The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> Noise Figure of double conversion system
https://designers-guide.org/forum/YaBB.pl?num=1317607193

Message started by vivarf on Oct 2^nd, 2011, 6:59pm

Title: Noise Figure of double conversion system
Post by vivarf on Oct 2^nd, 2011, 6:59pm

Hello

Does any want know how to simulate noise figure of double conversion system in Cadence using Spectre?

Is this possible to use Pnoise analysis? Here the system has two LO.

If the simulator does not support direct way to measure noise figure. Then could anyone suggest the alternative method?

Thank you.

Title: Re: Noise Figure of double conversion system
Post by rfidea on Oct 8^th, 2011, 7:04am

Interesting question! I think you can but the simulation time will be long. The brute force method would probably be to set up the PSS with two large tones, your two LO signals. You need to use the beat frequency as PSS fundamental. Maybe you should try to tweak your design to use a frequency for the high LO as an exact multipel of the low LO. This would save simulation time. The reference sideband for the pnoise would then be a high number since it is calculated reffering to the fundamental tone.

This is one way I can think of. Maybe Andrew Beckett can comment on this?

Title: Re: Noise Figure of double conversion system
Post by vivarf on Oct 9^th, 2011, 5:26pm

Thank you rfidea.

The problem is: two LO tones are so far different.
For example, if LO tones is 50G and 1G respectively.
In single conversion with 50G, the number of harmonic in PSS set-up should be at least 3 (then it will cover 150 GHz noise bands).
Now for double conversion with 50G and 1G, the beat frequency is 1G. So number of harmonic PSS set-up would be 150 to cover 150GHz noise bands.
The simulator could run forever :(.

Title: Re: Noise Figure of double conversion system
Post by rfidea on Oct 10^th, 2011, 11:20am

Yeah, I understand. Looks like this is a white empty spot in SpectreRF. Maybe you can use qpss but I think you are only allowed to have one large tone and I am not sure if you get the frequency translation in the second mixer if you use that LO as a moderate tone. But maybe you can try. At least qpss is able to just calulate the tones you like.