[sdiy] Silly question about followers
David G. Dixon
dixon at interchange.ubc.ca
Tue Jul 28 18:52:01 CEST 2009
The circuit is Ray Wilson's matrix keyboard circuit:
http://www.musicfromouterspace.com/analogsynth/KEYENCODERPCB/KBDMATRIX022005
.html
The extra follower is U9-B on schematic page 2. I believe that this has
just been done to use up an extra opamp on a quad, but I want to be sure.
As you can see, the opamp in question is LF444, which has very tiny input
bias currents (typically 10pA, compared to 65pA for TL074) and lower supply
current, but which otherwise looks to be a bit of a step down from TL074
(higher noise, lower slew rate, etc). I've bought a few LF444s for this
purpose, although I'm sure that TL074 would work just fine, as this input
bias current doesn't amount to the proverbial hill of beans in the context
of this circuit. I'll try both, and if TL074 works well, then I'll save the
LF444s for sample-and-hold circuits.
I've altered Ray's circuit to make it compatible with a 12 x N matrix
(instead of an 8 x N matrix). To do this, I've added an additional R/2R
ladder for voltage subtraction and completely changed the opamp
configuration. Each ladder has a follower, the main ladder's follower
output is inverted at unity gain, and then the two are summed in an
inverting summer at a ratio of 1/4 with a gain of roughly 2 (trimmed for
1V/octave output). That all requires a quad. I want to take the output of
the summer directly to the 100R resistor for CV3 (no follower), and then use
a dual (TL072) for the two portamento outputs as in Ray's circuit.
> > Here's a silly question, but I just want to make sure I'm correct before
> > eliminating an inconvenient buffer from a circuit I found on the web:
> >
> > Is there ever a reason to place a follower after an inverting summer?
> The
> > output impedance from an inverting summer is just as tiny as that from a
> > follower, so nothing would be gained by re-buffering the output,
> correct?
>
> Do you have a link to the circuit in question?
> For starters, some op-amps (like the CA3140) cannot drive a low
> impedance load (less than 2K), and are usually buffered when interfaced
> to the "outside world".
More information about the Synth-diy
mailing list