Just to summarise the what I’m working on: I’m trying to model ultrasonic transducers (CMUTs) using Virtuoso Spectre. The transducer has a small signal circuit equivalent and a large signal circuit equivalent. I have a working small signal simulation which utilises basic circuit components. The large signal circuit equivalent is more complex: for the Cadence simulation, I use a Verilog-A block to model the transducer and I need to use Harmonic Balance to simulate the large signal model. Both large signal and small signal models get radiation impedance data from n-port blocks. The radiation impedance is an S-parameter data file which governs the changing impedance of the transducer at specific frequency points in the form of “freq: z1:1(real,imag)”. The small signal simulation can employ this n-port without problem and gives correct AC simulation results (verified by comparison to existing results from other simulators).
My question is the following: When I run Harmonic Balance to simulate the large signal transducer model using a linear resistor instead of the n-port under small signal conditions, the results of the simulation match the small signal model simulations. In other words the Large signal model simulation works. However, once I plug in the n-port, the results do not match anymore. I tried using simple data in the n-port and it seems that the data is read by the Verilog-A block, but even with the simpler data large and small signal model results don’t match. I’m wondering if there is something I’m missing? For example is there some setting that that I'm overlooking which will enable the n-port data to be processed correctly? Or maybe the complex number impedances are causing trouble (though I can’t see how that could be seeing as the small signal model uses the same data?).