Date: Fri, 1 Sep 89 21:31:52 EDT
From: kludge%pyr@gatech.edu (Scott Dorsey)
Subject: Op Amp Microphone Amplifiers

For some time, I've kept a big bag of dual op-amps around, which I have 
been swapping in and out of my Magnavox CD player (an old 14-bitter,
sorry to say).  Overall I have been most impressed with te Motorola
34082.  Great part, though it has rather poor unity-gain operation.
It's slightly noisy too, but it seems to sound better in the Magnavox
than any of the others I have tried.  Comes in an uncompensated version
as well.

Recently, I have been playing with placing amplifiers directly at the
microphones for recording work, and running high-level (unbalanced,
unfortunately, but that will change) signals to the board.  Most of
the amps I have played with use 6CW4 nuvistors, and work well.  Power
supply wiring has been a problem, though I borrowed a lot of multiconductor
24 AWG stuff from work.  

The major problem has been with ribbon mikes, mostly some Shure units 
that I like to use.  The Shures have internal transformers (really not
of great quality, but I have not had the chance to remedy this), and
a 600 ohm balanced output.  They are pains.  They pick up all manner
of noise, and require immense amplification.  The 6CW4 is far too noisy.
In addition, there is a problem in that the common mode rejection of the 
amp has to be pretty high if a run of any length is to be used.  The
hum pickup even on a ten foot run is truly incredible.  So I got on the
phone and got some free samples.

I tool the Jung Audio Op-Amp book, and found a schematic for balanced
input amps.  Changed a couple of values; 60 dB of amplification is about
par for these mikes.  Didn't get much in the way of performance at all.
So, I used the dual op amp, and built a 30 dB differential amp, with
a following stage of 30 dB.  Worked very well, but the rejection of hum
could have used some work.  A call to Vishay brought some .001% resistors
which found their way into the input stage and out of the EEG laboratory.
The overall noise rejection was incredible; given a theoretically perfect
op amp, the input rejection was 120 dB.  About $5 per resistor though, and
I don't think they'll sample me any more for a long time.

The internal noise was still much better than the single-stage version,
but overall it stank.  Far too much for a widely-miked choral work,
which is what I usually like the beasts for.  So I dug out the op-amp
bg and tried the 5532, 5534, AD712, MC34082, MC34083, a TL part, and
a few other Motorola op-amps which were designed for low-noise operation.
(the 34082 is not an "audio" chip.  I used it for a 3.5 MHz radio 
transmitter modulator stage, and for video amplification too, but it's
not designed for these things.  If you can't tell, I like this thing).
Anyway, the upshot of it all was that the AD712 had the lowest noise
of all the ones in the box, with the exception of a LINCMOS part from TI
that had very poor high-frequency sounds).  It was crisp and it was
clean.  I't time to get some more samples of this one.
--scott