[Gnucap-devel] PSS analysis

[henrik johansson]

Hi Al,

I have played around some more with the Python extension for Gnucap to
see what can be done and what its limits are.

I've made a simple AC analysis where you can access the internal
matrices and vectors as numpy arrays.

The second example is an attempt to make a shooting Newton PSS
analysis. My reference was "Steady-State Methods for Simulating Analog
and Microwave Circuits" by Ken Kundert.

In order to build the Jacobian for the shooting newton iterations one
need the transient Jacobian and the trans-capacitance matrix for each
time instant. These are
obtained by using a subclassed Transient analysis where these are
saved. The transient Jacobian is easy but the trans-capacitance matrix
I cannot calculate so easily.
I tried running to calculate the AC-admittance matrix using doac and
acload and then taking the imaginary part  while setting the omega to
1rad/s but it must
be a better way to do this.

The second problem was to use the calculated voltage-vector as an
initial condition for the next transient run. Just setting the
SIM::vdc vector didn't help since the charges/fluxes  are stored
inside the reactive elements.

What do you think Al? Is there a better way to do this?

/Henrik