0
\$\begingroup\$

I want to analyze the effect of noise on a simple transistor circuit using HSPICE, while maintaining process variation.

The basic setup is as follows:

  • The HSPICE models for the transistors are given by the foundry, so I do not want to change anything in the model files (some of them are encrypted as well)
  • These model files require a Monte Carlo simulation to incorporate the process variation. I cannot choose the parameters but just performing a Monte Carlo simulation selects the parameters randomly.
  • I also want to include additional noise on the output nodes. To perform this, I am utilizing .TRANNOISE to add transient noise to the nodes.
  • The transient noise itself is provided through a Monte Carlo sampling

To include process variation through the model files, I need to use .TRAN 1ps 10ns SWEEP MONTE=100 SEED=19

To include the transient noise, I need to use .TRANNOISE V(out) SAMPLES=100 SEED=4

Now my question is this:

How do I combine both of these individual Monte Carlo simulations into one single simulation while maintaining separate control on the seeds?

If I try the following:

.TRAN 1ps 10ns SWEEP MONTE=100 SEED=19
.TRANNOISE V(out) SAMPLES=100 SEED=4

HSPICE throws an error.

I need individual seed control on the Monte Carlo simulation on both the process variation model and on the transient noise. So, it is imperative to have two independent seeds.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.