Mixer gain: Gain_Power? Gain_Vol? Gain_Vol_dB?

[yolande_yj]

If I have a spec of mixer conversion gain about 8. What does it mean?
Does it mean power conversion gain: 10×log(P_IF/P_RF) ?
Or voltage conversion gain: V_IF/V_RF or mag(V_IF/V_RF)?
Or voltage conversion gain in dB: 20×log(mag(V_IF/V_RF)) ?

What is the most common definition of mixer conversion gain?

Thanks.

 

[gszczesz]

It depends on the design environment for which it is being designed:
-In RF/Microwave circles, everything is specified in power. Gain is a power gain, and input/output impedances are specified (i.e. 50 Ohms input, 200 Ohms output, etc...)
-In slow speed analog circles (i.e. Power managment, etc...), things are always specified in terms of voltage gain. The output impedance is low and voltage gain is not affected by the load impedance. This is actually bad for mixers because they are open-loop devices with a high output impedance, which means your output impedance sets the gain (gm*Rout). So what some do is specify voltage gain for a given load impedance, else they specify the transconductance (gm) if the mixer does not have a low impedance output stage.

Greg

 

[yolande_yj]

In Low-IF or Zero-IF receiver, I have seen the down conversion voltage gain of 10db. Why they define voltage gain in db? Is it feel better than a voltage gain of 3?

 

[gszczesz]

Since these circuits tend to have high dynamic ranges and very different gain levels from part to part, dB works a lot better.

Also when the system guy takes the data and tries to figure out stuff like dynamic range, dB is much easier to work with.

 

[yolande_yj]

But since other blocks like LNA use power gain, how to calculate gain, NF, IIP3, P1dB... with power gain and voltage gain directly? If can not, it is no point to use db for voltage. Am I correct?