Designer’s Guide Community

Debopriyo Chowdhury Guest	Mixer Small Signal Model Oct 23^rd, 2005, 1:37am Hi, Could anyone just clarify if its possible to represent a mixer by a small signal model to calculate its conversion gain? Ifyes, how so I proceed to do it? Also how do I calculate theoretically its IIP3 or 1-dB compression point? Thanks and Regards, Debo
Back to top	IP Logged

vborich

Junior Member

Offline

Posts: 23

Re: Mixer Small Signal Model
Reply #1 - Oct 24^th, 2005, 1:12pm

S-parameter-like descriptions exist and are measurable by all RF simulators. But you have to be very careful with them. Unlike linear S-parameters, when you change the load and source conditions, the linear model changes because it's computed around the large-signal (LO) steady state, which is generally load- and source- dependent. For example, if you are designing an IF matching network based on models of this type, they are valid to the extent that the LO response is insensitive to changes at the IF port.

For details, see Microwave Mixers or Nonlinear Microwave Circuits by Steve Maas. The classic paper in this field is Egami's several decade old paper from IEEE Transcations on MTT.

Vuk

Back to top

IP Logged

Ken Kundert

Global Moderator

Offline

Posts: 2384
Silicon Valley

Re: Mixer Small Signal Model
Reply #2 - Oct 24^th, 2005, 2:23pm

Vuk seems to be referring to large signal s-parameters. However, it is also possible to apply true small-signal s-parameters to mixers. SpectreRF supports this with its PSP or periodic s-paramter analysis. In this case the s-parameters are like normal s-parameters, except the signal is a different frequencies at each port.

SpectreRF also provides the PAC and PXF analyses, which are also true small-signal analyses.

So yes, it is possible to have a small signal model for a mixer, but you must somehow account for the frequency translation.

For information on iIP3 and CP1, try http://www.designers-guide.org/Analysis/intercept-point.pdf.

-Ken

Back to top

IP Logged