The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> SpectreRF 'tran' and 'pss' problem
https://designers-guide.org/forum/YaBB.pl?num=1109960447

Message started by danda821 on Mar 4^th, 2005, 10:20am

Title: SpectreRF 'tran' and 'pss' problem
Post by danda821 on Mar 4^th, 2005, 10:20am

Hi,

I am using Cadence to simulate a oscillator. I got the following simulation log:

1. No DC path from node 'I24.CM4.1' to ground. Gmin installed to provide path. (this is cap connecting two transistor gates)

2. Notice from spectre during IC analysis during transient analysis 'tran' :
Initial condtion computed for node Vp is in error by 11.9793mV (1.19793k%)
.......
to reduce error in computed initial conditions, decrease 'rforce'. However setting rforce too small may result in convergence difficultis or in the matrix becoming sigular.

3. Warning from spectre at time = 16.0925ns during periodic steady state 'pss'
I17.T3: Vgs has exceeded the oxide breakdown voltage of 'vbox'=1.6V
....
I16.T5: The bulk-drain junction current exceeds 'imax'

The transient analysis can go through. But the pss failed to converge. I don't know what is wrong.

thanks.

Title: Re: SpectreRF 'tran' and 'pss' problem
Post by Ken Kundert on Mar 4^th, 2005, 11:48am

The first two warnings are generally nothing to worry about; they would not affect convergence of PSS.

The third is probably more an indication that PSS is having trouble converging rather than the cause of the trouble.

For PSS to converge, your circuit must have a periodic solution. If you are confident that it does, then it is generally a matter of setting tstab large enough. If you have set tstab so large that all the signals have reached their periodic steady-state behavior, and PSS still does not converge, then there is some other problem.

-Ken

Title: Re: SpectreRF 'tran' and 'pss' problem
Post by danda821 on Mar 7^th, 2005, 12:31pm

Thank you . Ken.

The problem is sovled. I think it is because there is on periodic solution of the circuit.