Xcos / Coselica Support

[QwillDune]

Follow-up from a previous topic:

Convert a XCOS model from Scilab 5.5.2 to Scilab 2024

In fact SIMM installs fine in Scilab 2024.0.0 (by on-the-fly compilation) but you have to install coselica before, as the dependency of SIMM upon coselica does not seem to be set in the metadata of the package. However, after satisfying this, there is a problem when SIMM loads: atomsLoad : Une erreur est survenue au cours du chargement de 'SIMM-0.6.2': exec: error on line #172: "Variable non définie : cos_utils_lib" S.

The tollbox SIMM needs to be updated to work with the newer versions of Xcos and Coselica.

While this may happen, we think our problem can be solved with only Xcos/Coselica.
We need help/training on how to do that.

The goal is to model a DC motor acting on a mechanism.
Input (example): Voltage PWM command, current sensing.
Output: rotation angle on the mechanism. Torque is dependant of position (angle)

If you know anyone who could help
Please let us know

[Pascal77]

Hello, I got the same problem on SCILAB 2024.1.0

COSELICA is installed correctly. Schematics can be created in XCOS.
However, SIMM creates problems.

It’s pretty annoying, all this talk of incompatible modules. DASSAULT SYSTEM, which took over the lead of SCILAB, should also have taken over the lead of the most widely used modules, namely :

• CPGE, COSELICA, SIMM at the very least.

As a result, I’m still using my SCILAB 5.5.2 version, which works perfectly with all these modules. It’s a real shame.

Translated with DeepL.com (free version)

[mottelet]

Hi Pascal,

I let @vincent.couvert answer you as he knows the Scilab story from the inside. Noboby decided one day to stop supporting these modules, the situation is more complicated…

S.

[Pascal77]

Hello Stéphane,

thank you for your answer. I hope something good will happen soon

Pascal

[vincent.couvert]

Hello @Pascal77 ,

Thanks for your feedback about Scilab (and particularly your post about uiimport).

We just updated the SIMM toolbox on ATOMS, this module should now work with Scilab 2024.0.0 (and recent 2024.1.0 version).

As @mottelet said, the situation is quite complicated for many ATOMS modules as the successive maintainers sometimes only posted binary versions on ATOMS (and did not contribute to the source code on Scilab forge for example) hence creating a difference between the available binaries and sources. Once in this situation, updating the toolbox is complicated and time consuming.

Development team is only composed of a few people and maintaining Scilab in already a huge work so we do not have lot of time for ATOMS toolboxes but we do our best. This is why contributions are welcomed from the users community and now Scilab (and most ATOMS toolboxes) source are available on Gitlab, contributing is very much easier.

Thanks again for using Scilab.
Vincent

[Pascal77]

Hello Vincent,

thank you very much for your message.

I’m well aware of the workload involved. I must say that it’s already a great satisfaction to see that SCILAB has been (re)taken in hand by competent and motivated people. The successive changes that have taken place in close succession are testimony to this. For my part, I’m not a developer, just a user, otherwise I’d have loved to be part of this renaissance.

And finally, how can I go about testing the new SIMM on the latest version of SCILAB?

Have a nice evening. Best regards,

Pascal

Translated with DeepL.com (free version)

[Pascal77]

I’ve just seen the update on the site https://atoms.scilab.org/toolboxes/SIMM/0.6.2.2

I will try…

Thanks again

[QwillDune]

Hello,

With SCILAB-2024.1.0 + Coselica 2024.0.0-1 + SIMM 0.6.2.2.
The toolbox loading runs fine and smoothly
Thank you

I tried again the model shared some months ago

Convert a XCOS model from Scilab 5.5.2 to Scilab 2024

Hello, A colleague of mine is using a Xcos model that will only run under Scilab 5.5.2. He and I tried to run it under Scilab 2024 without sucess. We are really rookie here. Can you help find what needs to be adjusted? Moteur à courant continu avec Scilab-Xcos - 20231006 - Scilab 5_5_2.zcos (5.9 KB) Moteur à courant continu avec Scilab-Xcos - 20231006 - scilab-2023.1.0 (64-bit) - SIMM NON COMPATIBLE.zcos (3.7 KB) Thank you F

I used the “5.5.2” model as the “2023” was giving error messages.

The work starts … compiling … linking … but stops with this message.
Link done.
WARNING: Diagram contains implicit blocks, compiling for implicit Solver.
WARNING: Diagram has been compiled for implicit solver
WARNING: switching to implicit Solver

• “SIMM post”*
• “Error in post_xcos_simulate: ending simulation.”*

Any clues ?
What can cause that?
Where should I look?

BR
F

[Pascal77]

I am trying the new version of SIMM in SCILAb 2024.1.0
Installation was fine. No encountered problems.

My first try is very simple :

All blocks come from SIMM palette.

We can see a problem regarding REP_TIME block.

Then I lauch the simulation, and I have an error message

When I click on OK button, I have errors messages in the console.

ATTENTION : Le diagramme contient des blocs implicites, compilation en cours pour le solveur implicite.
  "SIMM post"
  "Erreur dans post_xcos_simulate : Simulation interrompue."

So, it seems I get the same kind of errors than QwillDune.

Remark : As I am french, my SCILAB version is in french. Sorry for french texts in this post.

[Pascal77]

I just wanted to share a comment about file names regarding old SCILAB versions (for example in 6.1.1 version).
Indeed, files names are very important with COSELICA and SIMM pallettes.

Avoid accentuated characters and spaces in file names.

Otherwise, an error is generated and the simulation stops :

Once the name is correct, it’s possible to simulate a diagram (with a few restrictions). Here’s an example:

The working plots are given below: