Large Signal / Linearity data structure

35 views
Skip to first unread message

Curtis Mayberry

unread,
Jun 2, 2021, 2:28:31 PMJun 2
to scikit-rf
Would there be any interest in adding a large signal or linearity class that could take in harmonic data and then allow the user to plot or measure it in different ways?

The first use case I have in mind is reading in the results of a harmonic balance simulation and then plotting the 2nd and 3rd harmonic power level over input power at a specific frequency and VSWR.

Some future applications could include:
- Plot 2F0 / 3F0 harmonic power over frequency.
- Two Tone Tests
  - IMD histograms
  - IP3 measurement and plots.

Regards,
Curtis

Julien Hillairet

unread,
Jun 3, 2021, 4:11:28 AMJun 3
to scik...@googlegroups.com
Dear Curtis,

Certainly! I don't know much on this topic. Before implementing something, it may be useful to think twice about the best/more convenient way to implement this. You can open an issue on github to gather the discussions here.

Best regards,

Julien

--
You received this message because you are subscribed to the Google Groups "scikit-rf" group.
To unsubscribe from this group and stop receiving emails from it, send an email to scikit-rf+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/scikit-rf/df7a4ce5-98ef-4cfc-8274-7ebb171bdb89n%40googlegroups.com.

alexander arsenovic

unread,
Jun 3, 2021, 8:08:21 AMJun 3
to scik...@googlegroups.com
sounds like a good addition to me, and juliens github issue suggestion is good.  
also fyi, modular additions are easiest to integrate, so if you can dev this in its  own file, it would be most quickly merged.
look forward to it. 

alex

Vikram Sekar

unread,
Jun 3, 2021, 9:25:31 AMJun 3
to scik...@googlegroups.com
This would be an interesting addition to me since I work with this kind of data too. I'm curious if you are planning on making it tool specific for harmonic balance, because different tools tend to output this data in different formats.

-vikram

Matt Braunstein

unread,
Jun 3, 2021, 9:40:16 AMJun 3
to scik...@googlegroups.com
I agree that this would be a great addition.  But you will only really be able to plot the data that exists within what is loaded. To get other parameters typically requires running a harmonic balance simulation. 

Just keep in mind that this will be exponentially more difficult to implement. There’s a reason that it is called nonlinear. This includes just doing any harmonic balance solving. You will probably have to implement multiple solvers and provide access to many solver options in order to deal with any convergence issues. 

What would probably be much more useful without implementing a harmonic balance solver, would be managing, plotting and analyzing load pull data. I think that would be a better start towards nonlinear analysis. 

Matt

Curtis Mayberry

unread,
Jun 3, 2021, 11:02:42 PMJun 3
to scik...@googlegroups.com
Hey Everyone:
Thanks for the excellent suggestions!
I'll create an issue this weekend and we can discuss some more there.

-Curtis

You received this message because you are subscribed to a topic in the Google Groups "scikit-rf" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/scikit-rf/a9vIXGQ5pWo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to scikit-rf+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/scikit-rf/14bf319a-9cf4-4398-abf0-e6549ad4b7b5%40Spark.


--
Curtis Mayberry
Analog and RF Design Engineer
Reply all
Reply to author
Forward
0 new messages