The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> large signal noise figure
https://designers-guide.org/forum/YaBB.pl?num=1187188041

Message started by carlgrace on Aug 15^th, 2007, 7:27am

Title: large signal noise figure
Post by carlgrace on Aug 15^th, 2007, 7:27am

I am designing an LNA and I am interested in its large signal noise performance, that is, what is the NF of the LNA in the presence of a strong out-of-band blocker. I haven't done this type of simulation before. My thought is to use PSS in SpectreRF with the simulation set to a -23dBm blocker at 3MHz offset (for GSM). Then I will run a PNOISE analysis and divide the integrated output noise by the output-referred noise of the source resistor. Is this the correct way to address this issue?

thanks,
Carl

Title: Re: large signal noise figure
Post by ACWWong on Aug 16^th, 2007, 3:31am

Yes, if you run pss using the blocker, then pnoise (use pxf to ensure you get the LNA gain right, it should be reference sideband 0 for pnoise).
The pnoise sweep should be 3MHz offset from relative harmonic 1 (or absolute frequency pss freq+3MHz). Always check the spectre log as it tells you what its doing at the top of the pss or pnose run.
If you use your LNA source and load as ports (ususally of 50 source port and large load port resistance to mimic mixer load) then you can get spectre to calculate NF directly for you.
So you should see with a low power pss tone (ie small blocker) your NF will be the same as linear noise/sp analysis. When you increase the pss tone (for example to -23dBm or above) the NF will increase. you can print the noise summary (fill in the frequency relative if the pnoise sweep is relative) to see where the contributors are.

Title: Re: large signal noise figure
Post by Hunter_BJUT on Aug 27^th, 2007, 4:26am

no, muti-stimuli, use qpss

Title: Re: large signal noise figure
Post by ACWWong on Aug 27^th, 2007, 1:55pm

Hunter_BJUT wrote on Aug 27^th, 2007, 4:26am:

no, muti-stimuli, use qpss

In the case in question of an LNA with one big blocker, there is only 1 large tone, so qpss in not necessary.