[sdiy] backwards pots. Whats the deal?

Neil Johnson neil.johnson71 at gmail.com
Thu Apr 2 22:34:51 CEST 2015


Hi Tom,

> Ok, I'm convinced too. So why isn't this more widespread, if it's not actually a new idea? Most mixer circuits you see ae the "classic" design. If there's an already existing "improved" design, why isn't it more widely adopted?

Well, it's not so much an "improvement", rather a similar-looking
circuit with behaviour that in some applications is better, and in
others is not so good.

For example, if constant input impedance is critical to your circuit's
operation, then an input buffer would be far better.
If noise is your concern then this pot arrangement doesn't necessarily
give better (indeed, can be worse) noise performance than the
pot-first approach.  Although if noise was your concern then you'd be
considering different solutions.

Think of it as yet another spanner in the toolbox of solutions.  For
cost-sensitive CV summing I contend that it is better than the
traditional approach.  But, say, for audio mixing I would look for a
low noise solution (input buffer, then low-value pot in the
traditional connection).

To understand why some designers use it, while others don't, you need
to understand the design process.  And that also includes the context
of developing a commercial product.  The kinds of trade-offs change
when you're making 10,000s, or more, of something.  You can end up
playing funky tricks to get the cost down.

As for "rethinking the fundamentals", I think it is more a case of
there being still so much to learn about the basics.  For example, it
was only relatively recently (1991 - 30 years after the IC op-amp was
invented) that a paper was published that finally sorted out the
relationship between resistor tolerance and diff.amp CMRR - this page
at ADI gives a good overview:

http://www.analog.com/library/analogdialogue/archives/48-02/diff_amp.html

and lists Pallás-Areny and Webster's 1991 paper.  And for anyone
interested some of the key points in that paper are also discussed in
their textbook "Analog Signal Processing" which is not a bad read
(annoying in places, but has some interesting discussions and points).

Cheers,
Neil
--
http://www.njohnson.co.uk



More information about the Synth-diy mailing list