Low noise analog bandpass filter topology

Is amplifier noise multiplied by Q about the best that can be done?

Unh, yeah, just after I hit SEND I noticed the word "active" was left out.

For the direct implementation - yes. However you can use a switched cap or a superhet approach.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

Yeah, you get to choose f and Q.

Isn't that enough?

George H.

Oh, and the number of poles etc.

Switched caps seem noisy. Is superhet where you reset an RF oscillator and restart it from the signal, (or the noise)? If so, I've still got to try that.

George H.

Imagine a basic lowpass RC filter. Now put two capacitors instead of one and switch them with the center frequency of interest. This makes for entirely passive narrowband filter at arbitrary frequency. For the full filter, you will need two 90-degree shifted channels like that. Noise performance could be similar to that of the chopper stabilized opamps, that is less then 10nV/sqrt(Hz).

Downconvert - filter - upconvert back. That allows for lower Q filters.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

As I demonstrated in...

"ANALYSIS AND DESIGN OF INTEGRATED CIRCUITS" MOTOROLA Series in Solid-State Electronics ©1967 Library of Congress Catalog Card Number: 66-25486

Chapter 15 (Which I wrote :-) pp470-473 ...Jim Thompson