Advanced Search

It is EL6751 CANopen master from beckhoff with some sew lte-b+ vfd wchich is build in them directly. When I working on it via the TwinCAT3 enviroment, there are TX RX PDOs formated in the shape which I last described.. array[0...7] of bytes

I think there should be possibility to write them as complex with 64bit size and then slice it inside to correct s32 or u32s for the hal...

regards

@HansU
Have you tried patching/running this source code? I would appreciate some M/G code help.
forum.linuxcnc.org/gmoccapy/53025-setup-...-embedded-tab#303550

As far as driving the A-D chips from that  parallel port,
That should be possible even using the normal hal_parport driver
and a simple component with a basethread driven state machine.

(ChatGPT will not help here, it's just a time sink)

The main issue would be the encoder as the encoder resolution*speed
will limit the maximum speed significantly with parallel port sampling speeds.

Is parport fast? nope. Is it enough? Certainly is!! Some of the parts coming out of my Lcnc parport 4ax mill are beyond words - if measured for what it is.

No comments on the ChatGPT bit. Apart from being glad I won't be here in 20~50 years to witness a chronically suffering civilization whom had their natural neural plasticity annihilated for incentives by BigTech. Geez imagine the keyboard zombies coming up!!!

meister wrote:

I'm sorry, I was a bit bored
riocore/configs/MecanixDev/config-test.json
riocore/boards/MecanixDev/board.json

Stupid me... I could have cloned/checkout the 'dev' branch rather than main and use what you've done already.
Explains why I couldn't find it locally lol. Let's try that again. 

meister wrote:

copy a similar one and adjust the values (type/clock-speed and pin).

One word, Oliver; EPIC.

Let me learn what you've done for yourself and all of us first. This will require time on my side, but will be worth it I have a feeling - "it works" (see visual)!! I'll update on progress and hopefully can contribute what was done at my end. Interim, let me know if you need the sources/gerber/bom/anything for the dev board in use. Better yet, let me know what else could potentially go on it to make it a "RIO-BOARD" (RTL8111H?!) and I'll see what I can do. 

you check id uefi or not as rodw explain here: forum.linuxcnc.org/ethercat/45336-etherc...-how-to-step-by-step ??

In my case need (not uefi but have that error because not all libs needed was installed) to install libethercat 2 time second one using synaptic. After it goes ....

My xml ... no other ... for axis use one only and sim pin

I've set servo realtime only to 6ms ..... but during 6 hour working receive these error on dmesg:


mmmm .... EK1100 have only 0.5 mt of cable from pc .... is not near every tipe of electrical noise .... have separate power input ..... how is possible receive these warning?

where is the error?

regards
 

I just wanted to show that it is not that difficult to support new boards.
It is a bit more complicated to create the board with a picture and the correct pin positions for the setup tool, but this is only optional.

If the toolchain is already supported, it is quite easy,
copy a similar one and adjust the values (type/clock-speed and pin).

WOW, so many responses. I look away for a few minutes and already the topic is on pg.2. LOL

anyway. Aciera the reason I don't have the error message posted is because that happend days ago
and I found a work around for that instance. I posted this topics question so I know for future reference if
it happens again how best to deal with it. The way I handled it worked this time, but may not on
what ever machine I would be building in the future. I don't mean to have people frustrated trying to figure
out what I am trying to do but at the time of the problem I didn't think about saving the error report I
just tried something else and that worked. I will try to be more diligent next time.

tommylight when I get a chance this evening (my time anyway) I will give that a good look over.Thanks.  And yes
this because of the Foam cutter. I am just trying to find better ways of doing things.LOL

I've took it from here : (full)
github.com/scottalford75/Remora-NVEM/tre...%2Fremora-nvem-basic

I've done sometimes full chip erase,
I will try to erase it then put 1.3.3 with config you advised. But I'm not sure why at the moment I'm not getting anything from rs232

Second post here, configs attached, and so are STL so skip them
forum.linuxcnc.org/show-your-stuff/46752...ver-etcher-scratcher

for these pourpouse (one impot to multiple output) in 2014 i write a simple component .... think shared here pehaps ... but after really see Andy insert a new one for the same in Comp list of linuxcnc .... any how actually prefer to use logicladder ... more easy and efficent ..... yea block is short .... if is for a special pourpose you need write a new user comp .... at least is easy for these. .... if(x=1){.......} elseif(x=2){.....} etc etc ..... pherhaps compgenerator now it work well? into 2.7 and 2.8 not so ....

It is part of any servo config in LinuxCNC, even steppers with encoders have similar pins and all 3 tied to the same bus bar.
-
Is using the bus bar analogy helping or not?
I ask because using a joint(of wires) is a no go here due to joint being used for other stuff and causing confusion.
Edit:
Is this for the foam cutter?
If yes, you can download the configs for the "scratcher" and they use PWM and have logic for disabling PWM on stop or esc or e-stop, could save you headaches in use, a lot!

As I understand it you want the input signal to go to all the outputs. If so you don't need a component.

If however you want to select which output the input signal goes to (ie a demultiplexer) then you would need to insert hal components.

[edit]
Posting your hal file(s) and the error message would likely cut down on the guess work quite a bit