advanced Noise simulation in Complex polyphase filter

[Chr's]

Hi,
First of all, I found some old topic about the question regarding the noise simulation, like this one:
https://www.edaboard.com/threads/148584/

My question is about a process to generate a noise simulation without all these steps, because do it for the worst cases and typical it's quite boring and spend lot of time.

Regarding my filter, it's a Tow-Thomas with a current input and voltage output, central frequency at -1MHz.
I found a test bench which uses only one noise source and generate signal for I and Q automaticaly.
The attached drawing explains clearly (I beleive :smile
I don't find any cells (ideal, ahdl, veriloga, ...) to do this phase shiffter (a combiner) and the V2I. I'm trying with vcvs and ccvs but the phase is set by a time delay (not like AC paramter like in vdc,vsin,idc,isin,...) which change during the noise simulation (because the noise uses an AC simulation). I want a cell with a constant phase shift during the AC simulation. If someone has an idea to do these cells, he's welcome. :-D
I don't how this guy did it. :sad:

Thanks in advance.

 

[Chr's]

Hi,
I'm progressing on this test bench, i've some coherente results but I would like to improve these. My issue is about the dephasor box. It's done with RC combinaisons to get 90 degree, but this 90degree is done only for one frequency.
I'm trying to use dependent sources but they've no AC parameters. Anyoneone knows how to creat a dephasor with constant phase across the frequency ? even in verilogA.