Advanced Search

I would not focus that much on language.
From my point of view there are (at least) two other more important points:

- good documentation of the interface to EMC
- modular design with reusable elements

GUIs are often a matter of taste, and depending on the machine.
Which buttons are useful, which icons do you like, etc
But there are also common elements like 3D preview, g-code properties etc
Instead of starting the 17th GUI, maybe someone starts creating a toolchest, from which it could be more or less easy to design your own GUI.

regards, Gerd

cmorley wrote:

I have not heard of anyone running steppers (or step/direction servos) with encoder feedback that was used in a PID loop.

I think that people have used stepper motors in velocity mode with an encoder for closed-loop control, but I am not sure how well it has worked.

I assume that you could use a step-dir servo drive with a PID loop fairly readily. You would be asking for changes in position rather than velocity to close the loop, but I see no reason why it couldn't be made to work. In effect absolute position and homing would be handled by EMC2 PID with loop stability handled by drive PID.
There may be velocity-mode drives with up-down inputs too, where it would make sense to use a particular stepgen mode as the output function.
Stepgens are also usable as the output function for a frequency-voltage external stage, which I have seen suggested but don't know if it is often used. (freqgen is floating point so might not always be preferred).

Not that the possible existence of a particular setup justifies adding it to a tool like pncconf. I think it is perfectly reasonable to expect "unusual" configurations to be hand-coded in HAL. It makes no sense for you to spend a week saving one user an hour.

python: tnl/t_draw.c:203: bind_inputs: Assertion `inputs->BufferObj->Pointer' failed.
/usr/bin/emc: line 637: 5987 Aborted $EMCDISPLAY -ini "$INIFILE" $EMCDISPLAYARGS $EXTRA_ARGS
5862
PID TTY STAT TIME COMMAND
5986
PID TTY STAT TIME COMMAND
Stopping realtime threads
Unloading hal components


To those in the know __ it happened when I use the scroll wheel to zoom out.

C
C++
python

It needs to be a main stream language so a bigger base of people can maintain it.
For instance Mocca is pretty damn cool but if the originators stop maintaining it
how many people know pascal (I think it's pascal) well enough to update it?

Boy wouldn't a small manual dedicated to GUI developers be Handy.
probably see a few more GUI come along.

I have not heard of anyone running steppers (or step/direction servos) with encoder feedback that was used in a PID loop.
I do see in another thread someone is trying...
One guy monitored the feedback and got EMC to throw an E-stop if following error got to big. (this was through parallel port) .
If someone can show me a working config and it does not cause me too much trouble to add it -I will.
Personally I don't really see the idea.
If you want ease of setup use step / direction drives.
if you want ultimate control use servos with feedback to EMC's PID
Other then doing it just to see if it works.....
To each their own.

cmorley wrote:

You are confusing pncconf by asking for stepgens with encoders for each axes.
encoder feedback with steppers is not supported out of the box.

It possibly ought to be, as step-direction servo drives are not all that uncommon (even if they do put the control loop a little further from EMC2s control than is really ideal)

Did this problem get resolved - and if so how? I find I have the same PCI-parallel chip - and the same problem.

thanks for any information

rde

I think I have it fixed. It was actually fixed in stepconf quite a while ago. Just not added to pncconf.
It will be available in the next bug fix or now as source - not sure what version your using now...
Thanks again.

Check this out:
www.cnczone.com/forums/showthread.php?t=108329
which will send you to here:
build.spaceopera.org/site/spectralight02...rtadapter/index.html

You are confusing pncconf by asking for stepgens with encoders for each axes.
encoder feedback with steppers is not supported out of the box.
assign a custom name for the encoders (such as "x-feedback" don't use spaces) and configure / connect them in the custom HAL file.
not sure if your looking just to be able to read the feedback or use it to stop on following error or even try to correct following error.

Yes Pncconf has not been translated yet. It's still not finalized.
I will see If I can fix your request.

Cheers Chris

Here at school we have an old light machines comppany super spectra light mill that has a shady stepper. We also have a lathe by the same company that has been converted to a mill that I am planning to get a replacement stepper from. The only problem is the computer with the software on it doesnt work anymore, but I am trying to get emc to drive it...does anyone have any idea where i can find the pinout?

thanks

Rex

This is probably a question best asked in the new Pncconf section of the forums.
However, I would be tempted to create some dummy pwms and then hand-edit the HAL to remove them again.

You want whatever is linked to the PWM inputs to connect instead to the stepgen inputs.

Looks like pncconf knows only English.
In Russian language "Desktop" written not so as in English.
After all configured, pncconf trys to write in "~/Desktop" not in Russian meaning of desktop.
I need to make simlink to Russian call of desktop and the "Desktop".
After that pncconf can write result ok.

Hi!
I need to configure 5i20 card. I have servo with step/dir pulses, but at the same time I have
3 signal (A,B,Z) from this drive.
I used pncconf. Configure 2 encoders (lathe), 2 step generators. But pncconf tell me "You foget assign pwm for X,Z".
What I need to do?