Zero feedback push-pull buffer : piezo driver idea

[rjm]

REVIEW OF SCIENTIFIC INSTRUMENTS 80, 104701 2009 permalink

(the paper is presently a free download, which is cool)

I came across this as part of my "day job", but I suspect the author does audio design on the side, much like I do. It's an amplifier design for driving a piezo stack typically used in micropositioning applications. It has to output significant power into high capacitance loads at high frequencies. The author's solution was to bolt on a large current buffer onto the output of a simple op amp, what struck me as interesting was the relative elegance of the circuit, and the absence of feedback on the output stage. Feedback, it is claimed, actually reduces the circuit bandwidth... the analysis is a bit beyond me, I admit.



It is claimed that the circuit has high linearity, but would it make a good audio amplifier or not? The idea of a current buffer for an output stage is something I've played around with a bit in the form of MOSFET source followers, and liked very much... the circuit above would be the ultimate extension of that train of thought.

(a couple of ideas: use a power op amp like the LM3875 as the input, for a "power gainclone", or make a low power miniature version with a regular op amp for a pre-amp or headphone amplifier...)

[peranders]

You don't have to be an audio designer to design something like this because this only basic electronic knowledge.

The power stage is pretty conventional and the distortion will not be very low.

[rjm]

Originally Posted by peranders

... the distortion will not be very low.

That's what I was afraid of. Thanks!

[Frank Berry]

You can reduce the distortion by taking the NFB from the junction of R10/R13 rather than from pin 6 of U2. The available power output will be limited by the Vcc/Vdd voltage limit of the Opamp.