Designer’s Guide Community

uncle_ezra

Community Member

Offline

Posts: 90

Mixer Noise Figure vs LO
Jul 19^th, 2005, 8:21pm

So I configure PSS to sweep LO from 4.8G to 5.8G with output harmonic set to 0. Then I set Pnoise to do relative to 1 single point sweep @ 1M. For reference sideband I use -1. When I plot the NF it gives me some huge number.

I then run PSS with LO fixed @ 4.8G and sweep Pnoise from 1M to 15M. When I plot the NF it gives me what I expect.

So how to configure PSS and Pnoise to sweep LO signal and plot NF? Also how to configure to sweep LO power and plot NF?

Thanks

Back to top

IP Logged

Ken Kundert Global Moderator Offline Posts: 2384 Silicon Valley	Re: Mixer Noise Figure vs LO Reply #1 - Jul 20^th, 2005, 12:19am Did you use a relative sweep? If so you are measuring the noise figure for an up conversion mixer. Is that what you intended? If you did not use a relative sweep, then your settings look right for a low-side down conversion mixer. -Ken
Back to top	IP Logged

uncle_ezra

Community Member

Offline

Posts: 90

Re: Mixer Noise Figure vs LO
Reply #2 - Jul 20^th, 2005, 2:29am

So for downconversion you dont use relative in Pnoise? You just use absolute? Why is this so?

I tried setting 0 instead of 1 for relative harmonic and the plot looks reasonable. Also for conversion gain plot versus LO I set 1 for relative harmonic in PAC and the result seems right.

So what exactly is relative harmonic, when should you and should not use it?

Thanks

Back to top

IP Logged

Andrew Beckett

Senior Fellow

Offline

Life, don't talk to
me about Life...

Posts: 1742
Bracknell, UK

Re: Mixer Noise Figure vs LO
Reply #3 - Jul 20^th, 2005, 4:02am

If you sweep the pnoise in relative mode, relative to harmonic 1, it means that you will be looking at the noise at 4.8G+1M to 5.8G+1M.

What relative sweeptype on the small signal analysis does is specify that the frequency sweep you entered is shifted by n times the PSS fundamental where n is the relative harmonic number.

Since with pnoise and pxf you would be entering the output frequency range, this would only make sense with an up-conversion mixer. For pac you're entering the input frequency range, and so it's quite reasonable to use a relative sweeptype there for a downconversion mixer.

From IC5141 USR2 (out soon), the small signal forms have had a minor enhancement that I suggested, which is to add the word "Output" before Frequency Sweep Range on the pnoise/pxf form, and "Input" before the Frequency Sweep Range on the pac form - to make it more obvious to casual users...

Regards,

Andrew.

Back to top

IP Logged