Advanced Search

Hi Salame,

I live in Holland near Gorinchem and Rotterdam.
It would be great to have a servo drive and motor and test it to be auto configurated.

it means that any servo drive will be auto configured?
At this stage, i can not confirm this.
When i retrieve the info from the ethercat bus, and pdo config is made for the drive.

Then i can test the servo drive to send commands.

It would not mean every drive would have the same pdo config, these would be different,
but we can automaticly sort data out and do suggestions for that data like:

If it comes out that a servo position command is always a 64 bit value. Then we can already map this to
a position command. It's just figuring out a few things.

If it turn's out there isn't any logical explanation in the servo pdo config. Then we can add a 
template for that particular servo drive vendor id.

@Fakename
The links are broken indeed, the source has moved to :
codeberg.org/skynet/realtime_app/src/bra...dor/ethercat_twincat

It is part of a bigger project that used opencascade.
 

thanks
for the gui you need to install some python packages.


@all
oss-cad-suite stuff runs also on windows now

 

so it is not so hard to run the other toolchains also

EDIT: gowin works also
 

I can run it under Windows now, but the graphical stuff does not work. 

In my case, it would be more convenient, as I only have Linux on the Raspberrry Pi 5.

Anyway, it is better that you use your time on the important stuff. The whole project is very impressive. Congratulation!!!

Thanks for the images. I'm encouraged to see some have the low data rate type A interface, so it appears the data requirements to the MESA card are not super demanding. USB data throughput has really evolved from the original 1.0 (12Mb/s) to 3.3 (20Gb/s) or 1600x faster. From what I have read 95% of the cheap so called "fast" 3.2 cables don't even have the conductors for the fast laneways so the port reverts back to the lower original slow laneway and the non technical people wonder why the device doesn't seems all that much faster?

I found an original IBM Thinkpad "OnePort" docking station on eBay for $10. In addition to the Ethernet port it has 2x USB 3.0 ports (5Gb/s) so the "OnePort" interface must be pretty capable, that "OnePort" docking station interface was replaced with a USB 3.x port the following year.

I'm still at the beginning of my LCNC experience, it means that any servo drive will be auto configured? If you want to make some test I can send you one, where are you located?

Thank you Collins.

Attached video shows the 3 interpreters tools moving around.

There is no motion component or whatever. The tool is just jumping to the consumed gcode position.

One of the problem's is when adding gcode segments to a infinite program, it has to check for duplicates,
and there must be a limit to the gcode preview. Otherwise we get a buffer overflow.

This run's quite stable, fixed some runtime bug's.

Was looking for info about raspberry, do you suggest to make a new buil on pi 4 because it's more stable? Or it's worth go with the 5?

you can use this Spartan6 module, but have to setup/modify 'riocore/boards/Spartan6/board.json' and change package and type for your needs.

programming will not work over w5500 you need to write the bitfile into the flash.

---

@all
WINDOWS-WARNING:
rio was not intended to run on windows, at the moment i can't give windows/mac support for it.
I just made the system a bit more platform independent.

i normally don't use windows at all, but i think if you customise the generated makefiles a bit, you could generate the bitfiles there too, but very experimental .

Another question for Meister:

FPGA Module 2.00: Spartan 6 FPGA Board.

I would like to use that FPGA board. It has SPI / JTAG connectors for programming EEPROMs. Is it suitable?

I assume that the W5500 uses another SPI module inside the FPGA after programming for the data communicaktion with LinuxCNC. Therefore, I think it the configuaration EEPROMS cannot be flashed via the W5500. Is it correct?

Does the RIO System write the FPGA everytime when getting online or is the data permanently flashed on the EEPROMs?
 

i think you are on the wrong branch:

git checkout dev


BTW: very experimental, only for the gui and to generate the config-files,
i think, building bitfiles will not work yet on windows


 

Hello community,
I have another problem to solve, namely my feed potentiometer only works at mm/min.
It has no function at mm/rev. The potentiometer for rapid traverse always works.
Is there no other way or can it be influenced?
If you have to run in a program, a potentiometer like this is very helpful

Thanks and best regards
Olli

Hi Meister,

Unfortunately, I cannot make it run under Windows.

I have changed:

file_path = r".\riocore\__init__.py"
if os.path.isfile(file_path):
sys.path.insert(0, os.path.abspath("."))

When running rio_setup, the first problem, I cannot import halpins from riocore.
Directory is OK and the __init__.py file is there.

Perhaps you can help me out. Thanks a lot!

Thank you. I will study more.

And for LAN9252, it is better to copy the whole lib/soes from Easercat 2000 and go from there. That should work right away.
 

You can also add a PDO just for debugging, I had several of those during development.
Just to assign Obj.D1 = interesting_value and watch that value.
An alternative to the index shift is to have a counter counting up when the callback happens, there is
actually one already in my code "encCnt" which increments every time the index callback is called.
Just quickly checked in the debugger, here the callback is triggered when index passes by.