QucsStudio

[djsb]

Thought people might be interested in this experimental project to combine QUCS (a free spice like simulator) with Kicad.

http://www.mydarc.de/DD6UM/QucsStudio/qucsstudio.html

It's in it's early stages but maybe worth a try (and expressions of interest, encouragement to the author)

I feel more of these projects should get together.

David

[djsb]

Now at version 1.3.1

http://www.mydarc.de/DD6UM/QucsStudio/qucsstudio.html

Check it out.

David.

[TerminalJack505]

I haven't heard of this project.  I'll have to check it out.

I've been using QUCS for a while and I'm fairly happy with it.  It does have its problems, though. 

Sometimes a circuit will fail to simulate and either just hang or give a 'jacobian singularity.'  Oftentimes, the 'fix' is to simply add one to the the number of steps in the simulation and try again.  And keep adding one to the number of steps until either the simulation runs or you give up.  If that doesn't work then most likely there is a floating node somewhere that needs to be pulled to a either a voltage source or ground through a largish resistor.

Also, a lot of circuits that would work in the real world won't actually simulate in QUCS without putting in a bunch of BS components and/or changing component values to not-so-real-world values.

I get the impression that the QUCS project has been abandoned.  The last update was over a year ago.  If that's the case then hopefully this new project will continue where QUCS left off.

[amspire]

David, thanks. That looks excellent.

I will give it a try sometime in the next week.

Can you use Octave with the results of an analog simulation, or are they just separate programs run from the one IDE?

Richard.

[djsb]

Just thought I'd post an update.

Currently at 1.4.4 and there are 2 video's showing how to use it here.



Changelog here

http://www.mydarc.de/dd6um/QucsStudio/ChangeLog

Download from here

http://www.mydarc.de/dd6um/QucsStudio/qucsstudio.html

David

PS Here is a pdf of a presentation presented in India

http://www.mos-ak.org/india/presentations/Brinson_MOS-AK_India12.pdf

[amspire]

Initially this looks very nice.

It has file based voltage and current sources.

I know it is possible to use file based sources in LTSpice - as long as you convert them to a WAV file first, but this directly accepts CSV files, and it offers a selection of interpolations (linear, cubic spline, etc) if you need it. The WAV format is integer only that I think LTSpice then automatically scales to make the final amplitude 1.0 Volts or something., so it is not great at representing explicit floating point values.



This means you can easily take a DSO waveform capture, and immediately use it as a source voltage in a Spice simulation. It also means you have the full mathamatical and analysis capability available to analyze a captured waveform.

Now that is definitely good.

Pity the library is so poor. If I decide to use it regularly, I think I will have to spend a day adding a lot of common parts like 2N2222A, 2N2907A, 2N3904, TL074 opamps, LM358, TL431 references, etc. The library only contains a total of 7 opamps.

Richard.

[amspire]

I just tried QucsStudio. It has promise, but it still shows signs of opensource madness.

Well, I am not even sure if it is opensource. There are no links to the source that I could find, and it is not compatible with the Qucs project it is based on.

I tried the file based voltage sources - they could not read dat or cvs files. "No data" apparently. So either the feature is broken, or they have changed the format of both the dat and cvs files with absolutely no documentation, or even source code I can go to.

The Qucs file based voltage source example does not load (because of changes in the QucsStudio format), and even when you fix it so it can load, QucsStudio definitely reduces to understand the DA and CVS files.

So I will probably have to go back to the clunkier open source Qucs to do my tests. At least it works.

It is not a version of Spice, by the way - they explain why trying to base an opensource simulator on Spice is a problem. So it is not Spice netlist compatible, but they do have converters. The potential to combine analog simulation, VHDL compatible digital, S parameters, electric field simulation along with the mathematical power of packages like Octave is awesome, as long as they can end up with something that doesn't take 6 months to master.

Right now, it is good, but it does not come close to LTSpice for ease of use, documentation and graphics.  The KiCad-based schematic interface is pretty good. Could even be a little better then LTSpice.

Richard

[timofonic]

10th   April   2015     QucsStudio Version 2.3.0 released and examples updated

Any experienced EE tried it or would like to give it a look?

Moderators: Please move it to Open Souce tools

[timofonic]

8th   August   2015     QucsStudio Version 2.4.0 released and homepage updated

Any reply? Please move it to Open Source tools section...

[AF6LJ]

Is this posted somewhere one need not create an account in order to download? 

[timofonic]

Is this posted somewhere one need not create an account in order to download? 

You can download the Windows binary, no binaries provided for other platforms. The source code seems to not be publicly available, but maybe someone knows where to locate it.

This is the new site URL:

http://dd6um.darc.de/QucsStudio/qucsstudio.html

Anyway, the author forked QUCS but doesn't provide the source code. I wonder if this is legal or not.

ADMS (a compiler of Verilog-AMS in XML form to C with Spice compatible API, to be able to use Spice-based simulators) is LGPL 2.1 and QUCS is GPL 2.

Can anyone do some analysis? If it's true, I would be very interested that the QUCS Team contacts with some organization such as FSFE's Freedom Task Force , Software Freedom Conservancy or Groklaw


http://www.gnu.org/licenses/gpl-violation.html
https://fsfe.org/activities/ftf/reporting-fixing-violations.en.html
http://sfconservancy.org/blog/2012/feb/01/gpl-enforcement/
https://softwarefreedom.org/resources/2008/compliance-guide.html
http://www.gnu.org/philosophy/enforcing-gpl.html
https://sfconservancy.org/news/2015/jun/01/conservancy-seeks-linux-enforcement-questions/

[AF6LJ]

Thanks..

[timofonic]

Thanks..

You're very welcome!