Hello,

I am running a simple simulation with the Amplifier2 model. It has a 50 ohm single frequency source and and a 50 ohm output. I am running an HB simulation and would like to have the GainCompPower term change as a function of frequency that utilizes the "freq" variable. I have tried using equations like pwl(freq,...) and cpt = f(freq) and also tried reading data from a DAC to no avail.

The simulation always wants to grab the low frequency (1 MHz) point and use that in the harmonic balance as if it were at another frequency. It does not want to use the GainCompPower term given for any given frequency.

Also, it does appear to work fine if I set

S21 = f(freq). The correct gain is sent to Amplifier2. This is not the case for GainCompPower.

Thanks for your help,

Lou

I am running a simple simulation with the Amplifier2 model. It has a 50 ohm single frequency source and and a 50 ohm output. I am running an HB simulation and would like to have the GainCompPower term change as a function of frequency that utilizes the "freq" variable. I have tried using equations like pwl(freq,...) and cpt = f(freq) and also tried reading data from a DAC to no avail.

The simulation always wants to grab the low frequency (1 MHz) point and use that in the harmonic balance as if it were at another frequency. It does not want to use the GainCompPower term given for any given frequency.

Also, it does appear to work fine if I set

S21 = f(freq). The correct gain is sent to Amplifier2. This is not the case for GainCompPower.

Thanks for your help,

Lou

Amplifier2 is replaced with a 3 separate Amplifier2's. The first one has

S21 = -(Compression point)

The second one is Compression =0 referenced to INPUT and s21 =Gain

The third one has Gain = Compression point.

I can then specify, CP = pwl(freq, ...)

as well as Gain=pwl(freq,...) and I get the expected output

Referencing to the INPUT must force the pwl equations to be evaluated first. (that is prior running the HB simulation)

Why the difference ??Lou