Designer’s Guide Community

Charles CHENG

Guest

Noise Figure in the LNA+Mixer+VCO
Nov 11^th, 2003, 1:32am

Dear all,

I have tried to use Spectre to simulate the Noise Figure of the the LNA+Mixer+VCO together using the osc. model with phase noise. There are some effect I am not clear. I hope anyone could answer me. Thanks. After the simulation, I have plotted the NoiseFigure and get the NF=7.1dB at 100MHz (I am using RF=1.8GHz and LO=1.9GHz with phase noise). However, there are a small hill at 200MHz in the NoiseFigure plot. If I am just simulating the LNA+Mixer with ideal input LO and RF source. There are no small hill at 200MHz.
Actually, why there are small hill at 200MHz. Could anyone explain it? I am very thanks for your helping.

Best Regards,

Charles CHENG

Back to top

IP Logged

BSIM4

New Member

Offline

Posts: 8

Re: Noise Figure in the LNA+Mixer+VCO
Reply #1 - Nov 17^th, 2003, 7:36pm

I will take a guess here since more details would help to identify the problem.

The little bump might be caused by the image of the mixer. Why you do not see it with perfect LO? I think it is because with larger steps and a very, very sharp LO, your frequency sweep simply stepped through the peak at exactly 200MHz. When you used a real VCO, reciprocal mixing widened the peak at 200MHz, so you can see it with your current sweeping set-up.

To check if this is true. add 200MHz point to your frequency sweep setup and see if you can see the peak caused by imaging.

Back to top

IP Logged

Ken Kundert

Global Moderator

Offline

Posts: 2384
Silicon Valley

Re: Noise Figure in the LNA+Mixer+VCO
Reply #2 - Nov 18^th, 2003, 10:18am

The VCO phase noise will modulate all of the harmonics and spurs present in the circuit. So the question is not "why the little hill", but rather, why is there a signal present at the 2nd harmonic of the output signal.

BSIM4 is right to ask about your frequency point selection. Often these little hills exist and are not seen because they fall between the analysis points.

It is not surprising that the output has a 2nd harmonic. In fact if you adjust the frequency point selection, you will probably see little hills at the 3rd harmonic as well.

-Ken

Back to top

IP Logged