[measurement] LNA+buffer, how is actual power gain of LNA?

[wccheng]

actual result lna

Dear all,

I have a question about the measurement of LNA. I hope anyone give me a help. For my design, I just construct a CMOS LNA. Because of measurement, I have added one buffer at output and get a output in 50 ohms match. So, my circuit is LNA+buffer with input and output are 50ohms matching for measurement. If I want to know the power or voltage gain of the LNA alone, how could I do it?

I think that I should make one standalone buffer. Then measure its' gain. However, the input of the buffer is not 50ohms. For the case of LNA+buffer, the LNA output is not 50ohms. I cannot pretend the network analyzer source (50ohms) + buffer. Therefore, how could i cancel out the effect of buffer in order to get the real LNA performance?

Thanks

wccheng

 

[Mazz]

This is a classicl problem for RFICs.

I think that the only think you can do is to accurately fit the measurement with simulation (so modify the design if they doesn't fit), maybe with more than one single case (for example with different supply voltages, at different temperatures, measuring also linearity and so on..).
Once the meas and sim fit, your LNA is known.

I hope it can help.
Mazz