Stereo Memory Man puzzler

[bean]

I'm doing a re-draw of the SMM schem and ran into something a bit odd which I was hoping to get some insight on. I'm including the relevant pic below.

My question is about the Trim3 gain control before the output filter. It looks like they set it up to bleed signal off to virtual ground and I'm wondering why they did it this way at all. ISTM, it would make more sense to have it be a voltage divider tied to +7.5 and eliminate R32 altogether (with R33 tied to the wiper and adjusted the value as needed). Or just swap it out for R35 or R36 to control the output of U2B. Or, maybe it's that they wanted the filter stage to always be fixed? It just seems odd to me so maybe there's something I'm missing here.

[PRR]

> swap it out for R35 or R36 to control the output of U2B.

R35:R36 sets the Q of the filter formed by R22 R34 C25 C26. You don't want to also control gain there.

The 100k:10k divider may be more "gentle" than a straight potentiometric connection. Since half the plan I cut off I dunno what they are trimming for, small errors or large.

[Scruffie]

The choice seems to be based on 1) Giving the filter a bit more headroom by dividing before it 2) So the filter isn't dependent on the BBD output voltage for biasing, again, more headroom (otherwise it might be controlled by R31) and 3) R32/33 are part of the filter, so it's always going to be at least ~111k and you'd want it as close to 120k as possible for that filter to be set where it was obviously designed (equal value filters making Q calculations simpler).

[bean]

Got it, that's making sense. Thanks guys.