Common-Source JFET noise

The voltage noise density in nV/sqrt(Hz) is referenced to the input, so you need to multiply by the stage voltage gain e.g. gm*Rd

Thanks, That's in reference to R (resistance) and not reactance, right? So if I only have 22 ohms resistance and 70000 ohms reactance and gain of 100 then the total output voltage noise would equal thermal noise of the 22 ohm resister, which would be

Vn = sqrt(4 k T B 22) * 100 = 60 nV/SqrtHz

Darn, that's noisier than I thought it would be.

Paul

OK, assuming my above calculated 60nV/SqrtHz noise is correct, that's amplified noise from the source resistance (not reactance), but what about the JFET's noise? If it's 1 nV/SqrtHz then is that also amplified?

Thanks, Paul

The reactance doesn't contribute noise.

Originally, I thought you were talking about the equivalent input noise voltage of the JFET itself, as quoted on the manufacturer's datasheet. This will contribute more noise than your 22-ohm resistor. For a 2N3819 it's about 6nV/sqrt(Hz).

Input noise due to 22-ohm resistor = 0.6e-9 e_n for JFET = 6e-9

Total input noise = sqrt(0.6*0.6 + 6*6) = 6.03 nV/sqrt(Hz)

Ouput noise = 603 nV/sqrt(Hz)