The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> SpectreRF Convergence problem
https://designers-guide.org/forum/YaBB.pl?num=1109363567

Message started by ic_engr on Feb 25^th, 2005, 12:32pm

Title: SpectreRF Convergence problem
Post by ic_engr on Feb 25^th, 2005, 12:32pm

I am encountering convergence problem for simulating my ADC which is a 2nd order SIgma Delta Modulator. There are two integrators and there is a quantizer involved. The quantizer output feeds back into the first integrator.

The convergence problem is seen in PSS. I am monitoring the internal nodes, namely the output of the 1st and 2nd integrators and there seems to be okay. They are not clipping towards rail.

Can someone suggest me a work around for this.

Thx

ic_engr

Title: Re: SpectreRF Convergence problem
Post by Ken Kundert on Feb 25^th, 2005, 2:37pm

To use PSS the circuit must be periodic. Sigma Delta converters are not periodic, they are chaotic. To simulate this circuit with SpectreRF you must break it into components and use SpectreRF to individually characterize each component.

-Ken

Title: Re: SpectreRF Convergence problem
Post by ic_engr on Feb 28^th, 2005, 6:10am

Ken,

Doesnt your paper on DEvice Noise Simulation of Sigma Delta Modulators and the paper "Simulating Switched-Capacitors" show examples of 2-nd order Modulators as simulated using SpectreRF. There was no mention of breaking the blocks.

Having a clock at 2MHz and being synchronous with clock I though PSS should have handeled this.

ic_engr.

Title: Re: SpectreRF Convergence problem
Post by Ken Kundert on Feb 28^th, 2005, 9:33am

Actually it does require that you break the Delta-Sigma converter into its component pieces, analyze them separately and then combine the results.

-Ken

Title: Re: SpectreRF Convergence problem
Post by ic_engr on Feb 28^th, 2005, 12:41pm

Ken,

Can you seuggest me what I should do for the feedback from the quantizer to the integrator once the loop is broken.

Thx

ic_engr

Title: Re: SpectreRF Convergence problem
Post by Ken Kundert on Feb 28^th, 2005, 4:19pm

The best suggestion I have is to carefully read the paper.

-Ken