# [sdiy] CV mixer trouble

David G Dixon dixon at mail.ubc.ca
Wed May 16 01:04:03 CEST 2012

```With a circuit like this, you really complicate things by making the offset
adjustable.  I would just use 1% resistors, assume you've got a reasonably
accurate rail, and take what you get for 5V.

So, now, the question is, how do you get a 5V reference from the -15V rail?
Well, the answer is very simple!  Use three resistors equal to the feedback
resistor in a Y configuration: one to -15V, one to GND, and one to the
inverting input of the summer.  This will give you +5V out of the summer.
(That Thevenin shit really works!)

However, in this case, you should not really be all that concerned with
getting exactly +5V out of the summer.  What you should really be concerned
with is offsetting the CV input by the equivalent of -5V from the
perspective of the CV input itself.  In other words, if your CV is 3V, you
really want the summer to be seeing the equivalent of -2V. Hence, if you use
100k 1% resistors for the Y and for the CV input, you will have accomplished
the job already, irrespective of the actual feedback resistance (and,
therefore, the gain factor).  If the precision of the rail is truly a
concern, you can put a 1k or 2k trimmer in the Y and use that to make fine

As far as the gain trimmer is concerned, I would use a much smaller trimmer
relative to the fixed resistor, as this will give you a more precise
adjustment.  Hence, instead of a 100k trimmer and a 47k fixed resistor, I
would use a 20k trimmer and a 91k fixed resistor.  Or, better yet, if you
are sticking with 1% resistors (which you really should for this sort of
application, since they have better temperature specs), then you can use a
10k trimmer and a 95.3k resistor, or (better still) a 5k trimmer and a 97.6k
resistor.

My \$0.02 for what it's worth.

```