The microphone input is a _VERY_ sensitive circuit as audio inputs
go.  The optimum level is probably around 5 millivolts while the stereo
headphone output produces around 1-hundred times that amount without really
even trying that hard.  The name of the game is attenuation.  The pot is the
right idea but there are about 20 decibells or more too much audio.
What is needed is about 1 1-hundredth of that signal level at the maximum.

        Try the following circuit for starters:

        Find a fixed resistor that is about 100 times the resistance of the
element of the record level pot.  The audio from the stereo should connect
to one end of the resistor while the other end goes to the high end of the
pot.  The low end of the pot and the ground lead of the headphone output
need to both connect directly to the signal ground of the voice recorder
circuit.  The wiper of the pot should be capacitively coupled to whatever
circuitry the data sheet for the voice storage chip requires.  This gives
a variable record level that goes all the way from no signal at all to a
maximum of about 1/100 of the audio from the headphone output.  If you
omit the series resistor, there will be so much audio present that you will
probably have too much level almost as soon as the level pot is moved away
from the low stop.

        You may test this part of the circuit if you have a cassette tape
recorder that has an external microphone input.  You don't even need to power
your voice chip for this test.  Just connect the microphone input of the
cassette recorder between the wiper of the pot and ground and run some audio
to see if you can get a good but not too high level on the tape recorder.
Turning the pot should be just like turning any volume control so you can
check your input circuit before trying or frying anything else.

        A 10 K fixed resistor in series with a 100 ohm volume pot would give
you about the right amount of attenuation plus a low impedance output to the
voice recorder chip's input.  Your stereo would see essentially a 10 K load.


Martin McCormick WB5AGZ  Stillwater, OK
OSU Center for Computing and Information Services Data Communications Group