Baseband phase noise plot ADS

[manchandaprabhav]

Hello Everyone,

I want to plot the baseband phase noise plot in ADS, i.e. I have two sources at the same frequency (with corresponding phase noises), which mix and give a noise at baseband.
For some reason, ADS is showing the baseband noise as zero.
I have attached the images of the noisecon setup and the output that I get.
PS: I use an ideal mixer block (mixer)

 

[RealAEL]

What do you mean by ideal mixer block? Do you mean the Mixer2 component? if so what do you have set as the mixer noise figure (default for this is 0db)? More detail about your test circuit would help to diagnose any issue.

 

[G4BCH]

Have you set the phase of the two mixer inputs to 90 degrees? if not then the output will not be sensitive to phase variations.

 

[manchandaprabhav]

@RealAEL I meant the MIXER block not the MIXER2 block. Yes the noise figure of the mixer is set to 0 dB, as the aim is to study the noise from the sources rather than the mixer itself.
@G4BCH, why is that? phase noise is a stochastic process, thus should show at the output irrespective of the phase difference between my input signals.

 

[G4BCH]

why is that? phase noise is a stochastic process, thus should show at the output irrespective of the phase difference between my input signals.

For two signals at the same frequency the difference frequency from a mixer or product detector is
Vo = K Cos (Phi)
where:
Vo is output voltage
Phi is phase difference between the input signals,
K is proportional to mixer gain and the two input levels
You want to detect the phase difference between the two inputs so you need the function
d Vo / d Phi = K Sin (Phi) , hence the 90 degree phase shift.
With both signals in phase the output voltage is a is maximum and proportional to the two inputs, I.e an amplitude detector. When the two signals are in quadrature the output zero independent of amplitude and any deviation is proportional to the phase difference.

 

[manchandaprabhav]

@G4BCH Yes, that is correct for the deterministic phase. But when we talk about phase noise, it should be random and not affected by the phase of the two signals. The output phase noise spectrum is displayed separately from the output power/voltage of the mixing products (pnmx, anmx).

 

[manchandaprabhav]

Closing this thread as I found the error. It was to do with the default values of the isolation between the LO to RF, RF to output, and LO to output. Thank you all.