2031
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] 7c81 Questions

[Chris Albertson]

A few comments,  Don't wait untillyu have everything figured out to
start.  Do just the bare minimum first.  Leave the spindle manually
controlled and make the x, y, z axis work but do the limit switches later.
Also impletmt some kind of "bomb-proof E-stop" that works even if the
software is dead.  An A/C power switch works, or maybe better is to disable
the motors.  But keep it independent of the computer for now.

This is the minimum to make thing run.  Then AFTER this works add one thing
at a time.

After you have the minimum working, you next step is to make an actual
part.   This will debug you process for making G-code and that moters turn
in the correct direction and so on. This is non-trivial.  Are you using
CAD to make tool paths?   It all has to be correct from screen to metal
cutting.

Then later after you are using this add the A axis and home/limit and
figure out if a remote display will work and so on.  Then add a variable
speed spindle motor and the the encoders and do a step at a time.

On Sun, Feb 12, 2023 at 1:59 PM Gene Weber via Emc-users <
emc-users@lists.sourceforge.net> wrote:

>
> Hi,
> This is also posted on the forum at:
> https://forum.linuxcnc.org/27-driver-boards/48217-mesa-7c81-questions#264314
> I wasn't sure where the best place to ask these questions was.
>
> I'm converting my new PM-728VT mill to CNC. I have a background in
> Integrated Circuits and software, but zero machining knowledge. So Linux
> and FPGA's make more sense to me than all the machining acronyms. I opted
> to go with a Raspberry Pi 4B / Mesa 7C81 setup because I've messed with
> Raspberry Pi's a bit, figured it would let me get away with running the UI
> on an older laptop, and the 7C81 was one of the few Mesa cards in stock
> these days (LOL).
>
> The OS has not yet been installed on the Rpi, nor has any bit file been
> installed on the 7C81. I would like to have all of the electronics figured
> out, and a complete plan for the software installation before I do anything.
> Kudos to the people who wrote the LinuxCNC 2.8 documentation. It's very
> good, and if I knew anything about machining I'd probably be off and
> running. I'd like the LinuxCNC software/hardware setup to support:
> - Limit and home switches on all 3 axes.
> - CNC spindle speed and direction control to coordinate feed and speed
> during milling.
> - A spindle speed encoder to coordinate low speed spindle operation and Z
> axis feed for threading operations.
> - A fourth axis for "A" rotation (Future addition).
> - Automatic flood or mist coolant control (Future addition).
>
> I'm looking at all the pin files for the 7C81 and I have some question
> regarding the modules:
> PLL - I assume a digital phase locked loop. I assume this is used
> internally in the FPGA for circuit timing and I don't need to concern
> myself with it.
> WatchDog - Timeout circuit. Sounds like this is used to ensure LinuxCNC
> and the FPGA card haven't lost contact.
> IOPort - Is my understanding correct that any pin not defined can be
> programmed as a general purpose I/O? Are these what I connect the limit and
> home switches to?
> QCount - I looked at the VHDL code an it looks like this is a timer for
> something.
> SSerial - The LinuxCNC documentation says this is for connecting to a MESA
> smart serial daughter card. So I assume I won't use it.
> StepGen - Step and direction generation for the various axes. All of the
> pin files are showing only step and direction pins. The Mesa HostMot2
> Driver Section in LinuxCNC documentation under 13.1 actually lists seven
> possible pins. I'm using ClearPath-SD servos which have Step, Direction,
> Enable, and  HLFB (High-Level Feedback). Attached is a screenshot of what
> HLFB can be used for. So I'm going to need to figure out how to access more
> than just step and direction.
> PWM - Pulse Width Modulation. I have a KBSI-240D signal isolation board
> and understand how to connect that into my spindle control. But I guess I
> need two additional boards. A PWM DAC that outputs 0-10v. (There seem to be
> two basic cards from China under many names available in several places.)
> Also need a relay board. I was thinking about  this one.
> LED - I guess for LED's but not sure of the specifics. I don't think I
> need to care.
> ENCODER - This is listed in the Mesa HostMot2 Driver Section in LinuxCNC
> documentation under 15, but is not listed as available in any of the 7C81
> pin files. That seems like a problem for me since I would like to connect
> an encoder to the spindle. I know this is a lot of direct and implied
> questions, and I really do appreciate any help. I'm trying to understand
> this in depth so I can proceed logically.
>
> Thanks & Best
> Ge

Re: [Emc-users] Robot Arm control via LCNC, Pi4 and MESA.

[Chris Albertson]

You can make really good printed timing belt pulleys.  Download the CAD
model from one of the pulley distributors and in CAD, cut big hole about 30
mm diameter.  Print the ring.Then turn a 30 mm steel disk and epoxy the
steel disk in the hole.  You get a steel hub (that will hold set screws)
and plastic teeth that would be very hard to make in metal.   Even plastic
teeth are shared then rubber belts so they last a long time.

After printing the ring, I trued it up with a boring bar. to make a perfect
fit over the steel hub.  If you try to print the entire gear, always the
hubs fail and the gear spins on the shaft.  Bt the plastic never spins on
the 30 mm disk.

On Sun, Feb 5, 2023 at 8:23 PM John Dammeyer  wrote:

> Thanks Chris.
> That's why I've been slicing the parts.  Some of them I'd just turn from
> metal.   Not worth first doing it in plastic.  But some of the others
> definitely.  Many of them are up to 9 hours printing time.  And I need to
> order bearings and belts.  This time I think I'll take a crack at making my
> own pulleys.  Mostly because to start with this is going to be a toy
> (spelling mistake... should be tool).
> John
>
>
> > -Original Message-
> > From: Chris Albertson [mailto:albertson.ch...@gmail.com]
> > Sent: February-05-23 7:25 PM
> > To: Enhanced Machine Controller (EMC)
> > Subject: Re: [Emc-users] Robot Arm control via LCNC, Pi4 and MESA.
> >
> > About the hardware.  I would print all the parts and assemble a working
> > robot, then go back and make then with metal.  If yu are going to find
> > problems it is best to do this with a material that costs only 2 cents
> per
> > gram.
> >
> > If you don't want to paint yourself into a corner with software, then
> > follow the conventions that "everyone" uses.
> >
> > The convention is to first create what is called simply enough called a
> > "robot controller".  This uses a specific kind of interface called "Joint
> > trajectory Action" and when you look inside it is a series of points
> where
> > each point has
> >
> > 1) a position for each joint (of the joint in radians or in meters
> > depending if it is a linear or rotary actuator)
> > 2) an optional velocity for each joint
> > 3) an optional acceleration for each joint
> > 4) a time (in seconds from some "start time" when the goal needs to be
> met
> > All of the above are 64-bit floating point values
> > The robot controller sends back the actual points that it was able to
> > make.
> >
> > These points can come ahead of time and might get buffered and then the
> > robot controller can do some lookahead planning or they might com one
> point
> > just in time to be executed.
> >
> > Note that the robot controller does not worry about coordinated movement.
> >  If there are 6 joints, each is controlled independently.  Coordinated
> > control is done at the next higher level.   This next higher level needs
> to
> > understand forward and reverse kinematics and and motion planning that
> > avoids collisions
> >
> > If you keep to this loose standard then you will be able to swap
> high-level
> > controllers
> >
> > The higher level controller might be LCNC if your robot is going simple
> > tasks like tool changing or loading stock onto a CNC machine.   But if
> you
> > robot is fetching beer from an unorganized refrigerator, LCNC would be a
> > very poor choice for a controller.
> >
> > I think most robots use "MoveIt."  you can see how it works here:
> > https://moveit.ros.org/
> >
> > On Sun, Feb 5, 2023 at 6:24 PM John Dammeyer 
> wrote:
> >
> > > Actually yes, the plan is to create a series of patterns and cast
> aluminum
> > > blanks and then CNC machine them.
> > >
> > > Because the mechanical drawings are there too, and I've not been able
> to
> > > use FreeCAD 0.19 to convert to STP file format,  it's easier to
> recreate
> > > the Alibre Part files from the drawings.
> > >
> > > For now what I've started is generating the G-Code for my 3D printer
> so I
> > > can get a feel for pattern sizes.
> > >
> > > But it's the software in particular that is the interesting part.  The
> > > rest is really just mechanical assembly.
> > >
> > > Like most Arduino programs it's very linear and gets away with that
> > > because of the processor speed and memory.  But, as one comment to his
> > > video stated, it's very slow and the suggestion was a change to Servos.
> > > And yes, that would probabl

Re: [Emc-users] Robot Arm control via LCNC, Pi4 and MESA.

[Chris Albertson]

About the hardware.  I would print all the parts and assemble a working
robot, then go back and make then with metal.  If yu are going to find
problems it is best to do this with a material that costs only 2 cents per
gram.

If you don't want to paint yourself into a corner with software, then
follow the conventions that "everyone" uses.

The convention is to first create what is called simply enough called a
"robot controller".  This uses a specific kind of interface called "Joint
trajectory Action" and when you look inside it is a series of points where
each point has

1) a position for each joint (of the joint in radians or in meters
depending if it is a linear or rotary actuator)
2) an optional velocity for each joint
3) an optional acceleration for each joint
4) a time (in seconds from some "start time" when the goal needs to be met
All of the above are 64-bit floating point values
The robot controller sends back the actual points that it was able to
make.

These points can come ahead of time and might get buffered and then the
robot controller can do some lookahead planning or they might com one point
just in time to be executed.

Note that the robot controller does not worry about coordinated movement.
 If there are 6 joints, each is controlled independently.  Coordinated
control is done at the next higher level.   This next higher level needs to
understand forward and reverse kinematics and and motion planning that
avoids collisions

If you keep to this loose standard then you will be able to swap high-level
controllers

The higher level controller might be LCNC if your robot is going simple
tasks like tool changing or loading stock onto a CNC machine.   But if you
robot is fetching beer from an unorganized refrigerator, LCNC would be a
very poor choice for a controller.

I think most robots use "MoveIt."  you can see how it works here:
https://moveit.ros.org/

On Sun, Feb 5, 2023 at 6:24 PM John Dammeyer  wrote:

> Actually yes, the plan is to create a series of patterns and cast aluminum
> blanks and then CNC machine them.
>
> Because the mechanical drawings are there too, and I've not been able to
> use FreeCAD 0.19 to convert to STP file format,  it's easier to recreate
> the Alibre Part files from the drawings.
>
> For now what I've started is generating the G-Code for my 3D printer so I
> can get a feel for pattern sizes.
>
> But it's the software in particular that is the interesting part.  The
> rest is really just mechanical assembly.
>
> Like most Arduino programs it's very linear and gets away with that
> because of the processor speed and memory.  But, as one comment to his
> video stated, it's very slow and the suggestion was a change to Servos.
> And yes, that would probably increase the speed.  However, the way the
> steps are generated with physical waits between steps, where nothing else
> is done, is a major limiting factor too.
>
> I think a simple 10 MHz Spi bus shift register with the image of 8 step
> pulses would make a major difference in the amount of code space used and
> time wasted.  But then a special board needs to be made.  With the extra
> I/O on the Teensy freed up it's possible to implement CAN bus and then run
> other slower I/O devices with CANopen instead of a second USB port.
>
> So I'm back to wondering about the user interface and the teaching
> capabilities and to use LinuxCNC instead of hard coded Teensy Arduino style
> programming.
>
> John
>
>
> > -Original Message-
> > From: Chris Albertson [mailto:albertson.ch...@gmail.com]
> > Sent: February-05-23 5:20 PM
> > To: Enhanced Machine Controller (EMC)
> > Subject: Re: [Emc-users] Robot Arm control via LCNC, Pi4 and MESA.
> >
> > I have been following this for a long time.  He makes the STL files
> public,
> > do what you like with then but STL files are really only good for 3D
> > printing.He makes money by selling metal CNC'd version  of the same
> > parts.So you can build a plastic version of the arm for the cost of
> the
> > plastic (and motors and bearings)
> >
> > If you want to make your own CNC'd metal version, you could reverse
> > engineer the STL files.   The only parts that need to be redrawn are the
> > bearing pockets.It is easy to reverse-engineer STL files.   I am
> doing
> > this right now for an unrelated project in another open window with
> > Fusion360.
> >
> > What software to use to control the arm depends 100% on what you want the
> > arm to do.  With robot arms, there are several classes of things to do.
> > All of the interesting things can not be pre-scripted but use sensors to
> > "see" the world that the robot interacts with.
> >
> > On Sun, Feb 5, 2023

Re: [Emc-users] Robot Arm control via LCNC, Pi4 and MESA.

[Chris Albertson]

I have been following this for a long time.  He makes the STL files public,
do what you like with then but STL files are really only good for 3D
printing.He makes money by selling metal CNC'd version  of the same
parts.So you can build a plastic version of the arm for the cost of the
plastic (and motors and bearings)

If you want to make your own CNC'd metal version, you could reverse
engineer the STL files.   The only parts that need to be redrawn are the
bearing pockets.It is easy to reverse-engineer STL files.   I am doing
this right now for an unrelated project in another open window with
Fusion360.

What software to use to control the arm depends 100% on what you want the
arm to do.  With robot arms, there are several classes of things to do.
All of the interesting things can not be pre-scripted but use sensors to
"see" the world that the robot interacts with.

On Sun, Feb 5, 2023 at 4:08 PM John Dammeyer  wrote:

> Ran into this a few days ago.
> https://www.anninrobotics.com/
>
>
> It's fully open source with the usual disclaimer that one can build it for
> fun or education but can't make them to resell.
> Yeah right…
> https://www.aliexpress.com/item/1005003688277794.html
>
> The AR3 version used a Teensy 3.5 with the trajectory calculations done on
> the PC and strings of co-ordinates sent out to the Teensy for the step/dir
> and encoder feedback.
>
> The AR4 upgraded to the Teensy 4.1 and moved all the trajectory planning
> into the module itself.
>
> Although I've registered on his website to get into the forum so far no
> word.  The videos also seem to be about 11 months old.  Maybe the
> Aliexpress version isn't pirated but just production and he's abandoned the
> support for open source.
>
> Anyway, there have been discussions that LCNC does this sort of thing
> quite well.  I'm wondering if it's worth the effort to change over.  The PC
> program for teaching and running programs is pretty impressive.  At least
> to me.
>
> Comments?
> Thanks
> John
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Linuxcnc HAL and AI

[Chris Albertson]

GPT-3 is good enough to fool many people into thinking it is actually
intelligent.   But what it is really doing is looking for matches to stored
content.So if the solution was not present in the sample text GPT-3 was
trained on, it will not be able to give you the answer.

It is not better and maybe worse than simply "Googling" your question.

GTP seems smart because it can put the looked-up answer into
grammatically correct English.

As an example of how it works you might ask it "Hello, how are you doing?"
and it will find the common words that follow this question from thousands
of times It has read "how are you" and it will give the text that typically
follows like "Fine, thank you".   The thing to remember is that GPT-3 has
absolutely zero understanding of what the question means.

So in the case here GPT picked up on the words "file", "hal", "7i92" and
gave you text that matched.  It did not write the hal file for you.

On Fri, Feb 3, 2023 at 1:09 AM  wrote:

> On 2023-02-02 17:25, Roland Jollivet wrote:
> > Since most ?? of us have some difficulty with configuring the HAL
> > script, I
> > thought maybe ChatGPT could help.
>
> Excellent thought, Roland. But to what extent can ChatGPT help? I have
> just been reading that, at the moment, it essentially fetches content
> from a 2021 snapshot of the internet.
> So I am not clear whether the answer it has given has been lifted from a
> human-generated example, or whether it can go much further in devising
> an answer without, as you say, someone giving it a further description
> of the parameters/constraints and the way in which HAL programming is
> done (rules, I guess).
> This is an interesting question, though; and I have also been reading
> about the good success in having AI systems do coding.
> I recall the early code-generating systems which appeared in the early
> 1980s, which have made great strides since then, so it makes sense that
> AI systems like ChatGPT might make the benefits available to a much
> wider user base, like LinuxCNC users.
>
> Marcus
>
>
>
>
> >
> > So this is obviously very generic, but one can easily 'add' new
> > parameters/constraints to the query. By the way, I have no clue if it
> > is
> > correct or not.
> >
> > Maybe someone could try troubleshooting their existing script by
> > pasting it
> > into AI and ask to check it.
> >
> > Another option is for a developer to write a web page interface using
> > the
> > AI API to help users generate HAL scripts.
> > _
> >
> > AI  Q: write a HAL script for Linuxcnc using steppers and a 7i92
> > interface
> > card
> >
> > AI  A: Here is an example of a HAL script for LinuxCNC using stepper
> > motors
> > and a 7i92 interface card:
> >
> > # Load the 7i92 driver
> > loadusr -W hal_7i92 -i0
> >
> > # Declare the step and direction pins for each axis
> > setp stepgen.0.step_pinhalui.0.axis.0.step
> > setp stepgen.0.dir_pin halui.0.axis.0.dir
> > setp stepgen.0.enable_pin  halui.0.axis.0.enable
> >
> > setp stepgen.1.step_pinhalui.0.axis.1.step
> > setp stepgen.1.dir_pin halui.0.axis.1.dir
> > setp stepgen.1.enable_pin  halui.0.axis.1.enable
> >
> > # Set the maximum frequency for each axis
> > setp stepgen.0.maxvel  1000
> > setp stepgen.1.maxvel  1000
> >
> > # Connect stepgen outputs to 7i92 step and direction pins
> > net step0-out stepgen.0.step=> 7i92.0.step-0-out
> > net dir0-out stepgen.0.dir  => 7i92.0.dir-0-out
> >
> > net step1-out stepgen.1.step=> 7i92.0.step-1-out
> > net dir1-out stepgen.1.dir  => 7i92.0.dir-1-out
> >
> > # Enable the stepper motors
> > setp halui.0.axis.0.enable 1
> > setp halui.0.axis.1.enable 1
> >
> > This script assumes that you have already set up the hardware
> > connections
> > between your 7i92 card and stepper motors, and configured the 7i92 for
> > use
> > with LinuxCNC. Note that you may need to modify the pin assignments and
> > other parameters based on your specific setup.
> >
> > ___
> > 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
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Running PathPilot on non-Tormach Machines

[Chris Albertson]

On Thu, Jan 26, 2023 at 6:11 PM John Dammeyer 
wrote:

> > From: Chris Albertson [mailto:albertson.ch...@gmail.com]
> >
> > On Thu, Jan 26, 2023 at 5:03 PM John Dammeyer 
> > wrote:
> >
> > >
> > > I should probably buy another high speed 16GB or 32GB MicroSD in prep
> for
> > > creating a 2.9 system that runs Path Pilot.
> > >
> >
> > Skip the SD card.  Pi4 runs faster with a USB3 SSD.This used to be an
> > expensive luxury but now if you have a spare USB3 external drive it's
> free.
>
> That's not a bad idea.  I'll look into that.  I have a USB to 2.5" hard
> drive adaptor.
>
> >
> > Another option is a server to hold the Pi4's files.  It is not as slow as
> > you might think if you use 1Gb Ethernet.  Then it is fun to do a "df" on
> > the Pi4 and see that it has 12TB of free space, but the real advantage is
> > the files live on the server that uses a versioned file system and is
> > continuously backed up.
>
> No idea if 1 GB Ethernet can be put onto a USB3.  I'm using the Ethernet
> connector on the Pi for MESA 7i92H.
>

NFS does run over WiFi.  I'm not sure of the performance.I have a Pi4
inside a robot and all the Python code to run the robot lives on the server
and is NFS mounted over WiFi but.

In your case the files wuld live on the local USB drive and you's likely
only use NFS for moving files or common things that you want on every
computer (like a ~/Downloads, /gcode and such

> >
> > Chris Albertson
> > Redondo Beach, California
>
> John
>
>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Running PathPilot on non-Tormach Machines

[Chris Albertson]

On Thu, Jan 26, 2023 at 5:03 PM John Dammeyer 
wrote:

>
> I should probably buy another high speed 16GB or 32GB MicroSD in prep for
> creating a 2.9 system that runs Path Pilot.
>

Skip the SD card.  Pi4 runs faster with a USB3 SSD.This used to be an
expensive luxury but now if you have a spare USB3 external drive it's free.

Another option is a server to hold the Pi4's files.  It is not as slow as
you might think if you use 1Gb Ethernet.  Then it is fun to do a "df" on
the Pi4 and see that it has 12TB of free space, but the real advantage is
the files live on the server that uses a versioned file system and is
continuously backed up.

-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Running PathPilot on non-Tormach Machines

[Chris Albertson]

NC.
> >
> > At work and at home he uses Windows.  And he's bought into the free
> Fusion 360 con job for CAD/CAM and is
> > now looking at how to add CNC to his mill.
> >
> > OMG.  What a hodge podge of systems.
> >
> > So he starts on the various groups looking for what is available
> including even AliExpress which has CNC boxes
> > for $300.
> >
> > Not having used CNC before and not knowing anyone local the youtube
> videos become his other source of
> > information and education.  The most interesting ones appear to be ones
> with 10's of thousands of
> > subscribers that in fact are really long mostly to have multiple
> advertisers.  But at the time he's not aware that
> > the videos are short on information, long on con for supporting lots of
> advertising.
> >
> > IMHO, the more advertisements in really long videos the more likely the
> poster is there to monetize than to
> > actually provide useful help.   How interesting to watch a milling
> cutter for 2 minutes go back and forth
> > making chips.  Oh, and now another advertisement.
> >
> > So someone on a forum recommends MESA.
> >
> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fstore.mesanet.com%2Findex.php%3F
> >
> route%3Dproduct%2Fproduct%26path%3D83_84%26product_id%3D215=05%7C01%7Ctoddz%40pgra
> > hamdunn.com
> %7C040364ad9343477cf2fb08dafe3a6977%7C5758544c573f47cebee96c3e0806fb43%7C0%
> >
> 7C0%7C638101824085494729%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiL
> >
> CJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C=QsuaNDdgukaAklsVBzraskOqkJrgDYP%2B7
> > e9BcWa62nI%3D=0
> > Hmm.  Out of stock.  And even if it were, there's no diagram like the
> ACORN one.  So no idea where to start...
> >
> > Oh but then someone suggests
> >
> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fstore.mesanet.com%2Findex.php%3F
> >
> route%3Dproduct%2Fproduct%26product_id%3D381%26search%3D7i92=05%7C01%7Ctoddz%40pgra
> > hamdunn.com
> %7C040364ad9343477cf2fb08dafe3a6977%7C5758544c573f47cebee96c3e0806fb43%7C0%
> >
> 7C0%7C638101824085494729%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiL
> >
> CJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C=lNSWiCwuqqIMsYjHAEfR9JqJAjsWNK9DeF
> > 9GsKztVJg%3D=0
> >
> > Now I need to wire to it.  A suggestion comes up for two of these and
> cheap too.
> >
> https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.aliexpress.com%2Fitem%2F33
> > 015669728.html=05%7C01%7Ctoddz%40pgrahamdunn.com
> %7C040364ad9343477cf2fb08dafe3a697
> >
> 7%7C5758544c573f47cebee96c3e0806fb43%7C0%7C0%7C638101824085494729%7CUnknown%7CTWFpb
> >
> GZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7
> > C=LclFcxiqe05h7ZMNmrFS3%2Fri1wQ7cRfBl1KFG0z51OQ%3D=0
> >
> > Next motors and power supplies.  What to buy?  People rave about these
> new step-servos.  One brand in
> > particular.  But what size?  Direct coupled or with belts and pulleys?
> >
> > So back to the internet and youtube.  Ah, this guy has a mill like
> mine.  He used the Leadshine step servos and
> > ACORN.   Ordered, installed as per the youtube video.  Making chips.
> >
> > And he doesn't have to wade through deciding which of _all_ 12 or so
> LCNC user interfaces to use.
> >
> > The above is I think the typical thought processes.
> >
> > I'm still using AXIS.  Can't be bothered to spend the time trying to
> figure out whether I should spend $400 for a
> > touch screen HDMI 21" screen or some other user interface.  Really miss
> the MACROs from MACH3 though.  I
> > think that's what's called the conversational interface on Tormach?
> >
> > John
> >
> >
> >
> >
> > _______
> > Emc-users mailing list
> > Emc-users@lists.sourceforge.net
> >
> https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistin
> > fo%2Femc-
> > users=05%7C01%7Ctoddz%40pgrahamdunn.com
> %7C040364ad9343477cf2fb08dafe3a6977%7C57585
> >
> 44c573f47cebee96c3e0806fb43%7C0%7C0%7C638101824085494729%7CUnknown%7CTWFpbGZsb3d8eyJ
> >
> WIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C=R3
> > B7SVuYgzAgLRBVbhziV0q1g1QJi0lI1vbmlsoWvsA%3D=0
> >
> >
> > ___
> > 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
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Running PathPilot on non-Tormach Machines

[Chris Albertson]

On Sun, Jan 22, 2023 at 6:23 PM Ron Ginger  wrote:

>
> I see I can simply download a copy from the Tormach site.  what
> interface board do I need?
>

I doubt Pathpilot talks directly to any interface board.  It is a user
interface and likely generates g-code.

Why not try it and tell us what happens?

-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Linuxcnc debs for RPi4

[Chris Albertson]

>
> Why would anyone care about the latest Linux version?


Perhaps the computer is used for multiple purposes and other software is
run on the computer.   I only have one PC in the shop and it is used for
both milling and 3D printing.  It also runs a web interface for remote
operation of the printers and has a couple of USB webcams and
live-streams video, so I can check if stuff is working.   I'm lucky in that
I got it all to run on the same Linux version.

The other harder case is if someone wanted to use LCNC for a
robot controller.   You can think of a robot as a (maybe) 6-axis or 14-axis
machine where all the motors need to run in synchronized motion, and you
might try to use LCNC.   I think this is a reasonable idea for a large
industrial factory automation robot.In this case, there is a ton of
other software involved for things like real-time path planning and 3D
computer vision and whatnot, and it all needs to run on the same version of
Linux.  Actually, some people are doing this (with machine kit, I think)

After you put multiple software on the same PC and then get it to work, the
real problem is if the "latest" version of one part requires the latest
Linux version, and you need to upgrade to fix a bug.   The solution is
"docker" or virtual machines.  You end up running multiple copies of
Linux.   This is inefficient, but sometimes the only way.
-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Resolvers and 2 phase AC servos

[Chris Albertson]

lar mounts, your latitude
> above or below the equator determines that, and a large longitude.
>
> I have an old friend from back in the 70's that at that time was
> building a 12 bay steerable 2 meter get up. 16 element yagi's in a
> broadside array. But another job opening took me to NM from NE so I lost
> track of him. Employment brought me to WV 38 years ago, and I'm now 21
> years retired, and still here.
>
> And tell us how you did it when it working.
>
> 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, 1940)
> If we desire respect for the law, we must first make the law respectable.
>   - Louis D. Brandeis
> Genes Web page <http://geneslinuxbox.net:6309/>
>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Resolvers and 2 phase AC servos

[Chris Albertson]

This is a job best done on a $5 microcontroller, like the ESP32.   You can
even program the thing in Python to do what is needed.   You can generate
the AC as a 3-volt signal to an analog pin and then amplify it.

Microcontrollers have come a long way.  ESP32 is a dual-core 32-bit CPU
with hardware floating point, and yes the entire thing sells for $5 and it
can self-host a python system.  All you need to program it is a terminal
window on a PC.

Yes it is fast enough.  On my desk now I an using one to commutate the
coils on a BLDC motor using FOC control.  It computes the sinusoidal
voltage that goes the each of the three phases to keep the magnetic field
pointing the right way.   Quite a lot of calculations this is not simple
6-step algorithm.

Here is an demo-project.   An ESP32 is used to compute the voltage on each
of the three phases of a small motor so as to keep the ball on top of the
wheel.There is a second microcontroller with a built-in camera that
tells the ESP32 how far the ball is from the center point.   Again a $5
computer is doing this and if you look there is not 100 lines of code in
the loop.   The control is dramatically better then you can do with a Linux
PC and 10x simpler  and 100x cheaper.
https://youtu.be/fXxd8guAY7g

Tracking a green ball is a servo system is a lot like tracing the Moon,
except the Moon is easier to track








On Sat, Jan 21, 2023 at 3:26 AM Robin Szemeti via Emc-users <
emc-users@lists.sourceforge.net> wrote:

> So,  in a vaguely CNC related folly I have purchased a 2 axis military dish
> mount for radio stuff (moonbounce, if you must know)
>
> My plan is to control it from EMC, with a Mesa anything io card and the
> resolver interface, I already have various Mesa cards under EMC, so thats
> not an issue and I have a spare 5i23 PCI card and the isolated IO duaghter
> card, so with the resolver card I should be able to uses the resolvers, and
> pick up the limit switches on the IO card. I can just use some Python
> script to work out the moon position and generate "G1 X285.4, Y23.8" or
> whatever MDI motion commands every 15 seconds or so to track the moon,  not
> a problem.
>
> Now, the question is driving the two phase AC servos.   Generating vast
> quantities of 50 or 400Hz or whatever AC for the drives is simple enough, I
> can just use a couple of large Class D amplifiers, no worries. Generating
> the AC low level signals to drive these  basically, it needs a constant
> 50Hz signal on one coil of a fixed level, and a variable level sine wave on
> the other coil, at either +90 or -90 degrees depending on which way it
> needs to go.
>
> Is it possible to use HAL or something to generate the variable level 50Hz
> servo drive signals and output them from the PWM outputs on the resolver
> card?
>
> --
>
> Best regards,
>
> Robin Szemeti
>
> Redpoint Consulting Limited
>
> E: ro...@redpoint.org.uk
> T: +44 (0) 1299 405028
> M: +44 (0) 7971 883371
>
> CONFIDENTIALITY NOTICE
> The information contained in this e-mail is intended only for the
> confidential use of the above named recipient. If you are not the
> intended recipient or person responsible for delivering it to the
> intended recipient, you have received this communication in error
> and must not distribute or copy it.
> Please accept the sender's apologies, notify the sender immediately
> by return e-mail and delete this communication.
>
> Thank you.
>
> _______
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Available fanless PC for linuxcnc with parallel port

[Chris Albertson]

On Thu, Jan 19, 2023 at 11:20 AM Linden via Emc-users <
emc-users@lists.sourceforge.net> wrote:

>
> Any idea when the Ethernet cards will be produced and sold again??
>

I just put two of them in my shopping card, only one gave the "out of stock
notice"   One had the female pins, the other male pins, I forget which was
in stock.


It SHOULD be a lot easier to make a hardware motion controller.   Right now
I'm working on the design of a pulse generator that makes 16 PWM outputs
for a project unrelated to LCNC.But really it is the same thing,  a
bigger PC sends commands over a fast serial link (I'm using 1,000,000 bits
per second) then the command is interpreted and a pulse generator is set
up.   The pulses are hardware generated by diving down a crystal
oscillator.

So far I have $4 worth of hardware doing this and I was able to writ the
firmware in MicroPython.   The pulses are don in hardware so I don't need
the code to be fast.  I'm using an RP2040 chip.  The chip is a
ddual-core 32-bit M0 ARM processor and sells for $1.   There is just a
-load of peripheral hardware on this chip and again, the price is $1.
 But no one is motivated because it is easier to spend $110 and buy an FPGA
based one.

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Available fanless PC for linuxcnc with parallel port

[Chris Albertson]

The problem with buying a computer is that it not only needs a parallel
port, but it needs to have useable low latency.  You can not test latency
until AFTER you buy it.

I own several old PCs and have a parallel port card, none of them give good
scores on the latency test.  They are good enough to run the motors at low
step rates but not optimally.

On Thu, Jan 19, 2023 at 9:13 AM John Dammeyer 
wrote:

> https://www.neweggbusiness.com/product/product.aspx?item=9siv04gd8p7931
> Not sure why the parallel port doesn't show up in the description but is
> in the photo.
>
> > -Original Message-
> > From: andy pugh [mailto:bodge...@gmail.com]
> > Sent: January-19-23 4:23 AM
> > To: Enhanced Machine Controller (EMC)
> > Subject: Re: [Emc-users] Available fanless PC for linuxcnc with parallel
> port
> >
> > On Thu, 19 Jan 2023 at 09:51, Robin Szemeti via Emc-users
> >  wrote:
> > >
> > > I'll be surprised if you can find anythign with a parallel port,
> >
> > That's just not the case. (though many don't have them in the case)
> >
> > With parallel port:
> >
> https://www.mini-itx.com/store/category?type=motherboard=4GB-or-more=from-
> >
> 1=from-1=from-0=1-or-more
> > external=1=price=1
> >
> > With Parallel port header:
> >
> https://www.mini-itx.com/store/category?type=motherboard=4GB-or-more=from-
> >
> 1=from-1=from-0=1-or-more
> > internal=1=price=1
> >
> > (No help if you are in the US, this is a UK stockist)
> >
> >
> >
> > --
> > 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, 1912
> >
> >
> > ___
> > 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
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Available fanless PC for linuxcnc with parallel port

[Chris Albertson]

This card seems to be in-stock at the Mesa website.   I'd go with Ethernet
because it gives you the option later to use computers with no PCIe bus.
Maybe a Rapberry Pi or something like that.  Also the Ethernet cable gives
you flexibility as it can be up to 100 meters long.
 7I92TH Anything I/O Ethernet card
<http://store.mesanet.com/index.php?route=product/product_id=381>

On Thu, Jan 19, 2023 at 9:16 AM Chris Albertson 
wrote:

>
>
> On Thu, Jan 19, 2023 at 4:26 AM andy pugh  wrote:
>
>> On Thu, 19 Jan 2023 at 09:51, Robin Szemeti via Emc-users
>>  wrote:
>> >
>> > I'll be surprised if you can find anythign with a parallel port,
>>
>> That's just not the case. (though many don't have them in the case)
>>
>> With parallel port:
>>
>> https://www.mini-itx.com/store/category?type=motherboard=4GB-or-more=from-1=from-1=from-0=1-or-more=1=price=1
>>
>> With Parallel port header:
>>
>> https://www.mini-itx.com/store/category?type=motherboard=4GB-or-more=from-1=from-1=from-0=1-or-more=1=price=1
>>
>> (No help if you are in the US, this is a UK stockist)
>>
>>
>>
>> --
>> 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, 1912
>>
>>
>> _______
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>>
>
>
> --
>
> Chris Albertson
> Redondo Beach, California
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Available fanless PC for linuxcnc with parallel port

[Chris Albertson]

On Thu, Jan 19, 2023 at 4:26 AM andy pugh  wrote:

> On Thu, 19 Jan 2023 at 09:51, Robin Szemeti via Emc-users
>  wrote:
> >
> > I'll be surprised if you can find anythign with a parallel port,
>
> That's just not the case. (though many don't have them in the case)
>
> With parallel port:
>
> https://www.mini-itx.com/store/category?type=motherboard=4GB-or-more=from-1=from-1=from-0=1-or-more=1=price=1
>
> With Parallel port header:
>
> https://www.mini-itx.com/store/category?type=motherboard=4GB-or-more=from-1=from-1=from-0=1-or-more=1=price=1
>
> (No help if you are in the US, this is a UK stockist)
>
>
>
> --
> 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, 1912
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Available fanless PC for linuxcnc with parallel port

[Chris Albertson]

My go to place is Newegg.com  They sell all kinds of off lease used PCs but
getting one known to work with a parallel printer port is getting to be
hard.That is fine as Mesa cards are reasonably affordable.

I just checked and they have dozens of options for $80.  Typically an Intel
i3 with 4GB RAM for $80. If your budget is $300 then you can add a Mesa
fpga-based parallel port that works with your existing breakout and still
have money left over.

What I do before I buy one of these used systems is download the user
manual from the manufacturer.  HP and Dell and very good about keeping 10
years old manuals still available online.

On Wed, Jan 18, 2023 at 7:45 PM Linden via Emc-users <
emc-users@lists.sourceforge.net> wrote:

> Does any one have suggestions for a reasonably priced and available fan
> less pc with a parallel port that I can use with linux cnc to go with my
> gecko 540?
>
> Looking to spend under $300 if possible with shipping
>
> My original old dell gave up and died several years ago. The  pci 1
> parallel port card I have is not recognized by the bios of either of my
> other 2 computers and I cant find a header cable in my junk to go from
> the lp1 pins on my Asus mother board to 25 pin sub D. Be nice to have
> something off the shelf known to work as I don't have access to piles of
> old PCs any more and I would like to have my plasma table running again.
>
> Any suggestions are much appreciated thanks in advance,
>
>   Linden
>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Is this LCNC based?

[Chris Albertson]

Many times Google translations from technical Chinese make no sense.
 A simple example I see many times is "pressure" used to mean volts.  This
is a sign that Googel Translate was used.   A bilingual human with an
education in electronics would know to say "volts"  But Google does not
know which interpretation of the character to use.  The problem is that
can't invent new characters, so the same written characters have several
different meanings and you have to know the context to sort it out.

My wife has 8th grade education in Chinese.  Enough to read newspapers,
watch TV and read most fiction.  But if I ask her to look at some data
sheet that is in Chinese, she says "it does not make sense"   It is written
in a way that engineers have to learn.

I think this is one of those cases.  The "circular soft algorithm" is
likely a really bad translation. And it might mean "loop" or "cycle" for
all we know.

They say it runs on Linux.   I doubt anyone would write a CNC controller
from scratch.   What else might it be based on?

BTW, the company also sells a real clean looking "A + B" smiling axis that
has two rotary joints with 50:1 harmonic reduction and stepper motors.
 Who knows how well it works but it looks nice.   Price is too high for me,
I'd rather build one.This is a nice design to copy.  See how the 50:
reduction is integrated with the frame and to bolted to the frame.
amazon.com/Dividing-Rotational-Indexing-Harmonic-Reducing/dp/B09WLCTLCX
<https://www.amazon.com/Dividing-Rotational-Indexing-Harmonic-Reducing/dp/B09WLCTLCX/ref=sr_1_9?keywords=cnc+a+b+axis=167303=8-9>






On Mon, Jan 9, 2023 at 8:08 PM Thaddeus Waldner  wrote:

> 8) Interpolation Algorithm: S type, circular hard algorithem, circular
> soft algorithm:
>
> I wonder what is meant by interpolation algorithm.
>
> > On Jan 9, 2023, at 9:52 PM, Chris Albertson 
> wrote:
> >
> > Amazon has a turn-keyCNC controller for $480.  This includes the whole
> kit,
> > computer, FPGA, screen BoB and all except for power supply and the
> motors.
> >
> >
> > Question:  Is this based on LCNC?   Seem like it might be   Below is a
> > quote from their web page, I added the bold face.
> >
> >>
> >>   - The DDCS Expert is a *5 axis motion controller for open or close
> >>   stepper and servo systems* with 7" full color display screen. The
> >>   highest output pulse per axis is 1MHz. The users can self-define the
> >>   functional keys. This controller supports multiple spindle mode,
> support
> >>   straight Tool Magazine, gantry type Magazine, disk type magazine. The
> >>   Operation system interface even though very comprehensive, can be
> learned
> >>   in very short time.
> >>
> >>
> >>   - The DDCS Expert numerical control system *adopts the ARM+FPGA design
> >>   framework*. ARM controls the human-computer interface and code
> >>   analysis and the FPGA provides the underlying algorithms and creates
> the
> >>   control pulse. This guarantees reliable control and easy operation.
> >>
> >>
> >>   - *The internal operating system is Linux based.*
> >>
> >>
> > Product:
> amazon.com/Axis-Controller-Handwheel-MPG-Emergency/dp/B09WLDKW53
> > <
> https://www.amazon.com/Axis-Controller-Handwheel-MPG-Emergency/dp/B09WLDKW53?ref_=ast_sto_dp=1
> >
> > User Manual:
> > /nvcnc.net/pdf/DDCS%20Expert%20User's%20Manual%20V1%20(Part1)-nvcnc.pdf
> > <
> https://nvcnc.net/pdf/DDCS%20Expert%20User's%20Manual%20V1%20(Part1)-nvcnc.pdf
> >
> >
> >
> >
> > --
> >
> > Chris Albertson
> > Redondo Beach, California
> >
> > ___
> > 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
>


-- 

Chris Albertson
Redondo Beach, California

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

[Emc-users] Is this LCNC based?

[Chris Albertson]

Amazon has a turn-keyCNC controller for $480.  This includes the whole kit,
computer, FPGA, screen BoB and all except for power supply and the motors.


Question:  Is this based on LCNC?   Seem like it might be   Below is a
quote from their web page, I added the bold face.

>
>- The DDCS Expert is a *5 axis motion controller for open or close
>stepper and servo systems* with 7" full color display screen. The
>highest output pulse per axis is 1MHz. The users can self-define the
>functional keys. This controller supports multiple spindle mode, support
>straight Tool Magazine, gantry type Magazine, disk type magazine. The
>Operation system interface even though very comprehensive, can be learned
>in very short time.
>
>
>- The DDCS Expert numerical control system *adopts the ARM+FPGA design
>framework*. ARM controls the human-computer interface and code
>analysis and the FPGA provides the underlying algorithms and creates the
>control pulse. This guarantees reliable control and easy operation.
>
>
>- *The internal operating system is Linux based.*
>
>
Product: amazon.com/Axis-Controller-Handwheel-MPG-Emergency/dp/B09WLDKW53
<https://www.amazon.com/Axis-Controller-Handwheel-MPG-Emergency/dp/B09WLDKW53?ref_=ast_sto_dp=1>
User Manual:
/nvcnc.net/pdf/DDCS%20Expert%20User's%20Manual%20V1%20(Part1)-nvcnc.pdf
<https://nvcnc.net/pdf/DDCS%20Expert%20User's%20Manual%20V1%20(Part1)-nvcnc.pdf>



-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Sharpening teeny drill bits

[Chris Albertson]

If you are drilling less than 100 holes, I doubt the bit will get dull.
I buy the smaller 3mm and down bits from Chinese suppliers in bulk in
ziplock bags.  They are so cheap there is no point wasting time sharpening
them.The cost works out to 16 cents per bit.

I think this is why the sharpeners only go down to a minimum size, it is
not worth it for a 16 cent bit. Yes they are good quality

One example ebay.com/itm/134264963420
<https://www.ebay.com/itm/134264963420?hash=item1f42d0c15c:g:XMQAAOSwYcNjPOBi=enc%3AAQAH4KkQDN3Jz5CU5LGL9RDC7lKBFQlXdWnAYLweheWZh1jINk0kYnH%2FMPteG5FoxxG2XCdJRZV%2BozxczP5WigD25pfrjuoLk68FbFDNy%2Bdj51fWjd%2B%2FQgIskWUhiBUlHnHYnArgvvU4RjSWRet%2Bi7aNy3EbMj2xjXZIbzBJDkMUbXI7UGCim5MqWxrW9wo1CDhmlRU9ABpKLWdZ0YF2TA%2FVqwAyroKlxplqpTKZi2AYuQGqCd4HWOpVmUcZgoL2NzSh4euK4KbkIuaV%2B5OYDK6QgmiFe1VSOeeQpw8Dm%2FWiu%2Byt%7Ctkp%3ABFBMnK_88qRh>

On Sun, Dec 18, 2022 at 9:54 AM gene heskett  wrote:

> Greetings all;
>
> As many of you know a drill Dr is the bees knees, for drill bits above 6
> or so mm, but a disaster for smaller bits due to the tolerances of the
> stamped and plastic chuck giving you lop sided grinding. AKA BIGGER holes.
>
> So, what jig or machine do you all use to sharpen drill bits in the
> below 5mm categories? Or should I gently spiral down a 1/16" SC mill,
> but that runs out of DOC long before it reaches thru a 20mm square CF
> tube. I've about 35 such holes to drill in 1mm thick wall CF, 8 of which
> need to go all the way thru both sides of the tube. Those will all be
> thread clearance for 3mm screws, but 8 of them will need to be for a
> 3mmx.5 tap, in 4.1mm thick alu. CF will no doubt be hell on drill bits,
> so it (or they) will need resharpened frequently I expect.
>
> Thank you, stay warm and well 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, 1940)
> If we desire respect for the law, we must first make the law respectable.
>   - Louis D. Brandeis
> Genes Web page <http://geneslinuxbox.net:6309/>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Another of my silly? questions

[Chris Albertson]

I had only eyeballed the equations, not plugging in numbers and figured it
was a toss-up.  Put you can always modify the tube and turn it into a
solid.   Simply fill it with chopped carbon or glass fiber and epoxy paste,
pack it in and let it cure.   I've made parts with before and they are very
stiff and strong.  The random fiber orientation is not as bad as you might
think and even short chopped fiber still has 100+ length to diameter
ratio.  The key is is mix in as much fiber was physically possible while
still getting it all to wet-out.

This is a quick way to make parts, faster and stronger than 3D printing.
Model makers have been doing this for decades.  I was making small parts
for flat-water racing kayaks with fiber-filled resin and latex molds

So unless the tube needs to have light weight, just fill it and make it a
solid.  Or use a larger aluminum part

On Mon, Dec 5, 2022 at 4:03 AM andy pugh  wrote:

> On Mon, 5 Dec 2022 at 02:20, Chris Albertson 
> wrote:
>
> > and we don't even know the exact kind of steel in the rod.  Without
> knowing more it is a toss-up.
>
> You don't need to know, all iron alloys have the same stiffness (to within
> a couple of percent).
> The shear modulus (G) for steel is 80 and for carbon fibre is around 5.
> Steel is a lot stiffer than CF.
>
> Anyway, with a 2Nm motor you might see 1.7 degrees of twist in an 8mm steel
> shaft which is 500mm long.
> Playing with numbers it would be 3 x less with a steel tube 12mm dia and
> 1mm wall thickness.
> To match the stiffness of the steel bar with a _round_ carbon tube it would
> need to be 20mm dia and 1.5mm wall thickness.
>
> https://amesweb.info/Torsion/torsion-of-shaft-calculator.aspx
>
> Note that there is only any torsional load on the shaft when accelerating,
> so the extruder will generally not deviate from the path very much, but
> will be lagging along the correct path.
> (This is true of milling machines too, the f-error is typically
> approximately along the path, so has less effect on part geometry than the
> raw numbers would indicate)
>
> --
> 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, 1912
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Another of my silly? questions

[Chris Albertson]

On Sun, Dec 4, 2022 at 4:37 PM gene heskett  wrote:

>
> It seems to me that 19" of 20x20 square tunbing would have less twist
> per NM of applied torque than the same length of 8mm steel shaft, but
> IDK.  Hence the question.


This is a tough question.  You are in effect asking to compare the
torsional stiffness of two very different beams.

It can be calculated, but you don't want to see the math involved (Google
"second area of moment" if you do want to see.)

Of the two shapes, a thin wall square is dramatically worse than a solid
round.  But in the calculation, we raise the diameter to the 4th power, so
a larger size is very much favored.   Do they balance?We don't know the
mechanical properties of your epoxy/carbon composite tube and we don't even
know the exact kind of steel in the rod.  Without knowing more it is a
toss-up.

I'd say measure it.  Wrap string around each end in opposite directions and
hang weights on the string.  Or place a torque wrench on both and measure.


> --

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] hair brained idea?

[Chris Albertson]

a huge printer is not as usfull as you might think.  The nozzle is still
0.4 mm in diameter or close.  Print time is the cube of the linear
dimension of the parts you make,  Can you wait a week while the printer
runs.  What if a 1Kg spool of filament is not enough?

I got an idea, see what Cura says.  Define a printer that large and slice
something and look at the weight and print time.

A more practical idea is to print the parts in sections and then glue or
screw the parts together.  It helps if the parts have some kind of self
alignment or keyway.

or huge parts, casting is better.   Print a mold, then fill it with plastic
resin that is mixed with chopped fiber.It makes really strong parts.
Then disassemble the mold and take the part out.

A big printer is no different at all from a milling machine, except that it
moves faster and does not need to be build as rigid as there is not cutting
force.So why not look up some CNC router projects?   It is a common
LinuxCNCproject.

I would still use those closed loop steppers.   With a week-long print, you
do not want skipped steps.

On Sat, Dec 3, 2022 at 5:00 PM gene heskett  wrote:

> On 12/3/22 18:35, Chris Albertson wrote:
> > I don't know what your big-picture goal is,
>
> A 500mm Ender5+ and an 800+mm tronxy-400-pro, and maybe a 1000mm
> Saphire-5 plus eventually.
>
> but you can drive a normal
> > two-phase stepper in continuous, non-stepping mode.  In this mode they
> are
> > just like servomotors.  Then they attach a rotary encoder.  So
> > step-skipping is not going to happen as they don't step.
> >
> > Stepperonline sells these all setup and ready to go as drop-in
> replacements
> > for normal steppers.  The servo-loop is closed inside the driver
> > electronics so the computer just sees the usual sep/dir interface.
> >
> I am fam with that, I'm already using 4 of the nema 23 3 phase versions
> which are lots more efficient power wise, but didn't know it was
> available in nema 17's. I've asked for a quote for the bigger one in
> dual shaft.  Thank you, Chris. Take care & stay well.
>
> 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, 1940)
> If we desire respect for the law, we must first make the law respectable.
>   - Louis D. Brandeis
> Genes Web page <http://geneslinuxbox.net:6309/>
>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] hair brained idea?

[Chris Albertson]

I don't know what your big-picture goal is, but you can drive a normal
two-phase stepper in continuous, non-stepping mode.  In this mode they are
just like servomotors.  Then they attach a rotary encoder.  So
step-skipping is not going to happen as they don't step.

Stepperonline sells these all setup and ready to go as drop-in replacements
for normal steppers.  The servo-loop is closed inside the driver
electronics so the computer just sees the usual sep/dir interface.

On Sat, Dec 3, 2022 at 2:02 PM gene heskett  wrote:

> Greetings to all the other stepper experts here.
>
> I got the double ended motors for an Ender5+ today.
>
> Other than the paint job and wiring, leads instead of sockets, they look
> identical to the motor currently on the Ender5+ for Y drive. No
> perceived improvement.
>
> I can probably program klipper to drive then both with a separate 5160
> driver per motor, get another bracket, mount then butt to butt with one
> coil reversed, locked together with a 5mm to 5mm coupling,to maintain X
> squareness making them into one motor with twice the torque.
>
> The big IF is whether the drivers always power up at the same state
> before any move is commanded, I don't want x square to be a random
> driver power up fight.
>
> Its either make that work, or figure out some sort of a backlashless
> bevel gear from one of the 3 phase motors that are all nema-23's so far.
> That would be ideal if the fault line could be used to freeze things if
> it ever loses home by even one step. Something the 3 phase nema 23's can
> do automatically.
>
> Best antibacklash is not a bevel gear, but likely a jackshaft from the
> nema 23 to a short shaft where the nem-17 motor is now. Even a 1NM 3
> phase would be 200% of the torque of whats in there now.
>
> Which way, two nema-17 motors, or a nema 23 and a jackshaft with 30
> tooth pulleys, would you guys opt for?
>
> Thanks for your thoughts!
>
> 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, 1940)
> If we desire respect for the law, we must first make the law respectable.
>   - Louis D. Brandeis
> Genes Web page <http://geneslinuxbox.net:6309/>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] ? cutting a square carbon fiber tube ?

[Chris Albertson]

I bet you could print a better tool than this.
homedepot.com/p/Stanley-Miter-Box
<https://www.homedepot.com/p/Stanley-Miter-Box-STHT20360/311276301>
Use a fine-tooth hack saw.   It is easier to cut a fiber tube than a
same-size hardwood dowel or same-size steel tube.



On Thu, Dec 1, 2022 at 5:38 AM gene heskett  wrote:

> On 12/1/22 00:47, Chris Albertson wrote:
> > Cutting square tubing is no harder than cutting molding around a door
> > frame.   Use a miter saw.   A fine tooth hacksaw blade works well for
> > carbon fiber.
> >
>
> My mitre saw is a 12" Bosch chop saw, the fawncy high $ articulated one.
> And I've a new carbide blade. New & carbide=dull.
>
> > If you don't have a miter saw, print a plastic guide block that holds the
> > tube and has a slot for your hacksaw.
>
> I have considered that too, but in the form of an 18x18mm stuffer that's
> sacrificial.
> >
> > Finish with a sanding block to make the edge nice.   I've cut a lot of
> > fiber tubes, sharp saws work well. but glass and other fibers dull tools
> > really fast.
> >
> > You can make a really strong part by gluing fiber tube into 3D printed
> > blocks
> >
>
> That is essentially what I have in mind. But moving the tube sideways so
> one run of the X belt runs inside it. Shorten it maybe 10 mm, and mill a
> couple slots in the other ends idler so it remains adjustable.
> Shortening it a few mm will lengthen the belt and narrow the carriage,
> gaining back some X room lost by the basic design and even more lost by
> my 4 wheel carriage design.
>
> Somewhere along the line, move the x home sw to the left end where they
> were too cheap to use 6" more wire to reach it. That will require
> rebuilding marlin or klipper. Shrug. Turn the whole axis end for end and
> put it in some cable chain I already have. Neat but flying weight for
> the Y motor to throw around & its in trouble already on an Ender5plus.
> It loses Y home, I suspect because the way undersized psu is folding
> back when using gyroid infill at more than 60 for speed.. Everything
> else it can run at 200 or more. When I get done playing, the bed heat
> and motor load will be removed from that supply, bed heat will be 62VAC,
> and motor power will be a 650 watt 48 volt supply.
>
> Take care & stay well.
>
> 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, 1940)
> If we desire respect for the law, we must first make the law respectable.
>   - Louis D. Brandeis
> Genes Web page <http://geneslinuxbox.net:6309/>
>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] ? cutting a square carbon fiber tube ?

[Chris Albertson]

Cutting square tubing is no harder than cutting molding around a door
frame.   Use a miter saw.   A fine tooth hacksaw blade works well for
carbon fiber.

If you don't have a miter saw, print a plastic guide block that holds the
tube and has a slot for your hacksaw.

Finish with a sanding block to make the edge nice.   I've cut a lot of
fiber tubes, sharp saws work well. but glass and other fibers dull tools
really fast.

You can make a really strong part by gluing fiber tube into 3D printed
blocks

On Wed, Nov 30, 2022 at 8:08 PM gene heskett  wrote:

> Greetings all;
>
> An unusual question. How to cut, neatly, and squarely, a 20mm square
> carbon fiber tube to length, say 550mm out of an 800mm tube. I'm making
> a linear rail bearing X axis for two of my 3d printers,
>
> I have a die grinder that I can mount in a quick change holder, and a 2"
> by 30 thou CBN disk and 1/4" arbor, and a 4 jaw chuck on my Sheldon.
>
> Do I need to print an 18mm square plug to insert into this tubing to
> support the 1mm thick tubing walls, or can a relatively light grip be
> enough to hold it fixed while I turn slow and cut it off with the CBN
> wheel running at about half line voltage in the die grinder?
>
> All I can see in my minds eye is a hugely ragged edge cutting it by hand
> with a hacksaw.
>
> Better ideas anybody?
>
> Thanks All.
>
> 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, 1940)
> If we desire respect for the law, we must first make the law respectable.
>   - Louis D. Brandeis
> Genes Web page <http://geneslinuxbox.net:6309/>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] XHC HB04 ..

[Chris Albertson]

Ali Express seems to have many of them all just under $70.  They are
wireless and come with a dongle for a USB port.

On Fri, Nov 25, 2022 at 3:00 PM Robin Szemeti via Emc-users <
emc-users@lists.sourceforge.net> wrote:

> Annoyingly .. it seems you can no longer get the XHC HBO4 pendant with the
> single XYZA/sindle/feed knob ... all eBay has is the new ones with the two
> knobs  that lack the =1/2 and =0 buttons .. which I use a lot.
>
> I'm getting bored moving the handwheel between machines, any idea where to
> get one of the "old" ones from?
>
> --
>
> Best regards,
>
> Robin Szemeti
>
> Redpoint Consulting Limited
>
> E: ro...@redpoint.org.uk
> T: +44 (0) 1299 405028
> M: +44 (0) 7971 883371
>
> CONFIDENTIALITY NOTICE
> The information contained in this e-mail is intended only for the
> confidential use of the above named recipient. If you are not the
> intended recipient or person responsible for delivering it to the
> intended recipient, you have received this communication in error
> and must not distribute or copy it.
> Please accept the sender's apologies, notify the sender immediately
> by return e-mail and delete this communication.
>
> Thank you.
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Production machine running linuxcnc.

[Chris Albertson]

With video production quality, it is easy to move from "amateur-looking" to
"pretty good" without spending a fortune.  But then after that, each small
increment practically doubles the cost and effort.





On Wed, Nov 23, 2022 at 8:54 AM andy pugh  wrote:

> On Wed, 23 Nov 2022 at 01:07, Chris Albertson 
> wrote:
>
> > But as said, just a trip gets you 85% of the way to being
> >  professional looking,
>
> My own YouTube videos got a lot better once I bought a cheap tripod
> (and a real video camera, too)
>
> Still not even comparable to  Clickspring standard, or any other
> popular machining channel really, but at least watchable quality.
>
> --
> 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, 1912
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Production machine running linuxcnc.

[Chris Albertson]

I said "Buy a tripod" but this CNC company used motorized dolly shots
which, is basically CNC applied to videography.

This video should make it completely obvious to everyone here how it
works. https://youtu.be/IBZoRFJd7Cw?t=561

But as said, just a trip gets you 85% of the way to being
 professional looking,

The hard part to designing them is not getting the motion right, It's
making them silent.



On Tue, Nov 22, 2022 at 3:02 PM BRIAN GLACKIN 
wrote:

> For some high production value machining video.  Check out “Titans of CNC
> Machining” on YouTube.  These are very high impact videos that “advertise”
> the capabilities.  I think these utilize all the points Chris makes and
> then go well beyond.
>
> Not suggesting you have to go to thier level but offering additional
> inspiration.
>
>
> On Tue, Nov 22, 2022 at 2:01 PM Chris Albertson  >
> wrote:
>
> > The content looks good but if the goal is to advertise, then you need to
> > "up" the production values considerably.
> > This is more than good enough to show us, but "fair" video production
> > quality will put the company in the wrong light.
> > So in order is importance:
> >
> > 1) *BUY A TRIPOD.*  This is an absolute "must".
> > 2) Your video has what we call "artifacts".  It is VERY compressed,
> perhaps
> > you can change the setting. The compression artifacts are noticeable and
> do
> > not give the "clean high precision" "look" you need.   If it means
> spending
> > $350 on a good used camera, do it.  This work needs to be shot is a
> MINIMUM
> > of 1080p 60 FPS with the highest bit rate you can do.
> > 3) Possibly use some "EQ" on the audio. taking off some of the higher
> pitch
> > noise is unrealistic, but more pleasant to listen to.
> >
> > Yes, video is an entirely different skill set than machining.  The
> > current video is very good to show other machinists what you can do,
> > but advertising is a whole different ball game and does not depend on
> > logic,   You would think a potential customer would say "this guy is
> > obviously an expert CNC machinist, but only a fair videographer."  No,
> they
> > don't think like that, they "experience" the video and remember the
> > experience.
> >
> >
> >
> > On Tue, Nov 22, 2022 at 10:30 AM andrew beck 
> > wrote:
> >
> > > Hey guys I made this video last night of my machine running on a part
> > > thought you might like it.
> > >
> > > Running linuxcnc and preloading tool for toolchange
> > >
> > > Plus 4th axis etc
> > >
> > > I'm hoping to start making a lot of these to both advertise for my
> > machine
> > > shop and to show how good Linux CNC is.
> > >
> > > And that you can run big machines on it.
> > >
> > > https://youtu.be/kzaxBU0EVr0
> > >
> > > ___
> > > Emc-users mailing list
> > > Emc-users@lists.sourceforge.net
> > > https://lists.sourceforge.net/lists/listinfo/emc-users
> > >
> >
> >
> > --
> >
> > Chris Albertson
> > Redondo Beach, California
> >
> > ___
> > 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
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Production machine running linuxcnc.

[Chris Albertson]

The content looks good but if the goal is to advertise, then you need to
"up" the production values considerably.
This is more than good enough to show us, but "fair" video production
quality will put the company in the wrong light.
So in order is importance:

1) *BUY A TRIPOD.*  This is an absolute "must".
2) Your video has what we call "artifacts".  It is VERY compressed, perhaps
you can change the setting. The compression artifacts are noticeable and do
not give the "clean high precision" "look" you need.   If it means spending
$350 on a good used camera, do it.  This work needs to be shot is a MINIMUM
of 1080p 60 FPS with the highest bit rate you can do.
3) Possibly use some "EQ" on the audio. taking off some of the higher pitch
noise is unrealistic, but more pleasant to listen to.

Yes, video is an entirely different skill set than machining.  The
current video is very good to show other machinists what you can do,
but advertising is a whole different ball game and does not depend on
logic,   You would think a potential customer would say "this guy is
obviously an expert CNC machinist, but only a fair videographer."  No, they
don't think like that, they "experience" the video and remember the
experience.



On Tue, Nov 22, 2022 at 10:30 AM andrew beck 
wrote:

> Hey guys I made this video last night of my machine running on a part
> thought you might like it.
>
> Running linuxcnc and preloading tool for toolchange
>
> Plus 4th axis etc
>
> I'm hoping to start making a lot of these to both advertise for my machine
> shop and to show how good Linux CNC is.
>
> And that you can run big machines on it.
>
> https://youtu.be/kzaxBU0EVr0
>
> _______
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Recommended GUI dev environ?

[Chris Albertson]

The PiSim[le GUI library is built on top of TKinter and two other
libraries.  You get to choose.  One of them is web based.It is nice to
be able to spin off web-based version and control my gadgets with a cell
phone.  But the tk version has the most functionality at present.

Like I said, this is best used for doing fast work.   You can be done very
fast.In development many times at the beginning we want to try ideas
then over time settle on one design idea.  Then it is time to re-code it
all and make it fast and neat.  For prototyping and one-off work the simple
thing is good.

Also they do give you options for space and size.Maybe not as much as
you would like but you can get down inside the details if you like.



On Sun, Nov 6, 2022 at 3:20 AM gene heskett  wrote:

> On 11/5/22 21:14, Chris Albertson wrote:
> > For making a simple and quick GUI in Python I found
> > https://www.pysimplegui.org/
> >
> > With this, I can make a basic GUI app in 15 or 20 minutes with
> checkboxes,
> > sliders, and so on.  It uses dramatically fewer lines of code.  It is
> built
> > on top of tkiner which is the default Python GUI.   There is next to zero
> > learning curve.  Anyone can pick it up quickly.I use this for making
> > robot controller panels, where I have many sliders for parameter
> adjustment
> > and a few action buttons.   But it does not do things like OpenGL in a
> > window.
> >
> Looks interesting Chris, bookmarked, but ATM I like the looks of
> qt-dragon if
> I have to replace axis.
>
> tkinter wastes way too much screen real estate with its huge mandatory
> borders around text. But at the same time I have 20 years
> in making axis do what I want. gtk3 looks to be better than gtk original
> but at 88
> yo, learning a whole new language won't be all that easy.  pyvcp has
> done 90%
> of my wilder ideas, but does have its limits. And ATM I have other irons
> in the fire
> with the ultra short lifetimes of 3d printers. I now have about $1500 in
> a prusa mk3s,
> and its mosquito hot end is leaking, again.
>
> So I have a house and back porch ful of dead printers and am working on
> collecting the parts to build up both an Ender5Plus an a tronxy-400-pro
> into
> 500mm/sec printers.
>
> Like taking the build plate heaters up to 63 VAC power from a 48 lb 2/1
> toroid
> txformers output  just to shorten the 15 minute preheat times of their
> bigger beds.
>
> I'm planning on using the 24 volt bed power as triggers for a 4 pack of
> 60am SSR's,
> two in parallel to power the transformer when one of the printers call
> for bed heat,
> and two more to send the 62 volts from the transformer back to the bed
> that called
> for heat. No use powering the transformer when the printers are off or
> idle.
>
> And I have BTT's octopus-pro's and displays with the higher voltage motors
> coming along with a dozen 60 volt rated, 3 amp drivers for them, and a 4
> pack
> of banana-pi's to run octoprint. And the buggy Marlin firmware's are
> going to
> be gone in favor of klipper when I'm done.
>
> What ships with the tronxy has never made a print yet becausethe bed
> distance
> touch off does not "take" and the bed warp comp only works when it wants
> to,
> never both at the same time. With at 450mm x 450mm bed that warps up in
> the middle when heated just like an alu skillet on the cookstove, a working
> ABL is very important.
>
> So when I am done with the rebuilds, I'll have 63VAC bed heat, and 48 VDC
> on the motors. And will probably have carved up quite a pile of gussets to
> re-enforce frame corners on both the the big ender and on the tronxy-400.
>
> Not advertised is two more even bigger 500mm and 600mm versions that
> tronxy makes, including one with casters to roll it around on the floor
> as its
> envelope is 600mm by, and its frame is over a meter high wide and deep.
>
> And I think it ships with the same buggy and broken electronics, so
> if you want something that Just Works, stay away from tronxy.
>
> I may wind up with nema-23 3 phase motors for x on both of them if I
> don't miss morning roll call first. Those are amazing.
>
> Klipper, among other improvements, can use the output of a sparkfun
> accelerometer
> pasted on the head to correct in real time for frame and belt elasticity
> you see
> everytime you push marlin past about 50mm/sec. Active feedback like in
> linuxcnc.
> but linuxcnc can only use the encoder outputs. That I think is only
> because nobody
> has tried,  the tools are already there but it would take an
> accelerometer ($15 from sparkfun)
> on everything that moves, gantry and head for a 6040, or table and head
> for a 3 axis
> like the go704.
>

Re: [Emc-users] Recommended GUI dev environ?

[Chris Albertson]

For making a simple and quick GUI in Python I found
https://www.pysimplegui.org/

With this, I can make a basic GUI app in 15 or 20 minutes with checkboxes,
sliders, and so on.  It uses dramatically fewer lines of code.  It is built
on top of tkiner which is the default Python GUI.   There is next to zero
learning curve.  Anyone can pick it up quickly.I use this for making
robot controller panels, where I have many sliders for parameter adjustment
and a few action buttons.   But it does not do things like OpenGL in a
window.


On Sat, Nov 5, 2022 at 5:33 AM T Hyde  wrote:

> Good morning!
> I am working on a new project that will require me to develop a brand new
> customized GUI.
> This won't be Axis with an extra tab or panel but will need to be a
> complete GUI screen set. It will likely not have a G-Code preview so it is
> my understanding that Glade or pyqt are probably the most appropriate
> platforms to work with. I have old experience with Glade and the python
> connections required and am quite comfortable there however seem to recall
> some recent threads about glade no longer remaining compliant with Debian
> 10 or the current RT environment in regards to versions available in the
> repositories. Additionally seeing some install notes for pyQT apparently
> some of the environment must also be built from source due to lack of
> binaries and repositories.
> Is the above all old information and there are simple steps to get
> developing? Also is there a preferred platform to lean towards for new
> development purposes that will be supported for the next 3-4 years?
>
> I will be building in the current release of Linux CNC 2.8.4 Deb 10
> Preempt, stock iso install, and would like to keep that compatibility
> forward.
>
> Many thanks on the opinions,
> Ted.
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] CAM software

[Chris Albertson]

You are correct.  The free version only does 3-axis milling. You need
the paid version to get a 4th axis.

The best you can do with the free version is  is "3.5 axis."  That means
manual setup of the rotary table and multiple passes of 3-axis milling.
You can not move the 4th axis while cutting with the free version.   You
would need to design the parts with CAM limitations in mind.

If this is a hobby, then of course cost matters a lot. But for business
use, $545 per year is trivial.

I'm seriously looking at Open Source CAM solutions.  But really, the free
version of Fusion360 can do parts as complex as say a motorcycle transition
housing. Or that shoulder bulkhead assembly I posted.


Hi Chris,
>
> > -Original Message-----
> > From: Chris Albertson [mailto:albertson.ch...@gmail.com]
> > Sent: November-02-22 9:55 AM
> > To: Enhanced Machine Controller (EMC)
> > Subject: Re: [Emc-users] CAM software
> >
> > Here is why hand-coded g-code files can't be used for "real work".
> >
> > I have a part (see link below) that is one of dozens of parts in a
> > quadruped robot.
> > https://a360.co/3sQkHm3
> > The above is a simple web-based viewer that you get for "free" with
> > Fusion360.
>
> > Also, I doubt any normal person could write the g-code for this part and
> > get it correct.
> >
>
> The problem is the moment you want that 4th Axis that can't be hand
> coded.  At that point Fusion360 becomes way more expensive than say MecSoft
> Visual CAD/CAM.
>
> John
>
>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] CAM software

[Chris Albertson]

Here is why hand-coded g-code files can't be used for "real work".

I have a part (see link below) that is one of dozens of parts in a
quadruped robot.  The parts have to be modeled and moved in a simulation to
verify fit.  But also we use strees analysis to verify none of the design
is over-build and too heavy.  If I were to write g-code by hand, then I
introduce an "airgap".  I have no way to prove that what is being built is
what was simulated because I could have introduced an error.   I have every
design change stored in a version control system that tracks errors and
fixes.  Hand coding the g-code breaks this accountability.

You might say if this is only a hobby-level project, who cares about
accountability and formal design changes, but just try skipping the
bookkeeping, I think there are 135 threaded blind holes and a few dozen
keyways that interlock and changes have a ripple effect.

Here is a typical part I designed in Fusion360.  The part is obsolete now
as I found the motors seriously underperform.  The re-design will be much
more complex because I will eliminate the concept of "motor housing" and
build the rotors and stators into the structure and parts will need to be
metal for thermal reasons.

Even if I were able to write g-code by hand for this part, the "air gap"
problem is unacceptable.
https://a360.co/3sQkHm3
The above is a simple web-based viewer that you get for "free" with
Fusion360.  Its real use is to allow collaboration with others,  It uses a
different and simpler interface than does Fusion360.  So I could possibly
ask you for advice about the part, like for example "Could you do this on a
3-axis machine?" or you might look and ask what the three 12mm holes are
for.   I'd say "I 12mm carbon fiber tube is bonded and serves are the
robot's backbones.   So a customer can look at your work, and approve
changes and there is no air-gap, he gets what he saw in the above preview
link.

Also, I doubt any normal person could write the g-code for this part and
get it correct.




On Wed, Nov 2, 2022 at 5:43 AM gene heskett  wrote:

> On 11/2/22 05:50, stjohn gold wrote:
> > I use fusion360 for lathe. It was the only cad/cam program I could find
> at
> > a reasonable price with a post I could modify and get working for my
> > machine. I can't say I like it but it works and the price was right. For
> me
> > it gets the job done and does not cost 20k euros...
> >
> > cheers,
> > St.john
> While I write my own code.  It is not that hard.
>
> 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, 1940)
> If we desire respect for the law, we must first make the law respectable.
>   - Louis D. Brandeis
> Genes Web page <http://geneslinuxbox.net:6309/>
>
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] OT How to make fine splines

[Chris Albertson]

On Thu, Oct 27, 2022 at 6:24 PM andrew beck 
wrote:

> And it would be great to output a file you can save and reuse and combine
> manually later in needed
>

Yes, that was step #3.  I think you can not get around the need to write a
file

> ...
> >
> > The simplest one possible might be called "DrillAHole".  It would
> >
> >1. open up a dialog box and ask for the depth.
> >2. prompt the use to place the drill on the part i the correct
> location
> >and then "press go"
> >3. write out a simple g-code file to the disk that drills a hole to
> the
> >specificed depth
> >4. wait for the user to click "go"
> >5. start LCNC with the file it just wrote are a command line
> argument.
> >
>

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

Re: [Emc-users] OT How to make fine splines

[Chris Albertson]

I was just going to write that what is needed is a good "conversational"
interfasce for LCNC.

You do NOT need to modify LCNC to do this.  The new interface could be a
100% standalone program and run completely in user space

The simplest one possible might be called "DrillAHole".  It would

   1. open up a dialog box and ask for the depth.
   2. prompt the use to place the drill on the part i the correct location
   and then "press go"
   3. write out a simple g-code file to the disk that drill a hole to the
   specificed depth
   4. wait for the user to click "go"
   5. start LCNC the the files it just wrote are a command line argument.


The only difference between "DrillAHole" and some other program called
maybe "MillAPocket" is the content of the g-code and the prompts to the user

Other programs would be more complex.   Say for drilling a hole pattern.
It is limited by your imagination.  These are simple enough programs that
someone who was only an "intermediate level" Python programmer could do
this.   I'd recommend using some VERY simple to learn GUI toolkit so that
more programmers could contribute.  Take a look at this,
https://www.pysimplegui.org/en/latest/

Of course you would not make one program for every kind of operation.
 There would be a higher-level screen where you can select from face
milling, pockets, or holes.   But I'd start with a very basic "DrillAHole"
program to work out the overall plan and prove the concept

The program once written could be adapted to work with other CNC controlls,
not just LCNC.  The key to making this something "anyone" could write is to
NOT allow it to interface to any LCNC internals.  Keep the coupling to LCNC
as minimal as possible.





On Thu, Oct 27, 2022 at 5:34 PM John Dammeyer 
wrote:

> I like this user interface since here you can visualize what the end
> product will look like just like the HAAS or MACH3.
> https://forum.linuxcnc.org/41-guis/26550-lathe-macros?start=240#247195
> How does one go about changing from AXIS to something like the above?  Or
> is that only for the lathe.  Don't have a touch screen.
> John
>
>
>
> > -Original Message-
> > From: andy pugh [mailto:bodge...@gmail.com]
> > Sent: October-27-22 4:09 PM
> > To: Enhanced Machine Controller (EMC)
> > Subject: Re: [Emc-users] OT How to make fine splines
> >
> > On Thu, 27 Oct 2022 at 22:57, John Dammeyer 
> wrote:
> >
> > >His initial number of X moves on the example was over inflated for the
> number of passes he really made
> > when cutting.
> >
> > With LinuxCNC:
> >
> > #1 = 10
> > O100 WHILE [#1 LT 20]
> > G0 X#1
> > F1 F100 Z0
> > G0 X[#1 - 1]
> > #1 = [#1 + 0.025]
> > O100 ENDWHILE
> > M2
> >
> > > The video really shows though, what is missing in LinuxCNC is what
> MACH called Wizards.  Enter the
> > parameters on a screen and have it generate the G Code all within the
> user interface.
> >
> > There is ngcgui, Nativecam, my lathe macros, and a (dormant) Wizards
> framework:
> > And probably others that I have forgotten.
> >
> > https://linuxcnc.org/docs/html/gui/ngcgui.html
> > https://github.com/FernV/NativeCAM (Link to video tutorials at that
> link)
> > https://forum.linuxcnc.org/41-guis/26550-lathe-macros?start=240#247195
> > https://github.com/LinuxCNC/wizards
> >
> > --
> > 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, 1912
> >
> >
> > ___
> > 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
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Fusion 360

[Chris Albertson]

Yes, you physical computer must have rough hardware resources that it can
give the virtual machine enough to run Fusion and still have enough
leftover to run Lininx. So a dual-core 8GB, integrated
graphic laptop would be a poor choice.   If that is all you have.  You
options are to try FreeCAD or  OnShape.OnShape is a good option, it
"feels" like Solid works because it has made by people who left SolidWords
to start a new company.   But you have to pay for the CAM add-ons from 3rd
party developers.FreeCAD is limited but runs native and does well on
low-end hardware.  It is good enough for simple projects.

The question is if you are making a living with this or if it is a hobby.
  If you are running a business, then invest in what you need.  Typically
you'd invest 25% of an employee's salary in capital equipment, so buying a
new $10,000 workstation every four years is not unreasonable.



On Sat, Oct 22, 2022 at 10:14 AM Martin Dobbins  wrote:

> Your saying it is a "resource hog" does not bode well for running it on a
> virtual machine that doesn't have hardware horsepower.
>
> Thanks, Thaddeus
>
> 
> From: Thaddeus Waldner
>
> I’ve done Fusion 360 CAD/CAM classes with middle school kids (age 10-14).
> While it was mostly an exercise on how to follow instructions, many of them
> knew their way around the software and could begin making changes on their
> own by the time it was over.
>
> I reiterate that it is a resource hog, much more so than Solidworks or
> Onshape. It becomes painfully slow on anything with less than 16gb memory
> or with mediocre single-threaded CPU performance.  It seems that none of
> those CAD packages are optimized much for multi-core processors.
>
> > On Oct 21, 2022, at 6:04 PM, Matthew Herd  wrote:
> >
> > I use fusion on both windows and Mac. Mostly Mac.
> >
> > Matthew Herd
> >
> >
> >> On Oct 21, 2022, at 6:52 PM, Martin Dobbins  wrote:
> >>
> >> So, following on from the CAM discussion and all the love shown to
> Fusion 360
> >>
> >> Has anyone tried:
> >>
> >> https://all3dp.com/2/fusion-360-for-linux-how-to-install-it/
> >>
> >> or something similar?
> >>
> >> Or do you all use Windows or Mac?
> >>
> >> Martin
> >>
> >>
> >> ___
> >> 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
>


-- 

Chris Albertson
Redondo Beach, California

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

Re: [Emc-users] Fusion 360

[Chris Albertson]

A lot depends on what parts you are designing.  Simple, single parts with
mostly geometric surfaces can be done even on a low-end laptop PC.  But if
you want to model an entire CNC machine with every ball screw, bearing and
fastener and moving part with movement constraints for what parts are fixed
and which rotate, and which of them slide then you need a bigger PC.   32GB
RAM and 6 to 8 cores and importantly a GOOD GPU.  Just buy an Nvidia 3050
or something like that.   Most real projects have hundreds of parts.  But
for small single-part hobby projects, any PC will work

I've actually experimented.   I can add and remove CPU cores and RAM in
VMware and see the effect on performance.  The worst thing is when you
remove the Nvidia GPU and use software-graphics.  It becomes
unusable-slow.   Performance becomes good with 4-core and 24 GB.   But if
buying a new computer, I'd at least double those specs, and get the best
GPU I could afford.

When Autodesk finally gets off their buts and releases an Apple Silicon
native version of Fusion360, the obvious platform will be the Mac with
perhaps a couple of 4K monitors attached.  But we might be very old before
they do this.

On Sat, Oct 22, 2022 at 9:22 AM Thaddeus Waldner  wrote:

> I’ve done Fusion 360 CAD/CAM classes with middle school kids (age 10-14).
> While it was mostly an exercise on how to follow instructions, many of them
> knew their way around the software and could begin making changes on their
> own by the time it was over.
>
> I reiterate that it is a resource hog, much more so than Solidworks or
> Onshape. It becomes painfully slow on anything with less than 16gb memory
> or with mediocre single-threaded CPU performance.  It seems that none of
> those CAD packages are optimized much for multi-core processors.
>
> > On Oct 21, 2022, at 6:04 PM, Matthew Herd  wrote:
> >
> > I use fusion on both windows and Mac. Mostly Mac.
> >
> > Matthew Herd
> >
> >
> >> On Oct 21, 2022, at 6:52 PM, Martin Dobbins  wrote:
> >>
> >> So, following on from the CAM discussion and all the love shown to
> Fusion 360
> >>
> >> Has anyone tried:
> >>
> >> https://all3dp.com/2/fusion-360-for-linux-how-to-install-it/
> >>
> >> or something similar?
> >>
> >> Or do you all use Windows or Mac?
> >>
> >> Martin
> >>
> >>
> >> ___
> >> 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
>


-- 

Chris Albertson
Redondo Beach, California

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