The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Mixed-Signal Design >> Sigma Delta modelling
https://designers-guide.org/forum/YaBB.pl?num=1268408711

Message started by jefkat on Mar 12^th, 2010, 7:45am

Title: Sigma Delta modelling
Post by jefkat on Mar 12^th, 2010, 7:45am

Hi ,
I am working on modelling the sigma delta modulator (SDM). I used the simulink tool available on the Mathworks web. I am particularly interested in 2-1 MASH. This tool fails miserably for such a sim. It shows no improvment when you cascade the last 1st-order modulator.

Anyone with such experience ? any precious advice?
Thanks much for your time

Title: Re: Sigma Delta modelling
Post by loose-electron on Mar 12^th, 2010, 1:21pm

why dont you just build your own model from primitives and go from there (switches, ideal DAC's ideal comparators, etc)

Title: Re: Sigma Delta modelling
Post by Berti on Mar 15^th, 2010, 1:03am

What's this tool you are referring to?

Is it a Toolbox?

Title: Re: Sigma Delta modelling
Post by jefkat on Mar 15^th, 2010, 11:37am

Its a quasi tool I would say ..May be I am making some mistake. I am trying to find out..

Loose electron.. What you are saying is certainly one way of doing things but some high level sims in the tools like Matlab are quick and easy , as you already know.

any one? :)

Title: Re: Sigma Delta modelling
Post by loose-electron on Mar 15^th, 2010, 3:05pm

what I am suggesting is a high level matlab type simulation using ideal elements.

Its a sampled time system, so you can -->not<--do it in the S-plane or something like that.

You could do it as a Z-plane simulation, but frankly I would just do the ideal system as a behavioral time domain solution.

Title: Re: Sigma Delta modelling
Post by sheldon on Mar 15^th, 2010, 8:42pm

Jefkat,

Building high level models should be possible, it can also be
done ADE/Spectre. Only a few primitive cells are required:
delay, summer, gain, and quantizer should be enough, as
loose-electron pointed out. Using this approach you should
be able to directly realize the target transfer function and
build anything you need. In the attached sample is a second-order
loop. Using this approach, I have built first order, second order,
third-order MASH, band-pass, etc. They all work and show the
expected behavior.

Best Regards,

Sheldon

Title: Re: Sigma Delta modelling
Post by Berti on Mar 16^th, 2010, 1:45am

Dear Jefkat,

Quote:

Its a quasi tool I would say ..

Frankly, this is a very poor answer. You can't expect people to help you when you can't even explain which tool you used for the simulations.

I argree with Jerry and Sheldon that behavior simulations with e.g. Cadence is a good alternative to Matlab, even though I prefer Matlab because it provides many functions to nicely evaluate the results (FFT, plots etc.)

Cheers