PICnc with Machine Kit.

388 views
Skip to first unread message

John Dammeyer

unread,
Feb 27, 2020, 1:42:41 AM2/27/20
to machi...@googlegroups.com

This project seems to have lapsed 5 years ago. 

https://github.com/kinsamanka/PICnc-V2/wiki

 

Any interest in resurrecting it?

 

John Dammeyer

 

"ELS! Nothing else works as well for your Lathe"

Automation Artisans Inc.

www dot autoartisans dot com

 

Charles Steinkuehler

unread,
Feb 27, 2020, 8:16:21 AM2/27/20
to machi...@googlegroups.com
I am allergic to PIC chips and 8051 derivatives. ;-)
--
Charles Steinkuehler
cha...@steinkuehler.net

John Dammeyer

unread,
Feb 27, 2020, 1:08:52 PM2/27/20
to machi...@googlegroups.com
There's ARM and MIPS. There was 8086 and M68000. The 8086 won the war as Apple MACs now use Intel and Motorola is now part of NXP. I do wonder whether the support for MIPS and the PIC32 has waned with Microchip purchasing ATMEL.

Perhaps it's better to go TI with their product line?

John
> --
> website: http://www.machinekit.io blog: http://blog.machinekit.io github:
> https://github.com/machinekit
> ---
> You received this message because you are subscribed to the Google Groups
> "Machinekit" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to machinekit+...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/machinekit/ee24f2ef-890f-82eb-c667-
> 8f6290448ec3%40steinkuehler.net.

TJoseph Powderly

unread,
Feb 28, 2020, 12:35:17 AM2/28/20
to Machinekit
Ji Johm
please confirm that you can/cannot build it
and on what platform
I would not be the person to port to another chip or os
but i have built it , and used it before 2019
Thanks Tomp

John Dammeyer

unread,
Feb 28, 2020, 12:54:12 AM2/28/20
to TJoseph Powderly, Machinekit

Hi Tomp,

I would have absolutely no problem doing a schematic and PCB layout in Altium.  No problem in writing the software.  I have a bunch of PIC32 development systems and eval boards around here.  Just did a project about two years ago that captured CAN messages with a PIC32 and via very high speed SPI moved them into a Raspberry Pi with time/date and GPS stamps.

 

My web site home page shows one of my larger CAN based projects.  I also manufacture an Electronic Lead Screw kit.

 

What I've been mulling over is something like a Beaglebone Back with a larger LCD screen, perhaps even 48 buttons instead of 35, and more of a Break Out Board I/O interface.  The general idea is a DRO and Power Feed capability for all axis including spindle detection for power tapping.   Small box relative to PC, Keyboard, Mouse and Screen.

 

But, with an Ethernet port and a MESA like interface so it becomes the hardware interface to LinuxCNC or maybe also MACH3/4.

 

John Dammeyer

http://www.autoartisans.com

 

http://www.autoartisans.com/ELS/

 

 

 

 

 

 

--

website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit
---
You received this message because you are subscribed to the Google Groups "Machinekit" group.
To unsubscribe from this group and stop receiving emails from it, send an email to machinekit+...@googlegroups.com.

Mr Greg

unread,
Feb 28, 2020, 10:11:23 AM2/28/20
to Machinekit
@ JohnD

Re your OP
The answer is yes.
I have been using a Picnc + RPi2 -Jessy OS combo with machinekit for several yrs. Looking to upgrade/resurect/or move on if necessary.
I'm personally not the chap to do the detail code, but am working with a dev who has an interest and can contribute some limited time and effort.

I am currently exploring Pi4 + Buster with a view to either Mesa or Picnc? TBD

@ Tomp
I understand you have had Picnc successfully running on a Pi3 ?
I can't seem to get it to behave on a Pi3. Not with Jessie + MK anyways. All motors just spiral out of control soon as MK is started :(
All the SPI and dma addressing looks compatible They are both 2837
Any ideas?

@ All
By way of a little history. I did quite a bit of testing & verification for Kinsa' on the original Picnc v1.  ( circa 2012/13) This was on a Pic764 which had potential to run faster with more IO
I'm not sure of how much info there is left for that version. I may have some stuff archived?

Cheers
MrGreg

mngr

unread,
Feb 29, 2020, 5:45:48 AM2/29/20
to Machinekit
Hi,

In my humble opinion it should have both SPI and Ethernet connection.
I do not see any difference in the real-time capabilities of these two protocols.

cost would go up a bit, but this project could receive more attention form already existing communities.

Maybe it could re-use the MESA protocol over ethernet, allowing for mutiple slaves to be connected, reusing work on the machinekit side, and attracting the attention of the LinuxCNC community. I am in europe, I cannot buy MESA because it came out way too expensive.
There already are some open source motor driver, like Odrive or VESC. Maybe we could reuse something from here (BLDC are not stepper, i know, but hey, stepper are way easyer!)
How hard do you think it is to move the MESA protocol from an FPGA, VHDL, to a microcontroller?

I would use something like an ATSAME54, which is a cortex m4. I feel that atmel start libraries are well made, but I am not a seasoned developer.

I can work on this, but I need the guide of a experienced one.

Regards,

mngr

John Dammeyer

unread,
Feb 29, 2020, 1:27:54 PM2/29/20
to Machinekit

I do have a MESA 7i92H that works with LinuxCNC on a dual boot PC.  I have to switch to standard parallel port for MACH3 or use a USB Smooth Stepper (more expensive than the 7i92H).

I've also had my mill running with the BeagleBone Black, MachineKit and the Xylotex Cape.  There are issues with the cape, not with the BBB that had me change to a PC.

 

The BBB with Xylotex uses the PRUs for step/dir for up to 4 axis.  It can also take on Encoder input  Although there is a serial  port the CAN bus port is taken up by the Xylotex cape and I think SPI is too.  Partly, IIRC, because the HDMI also uses much of the I/O.

 

There are LCD displays for the BBB but once again they take up all the important I/O and support is minimal.  This screen was done in Python.

http://www.autoartisans.com/MachineKit/BB-W1-LCD4.jpg

 

Once you remove the HDMI and change which pins on the BBB can be used for step/dir and spindle encoder the potential of using SPI or I2C for keypad matrix and LCD display along with at least one Serial Port and one CAN port for CANopen you can address MODbus VFDs and CANopen based hardware for tool changers and other peripherals.

 

That means the basic Machinekit port to the BBB wouldn't have to be touched, the HAL and INI files would take care of remapping the step/dir pins but the AXIS user interface would have to go.  A new interface that handled say 48 buttons, a graphical LCD display and an encoder knob like I have on my ELS could provide the basics.

 

In fact, to start with and to not have to build large amounts of hardware, I'd probably use my ELS with a 20x4 character screen (instead of the 20x2) and the 35 buttons along with the MPG knob. 

For example:

http://www.autoartisans.com/ELS/photos/ELS110D.JPG

 

http://www.autoartisans.com/MachineKit/FourLines2.jpg

 

This photo shows the RS232 connection and if the Y axis step/dir pulses aren't used and the CAN driver chip is installed I've had it operating as a CANopen User Interface Device.

http://www.autoartisans.com/ELS/photos/ELSBottom.jpg

 

A prototype cape that has headers to plug in place of the PIC CPU on my ELS could be the starting point as in the past I've created companion boards that plug into jumper the header on the bottom of the ELS. 

http://www.autoartisans.com/MachineKit/IO_Interface-1.jpg

 

To expand a companion board with a ribbon cable to the 40 pin PIC IC socket then gives access to all the I/O and the header.  The BBB could plug into that.

 

And finally, since the BBB still has Ethernet it runs MachineKit but can perhaps talk MESA Ethernet protocol to a PC running LinuxCNC or MachineKit for the full blown AXIS display.  So you get the best of both worlds.  A small LCD/Keypad/MPG based module not much bigger than a DRO for pseudo manual operation of the mill but also attach a laptop or roll up a PC and you have full blown CNC.

 

This is all still kicking the tires.  Maybe a PIC32 would be an easier solution.

 

John Dammeyer

 

 

 

 

 

From: machi...@googlegroups.com [mailto:machi...@googlegroups.com] On Behalf Of mngr
Sent: February-29-20 2:46 AM
To: Machinekit
Subject: [Machinekit] Re: PICnc with Machine Kit.

 

Hi,

--

website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit
---
You received this message because you are subscribed to the Google Groups "Machinekit" group.
To unsubscribe from this group and stop receiving emails from it, send an email to machinekit+...@googlegroups.com.

ce...@tuta.io

unread,
Feb 29, 2020, 7:27:38 PM2/29/20
to John Dammeyer, Machinekit
Hi,

I am all in for cheap, low entry hardware. Don't care if it is PIC or ARM. I think that for most cases it would be enough to have just serial connection between PC/SBC and this board. But some simple  Ethernet based communication (maybe lower in the OSI model) would be nice too.

I have a feeling that to get traction, it has to be simple, it has to be documented and it has to have tutorials and pre-compiled usecases which are near the real-word wants and needs of machine builders. (Like the BBB and CRAMPS, simple one use device.)

Some type of free-cool-in social networking with Youtube videos would also help.

Then, I think, you have change to get involvement from general community.

And I think that the interest is there: Look for example on the YAPSC:10v project.

BTW, there are implementations for the smartserial protocol which could be used for mining operation. The STMBL guy have implementation and there are guys around the world who reverse-engineered it.

Cern.


Feb 29, 2020, 19:27 by jo...@autoartisans.com:

>
> I do have a MESA 7i92H that works with > LinuxCNC> on a dual boot PC.>   > I have to switch to standard parallel port for MACH3 or use a USB Smooth Stepper (more expensive than the 7i92H).
>
>
> I've also had my mill running with the > BeagleBone> Black, > MachineKit> and the > Xylotex> Cape.>   > There are issues with the cape, not with the BBB that had me change to a PC.
>
>
>  
>
>
> The BBB with > Xylotex> uses the PRUs for step/> dir> for up to 4 axis.>   > It can also take on Encoder input>   > Although there is a serial>   > port the CAN bus port is taken up by the > Xylotex> cape and I think SPI is too.>   > Partly, IIRC, because the HDMI also uses much of the I/O.
>
>
>  
>
>
> There are LCD displays for the BBB but once again they take up all the important I/O and support is minimal.>   > This screen was done in Python.
>
>
> http://www.autoartisans.com/MachineKit/BB-W1-LCD4.jpg
>
>
>  
>
>
> Once you remove the HDMI and change which pins on the BBB can be used for step/> dir> and spindle encoder the potential of using SPI or I2C for keypad matrix and LCD display along with at least one Serial Port and one CAN port for > CANopen> you can address > MODbus> VFDs and > CANopen> based hardware for tool changers and other peripherals.
>
>
>  
>
>
> That means the basic > Machinekit> port to the BBB wouldn't have to be touched, the HAL and INI files would take care of remapping the step/> dir> pins but the AXIS user interface would have to go.>   > A new interface that handled say 48 buttons, a graphical LCD display and an encoder knob like I have on my ELS could provide the basics.
>
>
>  
>
>
> In fact, to start with and to not have to build large amounts of hardware, I'd probably use my ELS with a 20x4 character screen (instead of the 20x2) and the 35 buttons along with the MPG knob.>  
>
>
> For example:
>
>
> http://www.autoartisans.com/ELS/photos/ELS110D.JPG
>
>
>  
>
>
> http://www.autoartisans.com/MachineKit/FourLines2.jpg
>
>
>  
>
>
> This photo shows the RS232 connection and if the Y axis step/> dir> pulses aren't used and the CAN driver chip is installed I've had it operating as a > CANopen> User Interface Device.
> To view this discussion on the web visit > https://groups.google.com/d/msgid/machinekit/228f471a-3292-4373-9fe8-08bf92f8fd3c%40googlegroups.com <https://groups.google.com/d/msgid/machinekit/228f471a-3292-4373-9fe8-08bf92f8fd3c%40googlegroups.com?utm_medium=email&utm_source=footer>> .
>
>
>
>
>
> --
> website: > http://www.machinekit.io> blog: > http://blog.machinekit.io> github: > https://github.com/machinekit
> ---
> You received this message because you are subscribed to the Google Groups "Machinekit" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to > machinekit+...@googlegroups.com> .
> To view this discussion on the web visit > https://groups.google.com/d/msgid/machinekit/17f901d5ef2d%24e86ecd00%24b94c6700%24%40autoartisans.com <https://groups.google.com/d/msgid/machinekit/17f901d5ef2d%24e86ecd00%24b94c6700%24%40autoartisans.com?utm_medium=email&utm_source=footer>> .
>

ce...@tuta.io

unread,
Mar 1, 2020, 4:25:51 PM3/1/20
to Chris Albertson, Machinekit
Hi,
the original author didn't include the <nachi...@googlegroups.com> email in his reply, so I am posting this as an all-in-one answer.

Mar 1, 2020, 06:20 by alberts...@gmail.com:

>
>
> On Sat, Feb 29, 2020 at 4:27 PM cern via Machinekit <> machi...@googlegroups.com> > wrote:
>
>> Hi,
>>
>> I am all in for cheap, low entry hardware. Don't care if it is PIC or ARM. I think that for most cases it would be enough to have just serial connection between PC/SBC and this board. But some simple  Ethernet based communication 
>>
>
> In today"s world and tomorrow's world, the only serial interface that makes sense in a new design is USB-C.  This goes duble if it has to connect to a PC.   Soon this will be the only data connector PCs have.     What is quickly becomming universal is USB 3.1 protocol over a USB-C connector.     It can do 10 Gigabits persecond.   But just as importantly it can be VERY low latency with > isochronous transfers and there are specs for buffer times on hubs.     > Anything designs not using USB-C will be > obsolete>  in a few years.   A limitation ss that USB-C cables can only be 3M long.    Ethernet can go for a few kiliometers if you use fiber but is slower and few computers will have Ethernet in the future.
>
I have one USB-C connector with ThunderBolt capability on my notebook, but I have no idea if it even works as so far I had no device to connect to it. To me it sounds little bit like the predictions that we will travel to Moon bases for our holidays.

But, to allow deterministic (up to a point) communication over USB-C sounds interesting. Do you have some link for papers (or other documents) on this subject? (For when the time allows or the strike of fancy happens.)

Cern.

>  
> --
>
> Chris Albertson
> Redondo Beach, California
>

ce...@tuta.io

unread,
Mar 1, 2020, 5:09:48 PM3/1/20
to John Dammeyer, Machinekit
Guys, please, make sure that you click "Reply to all" and that you include the machi...@googlegroups.com address in the "To" form.

Thanks,
Cern.


Mar 1, 2020, 22:53 by jo...@autoartisans.com:

> Personally I think that's a load of crap. One only has to look at the various microprocessor products out there to see that USB 2.0 as an interface will be with us for a very long time. One of my clients switched from using Panasonic Tough Books for their PC interface in their product to the dedicated industrial PC boxes. And their products are designed to have at least a 10 year life. And those boxes will have USB 2.0 for a long time along with Ethernet.
>
> Take a look out at the industry where suppliers and manufacturers are attempting to alleviate their clients fears about buying a specific technology so they promise at least a 10 year life span. I'm still using a 9S12 series in a product that as yet has not been issued an end of life buy request.
>
> Of course newer PCs will come out with USB-C. Of course over the years we will see products that are sold with that. And we'll probably also see a ton of products that are the USB-C connection to the laptop and have 4 to 8 USB 2.0 connectors to attach to all the real hardware that's out there.
>
> So yes we will see new embedded processors with USB-C. Of course. But to say that the only design choice is that is ignoring what is already there.
>
> John
>> --
>> website: http://www.machinekit.io blog: http://blog.machinekit.io github:
>> https://github.com/machinekit
>> ---
>> You received this message because you are subscribed to the Google Groups
>> "Machinekit" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to machinekit+...@googlegroups.com.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/machinekit/M1MzPd---3-2%40tuta.io.
>>

Mr Greg

unread,
Mar 1, 2020, 5:10:16 PM3/1/20
to Machinekit
Hello

Yes, reasonably priced / cheap "off the shelf hardware", simplicity, docu - manuals - tutorials etc in line with Kinsa's original picnc project.

I suggest it would be best to avoid using ethernet. wifi & BT want to be ruled out of the RT. This leaves ethernet as being the only practical option to keep the machine networked. Somewhat necessary in the majority of use cases.
SPI is most likely available in the likely candidate MCUs. ARM, PIC whatever...

The code is already for PIC. Developing the existing code for a current PIC Dev board does make sense. However..PIC developers seem to be a bit thin on the ground? The popularity of  and future availability PIC Dev boards is also open to question.

The ARM Cortex M0 series and upwards  would perhaps be a good place to start if there is the skill and interest, not to mention quite a lot of time to effect a Porting and development of existing code?
ARM Documentation states that (in theory at least) that code written for the M0 will run on any later Mx series. Eg STM32 series, Arduino Zero, Due, and any number of clones and similar ARM M0, M3 M(whatever boards) that are quite ubiquitous, cheap and a probable assured future as far as can be seen.
The previously mentioned developer (and myself)  started this project a year+ ago.


It was only a start, with limited headway made. External pressures and the enormity of the task brought about a halt. He may well be interested in contributing further TBD.
Guess you will hear from him if he is.....

Just a basic rework and update of the existing Picnc driver and firmware on the existing pic32MX150 that works with current hardware and OSs + Machinekit and Lcnc would be a most excellent solution in itself without porting it to anything else. It works well and is cheap. Gerber files for existing PCBs are available. It is already is capable of fulfilling the use cases of many potential users. Perhaps nobody has taken much notice of it.

For those readers here who haven't


I am prepared to do test & verification on PIs 2, 3 and 4. I have ARM M0 and M3 boards. Bench test cnc and 3/4 axis cnc machine(s) currently running Picnc.
I am happy to write docu, at least for the Pi stuff that I have tested. possibly more TBA.

So, yes I am interested in a resurrection, or perhaps rebirth sounds a bit more positive? of Picnc.

Food for thought and possible collaboration,

Cheers
MrGreg
> To unsubscribe from this group and stop receiving emails from it, send an email to > machi...@googlegroups.com> .
> To view this discussion on the web visit > https://groups.google.com/d/msgid/machinekit/228f471a-3292-4373-9fe8-08bf92f8fd3c%40googlegroups.com <https://groups.google.com/d/msgid/machinekit/228f471a-3292-4373-9fe8-08bf92f8fd3c%40googlegroups.com?utm_medium=email&utm_source=footer>> .
>
>
>
>
>
> --
>  website: > http://www.machinekit.io>  blog: > http://blog.machinekit.io>  github: > https://github.com/machinekit
>  ---
>  You received this message because you are subscribed to the Google Groups "Machinekit" group.
>  To unsubscribe from this group and stop receiving emails from it, send an email to > machi...@googlegroups.com> .

ce...@tuta.io

unread,
Mar 1, 2020, 5:38:15 PM3/1/20
to Mr Greg, Machinekit
Hi,
I have been looking from afar (as I don't need another project at the time) at the Zephyr RTOS project. It gets me thinking, something like PICnc needs mainly access to encoder and PWM peripheral timers, right? And then receive and send messages over some time-deterministic line, right?

So, wouldn't it be possible to use some RTOS "middleware" (for lack of better terminus technicus) for this? That way the support for the hardware would be outsourced to somebody else.

(Sorry for my [maybe] naive idea.)

Cern.

Mar 1, 2020, 23:10 by mr84...@gmail.com:
>> > To unsubscribe from this group and stop receiving emails from it, send an email to > >> machi...@>> googlegroups.com <>>> > .
>> > To view this discussion on the web visit > >> https://groups.google.com/d/>> msgid/machinekit/228f471a->> 3292-4373-9fe8-08bf92f8fd3c%>> 40googlegroups.com>> <>> https://groups.google.com/d/>> msgid/machinekit/228f471a->> 3292-4373-9fe8-08bf92f8fd3c%>> 40googlegroups.com?utm_medium=>> email&utm_source=footer>> >> .
>> >
>> >
>> >
>> >
>> >
>> > --
>> >  website: > >> http://www.machinekit.io>> >  blog: > >> http://blog.machinekit.io>> >  github: > >> https://github.com/machinekit>>
>> >  ---
>> >  You received this message because you are subscribed to the Google Groups "Machinekit" group.
>> >  To unsubscribe from this group and stop receiving emails from it, send an email to > >> machi...@>> googlegroups.com <>>> > .
>> >  To view this discussion on the web visit > >> https://groups.google.com/d/>> msgid/machinekit/17f901d5ef2d%>> 24e86ecd00%24b94c6700%24%>> 40autoartisans.com>> <>> https://groups.google.com/d/>> msgid/machinekit/17f901d5ef2d%>> 24e86ecd00%24b94c6700%24%>> 40autoartisans.com?utm_medium=>> email&utm_source=footer>> >> .
>> >
>>
>>
>
>
>
> --
> website: > http://www.machinekit.io> blog: > http://blog.machinekit.io> github: > https://github.com/machinekit
> ---
> You received this message because you are subscribed to the Google Groups "Machinekit" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to > machinekit+...@googlegroups.com> .
> To view this discussion on the web visit > https://groups.google.com/d/msgid/machinekit/efc06de1-01bd-48fa-881d-c233ac822017%40googlegroups.com <https://groups.google.com/d/msgid/machinekit/efc06de1-01bd-48fa-881d-c233ac822017%40googlegroups.com?utm_medium=email&utm_source=footer>> .
>

Chris Albertson

unread,
Mar 1, 2020, 6:57:48 PM3/1/20
to ce...@tuta.io, John Dammeyer, Machinekit
Yes, for a long time to come if you buy an industrial PC it can have lagacy interfaces on it like USB2, Ethernet and maybe even RS232 or Parallel ports.  But if you have an "ordinary" PC you got at Best Buy, in a few years they will be 100% USB-C and you will need a dongle to connect the older ports.

That said even the idea of using a PC is getting old.  People today use phones and tablets, mostly phones and don't want wires of ay kind.    Yes, of course, those industrial PCs will have wires but look at today's higher-end PC, I don't even see USB connected keyboards except at the low price points

Today WiFi can be much faster then even gigabit Eithernet.  The new 5G networks will also be faster than Eithernet.    A basic cell phone will be much more capable than a PC was only 6 or 8 years ago.

In terms of market share LinuxCNC/EMC does not even show up.  It is nearly zero.

The way to think about a new-start product is to look at the workflow.    Engineers will likley always work at computers with large screens and 3D modelsing CAD systems.   Their work product goes into some kind of version control system that tracks versions and comments and problems.  Then the CAD design gets converted to g-code and also version controlled ad then some sign-offs maybe.   Finally, there is a machine tool.  I thik all users want is to be able to point the machine to a version controled g-code file, set up the stock on clamps and "go".   

John Dammeyer

unread,
Mar 1, 2020, 7:10:59 PM3/1/20
to machi...@googlegroups.com
My idea is FreeRTOS on probably PIC32 because that's what I have. Or possibly use a BeagleBone but at the IDE level scrapping Linux and using just FreeRTOS.

It should have DROs, at least 4 axis power feed (only step/dir, no smart serial), Spindle quadrature for spindle RPM display along with Limit, home and estop inputs. Very much like my ELS but for a mill. And with an Ethernet connection and the appropriate software interface to LinuxCNC.

And around a 7" LCD display with a bunch of mechanical buttons and a quadrature input MPG for easier control.

So essentially an intelligent BoB with an Ethernet connection.

The intelligence comes from the ability to power feed any axis from a BEGIN_n location to an END_n location where _n is the axis letter.

There is always a desire to try and service all markets with a single product. Often they don't serve any of them well. For example the latest crop of electronic gearing for lathes using Arduinos at first appear pretty powerful until you want to run it on Sherline lathe at 3000 RPM. Or you want that custom thread or turn rate and discover that you have to start up the development IDE after keying new values into a spread sheet and then reprogram the processor with some new values. For some that's fun. For others not exactly turnkey. But at a very low cost and some low level participation you get a simple electronic gearing system.

OTOH, I don't know of too many people who manually turn their 3 jaw chuck a quarter turn on their LinuxCNC system and expect the carriage to track that. And if there's 0.025" backlash on the worn half nut that equates to a quarter turn you might just snap the edge off the carbide threading tool bit. So a simple electronic gearing might also work well for a mill.

But with an ELS_Mill the point would be to create a user interface that has the mill behave like a manual mill with DRO and power feed. One might even want the MPG knobs mounted on plates in the same location as the original handles so it has the feel of a manual lathe and a mental change of perspective/direction isn't needed.

With perhaps only 4 or 5 axis this would not work for that robot arm. Nor that specialized CNC system with odd axis. But who cares. As far as I can see 99% of the small mills sold fit the standard XYZ model and very few people even add an A+B axis trunnion setup. If they wanted that there's standard LinuxCNC and MachineKit.

And then run MachineKit or LinuxCNC on a PC or even a Pi4 strapped to the back of a monitor but instead of connecting via Ethernet to a MESA series of boards that ultimately are the Break Out Board they connect to this ELS-MILL. The opportunity of a project like this is that the many many users out there who do not subscribe to any of the CNC groups because they are adamant that they do _not_ want CNC might well add or change up their BoB to add the power feed and DRO features. Then tie in a quadrature encoder to the spindle to get spindle RPM since that new VFD they installed can now remove the need to change belts.

And then, when that project shows up that needs an actual CNC G-Code program to mill a spiral or something the opportunity to suddenly run CNC is there. Ready to go.

So the above is my basic SYNOPSIS or PROBLEM on what that sort of product would do. Under the MATERIALs heading perhaps start with something simple like
https://www.busybeetools.com/products/mini-mill-3-4hp-brushless-motor-cx-serie.html
or even cheaper
https://www.kmstools.com/king-industrial-king-industrial-mini-milling-drilling-machine-154139

Find the kit or youtube video that has a conversion to CNC for it and put on the motors, belts and far east stepper motor drivers. And maybe pick up a cheap BoB to start with so you can use say MachineKit on a BeagleBone Black to at least test the motion and get things rolling.

Next under materials is the display. The Nuts & Volts magazine has started a series on the Nextion displays so perhaps to start with, again to move forward rather than spend it all trying to design the perfect solution,
https://www.itead.cc/nextion-nx8048t070.html
for $75 US.

Wire up a panel with 42 keys in a 7x6 matrix and add an ESTOP button to it. A few serial/parallel chips with SPI selection to poll the switches. A bit of code to read them.

Now we're into METHOD some of which I've described above.
1. We can make our motors move,
2. We've added an encoder to the spindle to read quadrature and determine speed and direction.
3. We've wired up limit switches to help arrest motion at both ends of travel and mark as World Home Position.
4. Using the Nextion software we create a simple user interface with DROs for 4 axis, a spindle RPM and several fields for feed rate and setting spindle speed.
At the moment this looks very much like the Far East CNC boxes with 4.3" displays that you can buy for as little as $200. These are ARM processor based.

So what's next in the METHOD section?
Well you want to create the code to accelerate and move the motors so you need a method of entering values and testing motion for each axis. Take a close look at the method used n MACH 3 and given there's a 7" graphic display this is a good place to copy a concept. The file created by this screen interaction may well be written out (and read back when the screen is displayed) to look like this:
[AXIS_0]
TYPE = LINEAR
HOME = 0.0
MAX_VELOCITY = 3.3
MAX_ACCELERATION = 15.0
STEPGEN_MAXACCEL = 35.0
SCALE = 15000.0
FERROR = 0.05
MIN_FERROR = 0.01
MIN_LIMIT = -0.04
MAX_LIMIT = 13.5
BACKLASH = 0.0215
HOME_OFFSET = -0.1
HOME_SEARCH_VEL = -0.8
HOME_LATCH_VEL = 0.01
HOME_FINAL_VEL = 0.2
HOME_IGNORE_LIMITS = YES
HOME_VOLATILE = 1
HOME_SEQUENCE = 1

That would all be described in more detail in the DESCRIPTION section of this project. An initial layout that describes what is wanted and how it should work. Or just clone the tools that come with LinuxCNC/MachineKit.

Finally we get to the point where we write the code that accelerates and decelerates a motor and runs it a given speed. Track the stepping pulses, convert to engineering units and show them on the display. It's not that much more work to then add another screen with BEGIN/END points for each axis and expand the software to now move from BEGIN to END using the accel and vel values.

So far we're really only talking about each axis operating independently. There's nothing in there, nor maybe should there be anything that says our feed rate is F30 and the distance from BEGIN to END is a tuple like BEGIN(X,Y and Z) and move to END(X,Y and Z). That’s' so easily done with a single G code command that duplicating that means a user wants complex behaviour and so should invest in the time to learn a complex system.

But the reality is that the lowest acceleration value can be used for all axis and the max velocity of the slowest axis is also the limit. So it's now simple linear algebra to calculate the motion of the tool bit from BEGIN to END at F30 and supply an accel value, velocity value and distance value to the motion engine and move the motors so the tool bit travels at the expected rate from BEGIN to END.

The question is should it at this level of a tool. That question needs to be answered before a single line of code is written.

This is all just me thinking out loud. Perhaps the solution for a BeagleBone connected to a decent Cape is to run MachineKit but instead of the Axis Interface run the stand alone ELS-MILL user interface with the LCD screen and array of buttons. But plug in a monitor, keyboard and mouse and press the local/remote switch on the LCD user interface and suddenly you are running MachineKit.

John Dammeyer.


> -----Original Message-----
> From: cern via Machinekit [mailto:machi...@googlegroups.com]
> Sent: March-01-20 2:38 PM
> To: Mr Greg
> Cc: Machinekit
> Subject: Re: [Machinekit] Re: PICnc with Machine Kit.
>
> Hi,
> > On Sunday, 1 March 2020 00:27:38 UTC, ce...@tuta.io wrote:
> >
> >> Hi,
> >>
> >> I am all in for cheap, low entry hardware. Don't care if it is PIC or ARM. I
> think that for most cases it would be enough to have just serial connection
> between PC/SBC and this board. But some simple� Ethernet based
> >> > Subject:> �[Machinekit] Re: PICnc with Machine Kit.
> >> >
> >> > �
> >> >
> >> >
> >> > Hi,
> >> >
> >> >
> >> > �
> >> > �
> >> >
> >> >
> >> >
> >> > --
> >> > website: > >> http://www.machinekit.io>> > �blog: > >>
> http://blog.machinekit.io>> > �github: > >>
> https://github.com/machinekit>>
> >> > ---
> >> > You received this message because you are subscribed to the Google
> Groups "Machinekit" group.
> >> > To unsubscribe from this group and stop receiving emails from it, send
> an email to > >> machi...@>> googlegroups.com <>>> > .
> >> > To view this discussion on the web visit > >>
> https://groups.google.com/d/>> msgid/machinekit/228f471a->> 3292-4373-
> 9fe8-08bf92f8fd3c%>> 40googlegroups.com>> <>>
> https://groups.google.com/d/>> msgid/machinekit/228f471a->> 3292-4373-
> 9fe8-08bf92f8fd3c%>> 40googlegroups.com?utm_medium=>>
> email&utm_source=footer>> >> .
> >> >
> >> >
> >> >
> >> >
> >> >
> >> > --
> >> > �website: > >> http://www.machinekit.io>> > �blog: > >>
> http://blog.machinekit.io>> > �github: > >>
> https://github.com/machinekit>>
> >> > �---
> >> > �You received this message because you are subscribed to the Google
> Groups "Machinekit" group.
> >> > �To unsubscribe from this group and stop receiving emails from it, send
> an email to > >> machi...@>> googlegroups.com <>>> > .
> >> > �To view this discussion on the web visit > >>
> https://groups.google.com/d/>> msgid/machinekit/17f901d5ef2d%>>
> 24e86ecd00%24b94c6700%24%>> 40autoartisans.com>> <>>
> https://groups.google.com/d/>> msgid/machinekit/17f901d5ef2d%>>
> 24e86ecd00%24b94c6700%24%>> 40autoartisans.com?utm_medium=>>
> email&utm_source=footer>> >> .
> >> >
> >>
> >>
> >
> >
> >
> > --
> > website: > http://www.machinekit.io> blog: > http://blog.machinekit.io>
> github: > https://github.com/machinekit
> > ---
> > You received this message because you are subscribed to the Google
> Groups "Machinekit" group.
> > To unsubscribe from this group and stop receiving emails from it, send an
> email to > machinekit+...@googlegroups.com> .
> > To view this discussion on the web visit >
> https://groups.google.com/d/msgid/machinekit/efc06de1-01bd-48fa-881d-
> c233ac822017%40googlegroups.com
> <https://groups.google.com/d/msgid/machinekit/efc06de1-01bd-48fa-881d-
> c233ac822017%40googlegroups.com?utm_medium=email&utm_source=foot
> er>> .
> >
>
> --
> website: http://www.machinekit.io blog: http://blog.machinekit.io github:
> https://github.com/machinekit
> ---
> You received this message because you are subscribed to the Google Groups
> "Machinekit" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to machinekit+...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/machinekit/M1NEzHw--3-2%40tuta.io.

John Dammeyer

unread,
Mar 2, 2020, 7:16:38 PM3/2/20
to Machinekit

A bit of digging through my boxes after a double check of the data sheet shows I find that the PIC32 won't serve my needs.   I thought it had a quadrature encoder module but it doesn't.   I have PIC32MX boards and a bunch of dsPIC33F series boards.  

 

The problem is the PIC32MK which has motor control features and the quadrature encoder doesn't do Ethernet. However one could add the ENC28J60 to deal with Ethernet if it was wanted for connection to a MachineKit, LinuxCNC or MACH3 PC.  Or just use a Pi or Beagle running MachineKit or LinuxCNC and SPI and forget about Ethernet connectivity.

 

I have these modules.

http://www.autoartisans.com/PIC32/AutoBoard_CAN_EEROM.jpg

http://www.autoartisans.com/PIC32/PIC32_Sets.jpg

 

And a bunch of others with dsPIC33F series modules.

 

This will require some more research…

John Dammeyer

 

 

 

 

justin White

unread,
Mar 2, 2020, 8:56:01 PM3/2/20
to John Dammeyer, Machinekit
Yeah, there is no reason to use ethernet when almost everything has a uart. SPI is a viable rt interface and there is no reason to tie up the processor boards ethernet port. Mesa is using spi with the raspberry pi boards, the 7c80 and the 7c81, so you are pretty much guaranteed it works. Also being that spi comes off a gpio header there is no rj45 connector to add to the pic card.

I'm lost on the USB-c talk.....an rt interface is about transferring small packets quickly. USB regardless of the generation isn't required, or even really suitable. Mesa doesn't even use gigabit on its ethernet cards, its 10/100.

I've never heard of this PICnc before now but it seems like it could have been a good idea but that micro was too small, and imo too slow. Step rates were only mentioned at 40khz which is OK I suppose but mesa is doing them at 10mhz on a spartan 5/6. Stepgens and encoders only wind up running at a fraction of the speed of the controller and a 40mhz pic32 seems to get 40khz, while a 500-600mhz fpga is getting them at 10mhz.

I realize cost is a thing here but what is the price difference really between the original pic32 used and a higher end micro?

--
website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit
---
You received this message because you are subscribed to the Google Groups "Machinekit" group.
To unsubscribe from this group and stop receiving emails from it, send an email to machinekit+...@googlegroups.com.

John Dammeyer

unread,
Mar 2, 2020, 9:33:11 PM3/2/20
to justin White, Machinekit

Let me go back one step and outline my idea.  First I have machinekit running on a BBB with a Xylotex Cape that has a DB-25. There are some problems with the high verses low for switches NC/NO so I ended up not using it.

 

I'm not directly thinking of the MachineKit, LinuxCNC or MACH3/4 group of users.  I'm thinking of the users who don't even sign onto the CNC groups and aren't interested in CNC.  Often because they don't know what they don't know.

 

What I'm thinking of is an ELS-MILL that has DROs, Step/Dir signals, Spindle Speed control, and ideally spindle speed feedback along with the usual group of inputs and outputs like limit switches and coolant etc.

 

Really just a DRO setup with power feed on each axis.  For someone using a manual mill the addition of power feed to an axis is such a tremendous improvements and there are all sorts of ways of doing this.  Model Engineer's Workshop magazine etc all have had articles along that theme.

 

Now grab that laptop or PC neither of those have SPI or IIC at least not without an add on card of some sort.  But they pretty well all have Ethernet.  So BBB, Pi, PC, Laptop embedded PC box all with LCD Display or HDMI, USB for keyboard and mouse all running some form of CNC software with the motion commands out Ethernet to a target. 

 

And if that target is the ELS-MILL (which serves as a Break out Board too) then you can easily migrate a non-CNC user to the wonders of CNC.  And if you use a BBB or Raspberry Pi running machinekit in a way that appears to be an ELS-MILL but with a local/remote switch changes personalities you now have it all bundled into one board.

 

But that might be asking too much hence the idea of using a PIC32 or something like that for the ELS-MILL.

 

John

 

 

From: machi...@googlegroups.com [mailto:machi...@googlegroups.com] On Behalf Of justin White
Sent: March-02-20 5:56 PM
To: John Dammeyer
Cc: Machinekit
Subject: Re: [Machinekit] Re: PICnc with Machine Kit.

 

Yeah, there is no reason to use ethernet when almost everything has a uart. SPI is a viable rt interface and there is no reason to tie up the processor boards ethernet port. Mesa is using spi with the raspberry pi boards, the 7c80 and the 7c81, so you are pretty much guaranteed it works. Also being that spi comes off a gpio header there is no rj45 connector to add to the pic card.

justin White

unread,
Mar 3, 2020, 6:21:04 AM3/3/20
to Machinekit
I've seen the mention of these ideas but it's all over the place, there is mention of too many things that are in no way compatible with each other. The original post simply states that you want to resurrect the PICnc which the firmware and hal drivers already seem to be available for to run as the motion interface for a LinuxCNC/MK host, specifically the Rpi over SPI, and that doesn't work with Mach3  so now you're either writing another set of firmware or you're abandoning Mach3.

Then you seem to want to run this as a standalone electronic Lead screw thing that is going to require different firmware again, but it's supposed to drive a small LCD itself and be controlled by hardware buttons on a control panel I presume. But it is supposed to retain the functionality of being controlled by LCNC/MK? Is there a driver for that? what is the means of commanding it from LCNC/MK? If someone is going to plop a non purpose built laptop or PC in front of it to casually decide to do CNC today, that laptop needs to be running a RT kernel, I don't know that MK/LCNC will actually run a machine from a Live image.

All this when realistically, if the project is left with it's original intention of running as the motion interface for an rpi, the only thing that needs to happen you just create a simple GUI and config to do nothing more than be and electronic lead screw. The Rpi obviously can run a monitor out of the box and there will never be a need to reflash the firmware on the PICnc device. The Rpi itself already has an ethernet interface so if it's not being tied up by the PICnc thing as it's running on spi all someone has to do is plug the laptop into it and VNC to get a bigger monitor and keyboard going...........or you can just plug a bigger monitor and keyboard in to the pie lol.

So what is the purpose of doing it any other way? to avoid the cost of a rpi?

To unsubscribe from this group and stop receiving emails from it, send an email to machi...@googlegroups.com.

--
website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit
---
You received this message because you are subscribed to the Google Groups "Machinekit" group.

To unsubscribe from this group and stop receiving emails from it, send an email to machi...@googlegroups.com.

John Dammeyer

unread,
Mar 3, 2020, 12:02:38 PM3/3/20
to machi...@googlegroups.com

I thought I was quite clear about this.  This PICnc version requires Ethernet.  LinuxCNC through a MESA Ethernet 7i92H or MACH3-4 through an Ethernet Smooth Stepper control hardware.  I am going to guess that the MESA driver for the 7i92H also exists for MachineKit.

 

The MESA board I have has the stepping-PWM  engines implemented in an FPGA. What I´d like, but what may not be possible, is to replace the MESA 7i92H with the updated PICnc. 

 

If indeed the stepping engine is now within the PIC as well as the spindle quadrature etc. and the board is in effect the Break Out Board for the machine then there is absolutely no reason you can´t have a single input serve as a local-remote switch.  In remote it behaves like, say a MESA 7i92H and the buttons and display appear to behave like a pendent on LinuxCNC.  In local it´s a DRO & PowerFeed machine controller with a bunch of useful buttons to simplify manual operation.

 

And internally there's a remote config parameter that can be set to LinuxCNC or MACHx. Ie. MESA 7i92H or SmoothStepper.

 

Now maybe this won't be possible without also having an FPGA duplicate what is done with say the 792H.

 

Maybe this won't be possible at all with MachineKit.

 

Hope that clarifies things a bit.

 

John

 

 

To unsubscribe from this group and stop receiving emails from it, send an email to machinekit+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/machinekit/ebbcaad7-6dff-47d7-932c-742d4faf3e95%40googlegroups.com.

justin White

unread,
Mar 3, 2020, 8:34:49 PM3/3/20
to Machinekit
I thought I was quite clear about this. 
Not at all, probably because you are misunderstanding the purpose of the hardware you are mentioning in this post.
 
This PICnc version requires Ethernet.  LinuxCNC through a MESA Ethernet 7i92H or MACH3-4 through an Ethernet Smooth Stepper control hardware.  I am going to guess that the MESA driver for the 7i92H also exists for MachineKit.
You want to run "mesa" ethernet from the 7i92H on the PICnc? Doesn't sound easy and I doubt there is anything specific about the 7i92 in the hm2_eth driver other than the fact that the board types are probably added to the driver as they are released. MK is well behind LCNC in Mesa drivers  so no idea if it works....you should probably test it. Mesa cards don't run at all on windows with the hostmot2 firmware that the MK/LCNC projects interface with, and nobody around here can use a "smooth stepper" because of the architectural differences between Linux Preempt kernels and the Windows way of doing things. Windows devices tend to buffer things in hardware to avoid RT requirements, which is why USB hardware is a thing in Windows and not Linux. The "Linux way" is to run in a RT capable kernel and minimize the "load" by having the hardware do the heavy lifting, but it is still pegged in real-time. 

So you're hardware is not going to just be compatible with both, there is very little that is as it stands. This is probably not the place to discuss your Mach3/4 Windows needs (it's making me cringe just thinking about it). Mesa does however have a "SoftDMC" firmware that is meant for WIndows applications. An FPGA is far more versatile than a micro so there may be something you can do with that, but SoftDMC will not run under LinuxCNC and I seriously doubt anyone here knows much about it. 

The MESA board I have has the stepping-PWM  engines implemented in an FPGA. 

They all do, it's part of hostmot2 firmware, as are the encoder, Smart serial and a few other firmware modules.

What I´d like, but what may not be possible, is to replace the MESA 7i92H with the updated PICnc. 


They functionally do the same exact thing. The PICnc is doing the stepgen, encoder, PWM stuff as firmware modules in the micro as well. The difference is ethernet, support, and an FPGA vs SPI, a dead project, and a micro

If indeed the stepping engine is now within the PIC as well as the spindle quadrature etc.....
 It is

and the board is in effect the Break Out Board for the machine then there is absolutely no reason you can´t have a single input serve as a local-remote switch.  In remote it behaves like, say a MESA 7i92H and the buttons and display appear to behave like a pendent on LinuxCNC.  In local it´s a DRO & PowerFeed machine controller with a bunch of useful buttons to simplify manual operation.
That's not going to happen like you want it to, unless you are capable of writing complex firmware. Your best bet is to abandon the PICnc thing for what you want to do and call Mesa. Hostmot2 is "host-based-motion-control" it only runs with a host i.e. LinuxCNC. SoftDMC is something completely different and like I said, it's never really discussed in the Linux world. The Mesa cards generally have 2 EPROMs and one can contain Hostmot2 firmware and the other SoftDMC. SoftDMC may possibly be able to run by itself and take care of your "local" thing....I don't know.

Now maybe this won't be possible without also having an FPGA duplicate what is done with say the 792H.

 Maybe this won't be possible at all with MachineKit.

The PICnc already does functionally what the 7i92 does, just alot less of it and over spi.

The PICnc sounded interesting as I've never heard of it before. the board isn't very interesting and the micro is too small, but the fact that the original designer made firmware and drivers for it makes it something viable for resurrection, with SPI IMO.

Honestly your intended use doesn't sound very interesting at all because it requires way too much non-existent configuration and what you are actually trying to do is SUPER EASY just using an Rpi and a 7C80, 7C81, the original picnc...whatever. A DE10-Nano running MKSOCFPGA would do it all day.

I'm all in for a Rpi based SPI Resurrection of the original PICnc.

 


 

John Dammeyer

unread,
Mar 3, 2020, 9:22:44 PM3/3/20
to justin White, Machinekit

Hi Justin,

Clearly we're on different pages. 

Thanks for your input.

John

 

 

From: machi...@googlegroups.com [mailto:machi...@googlegroups.com] On Behalf Of justin White
Sent: March-03-20 5:35 PM
To: Machinekit
Subject: Re: [Machinekit] Re: PICnc with Machine Kit.

 

I thought I was quite clear about this. 

--

website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit
---
You received this message because you are subscribed to the Google Groups "Machinekit" group.
To unsubscribe from this group and stop receiving emails from it, send an email to machinekit+...@googlegroups.com.

justin White

unread,
Mar 3, 2020, 9:50:29 PM3/3/20
to Machinekit
Yep,

Actually I misspoke anyway, the PICnc never got an encoder module and the PWM stuff was "being worked on" so it's strictly a Stepgen/gpio project where it stalled. 

To unsubscribe from this group and stop receiving emails from it, send an email to machi...@googlegroups.com.

John Dammeyer

unread,
Mar 4, 2020, 1:21:38 AM3/4/20
to justin White, Machinekit

This is the only viable PIC32 is since it has the QEP module and PWM for motor control.  https://www.microchip.com/wwwproducts/en/PIC32MK1024MCF064

Unfortunately not Ethernet.

 

John

 

"ELS! Nothing else works as well for your Lathe"

Automation Artisans Inc.

To unsubscribe from this group and stop receiving emails from it, send an email to machinekit+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/machinekit/fa1ca289-f15b-422a-b875-cb7ead005dbe%40googlegroups.com.

ce...@tuta.io

unread,
Mar 4, 2020, 8:18:11 AM3/4/20
to John Dammeyer, justin White, Machinekit
Hi,

since we are having extensive talk about this kind of hardware - did anybody spare any thought to EtherCAT slave device which would be useful for home machine integrator/retrofiter? (Some basic IO, encoder counters, analog signal output generation and step/dir signals output [in one or two devices] - basically what MESA sells mostly to LinuxCNC crowd.)

My fancy taken me to XMC4800 from Infineon or LAN9252 from Microchip, there is even OpenSource project: https://github.com/DieBieEngineering/DieBieSlave - however the author overdid it with layers and it would still need another board.

Cern.

Mar 4, 2020, 07:21 by jo...@autoartisans.com:
> To view this discussion on the web visit > https://groups.google.com/d/msgid/machinekit/8da44c79-f44c-4aa1-ba94-b8185fe1bdf8%40googlegroups.com <https://groups.google.com/d/msgid/machinekit/8da44c79-f44c-4aa1-ba94-b8185fe1bdf8%40googlegroups.com?utm_medium=email&utm_source=footer>> .
>
>
>
>
> --
> website: > http://www.machinekit.io> blog: > http://blog.machinekit.io> github: > https://github.com/machinekit
> ---
> You received this message because you are subscribed to the Google Groups "Machinekit" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to > machinekit+...@googlegroups.com> .
> To view this discussion on the web visit > https://groups.google.com/d/msgid/machinekit/fa1ca289-f15b-422a-b875-cb7ead005dbe%40googlegroups.com <https://groups.google.com/d/msgid/machinekit/fa1ca289-f15b-422a-b875-cb7ead005dbe%40googlegroups.com?utm_medium=email&utm_source=footer>> .
>
>
>
>
>
> --
> website: > http://www.machinekit.io> blog: > http://blog.machinekit.io> github: > https://github.com/machinekit
> ---
> You received this message because you are subscribed to the Google Groups "Machinekit" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to > machinekit+...@googlegroups.com> .
> To view this discussion on the web visit > https://groups.google.com/d/msgid/machinekit/02e601d5f1ed%2422020ac0%2466062040%24%40autoartisans.com <https://groups.google.com/d/msgid/machinekit/02e601d5f1ed%2422020ac0%2466062040%24%40autoartisans.com?utm_medium=email&utm_source=footer>> .
>

dave engvall

unread,
Mar 5, 2020, 1:52:22 PM3/5/20
to machi...@googlegroups.com


It is my understanding that Peter's 'softdmc' is motion only. Nice 50 us cycle time, axis gearing, takes care of jerk, etc but needs all the other stuff wrapped around it. Years ago there were rumors that someone was writing a pkg for softdmc that  would use all the lcnc development except motion. I never found out the source of the rumor. I do think a very bright team of about 4 to 6 coders could get a package working. Softdmc has always been free but Mesa now lists it a free in some places on the site and $100 in the price list. Just lifting out the lcnc motion and dropping in softdmc sound easy but as ususal the devil is in the details.

My D525 cpu rolled over and died so I'm working on a Rpi4B, 7i90, 7i33  servo setup using both rotary encoders and glass scales (X and Y). Z will remain rotary encoder. No promises on time line.

HTH

Dave


 

 

--
website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit
---
You received this message because you are subscribed to the Google Groups "Machinekit" group.
To unsubscribe from this group and stop receiving emails from it, send an email to machinekit+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/machinekit/8da44c79-f44c-4aa1-ba94-b8185fe1bdf8%40googlegroups.com.

--
website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit
---
You received this message because you are subscribed to the Google Groups "Machinekit" group.
To unsubscribe from this group and stop receiving emails from it, send an email to machinekit+...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages