I am using a MAX680ACP to generate +/- 10V supplies for
an LF383 op-amp, the output of which I am feeding to
an LM567 tone decoder.  It works fairly well, but the
ripple on the +10V supply (at about 12Khz) gets amplified
and is impressed on the output whether the photo transistor
is covered or not.  The strange thing is that there is no
measurable ripple on the -10V supply.  Things go down hill
fast when I try to boost my gain by using the second op-amp
as a 1:100 gain amp.  Needless to say there is a nice square
on the output with, surprise!, the same period as the
charge pump ripple.  I have bypassed the +10V supply to
the op-amp with a ferrite bead and 4.7uf cap.  It helps
a little, but the output is still a wild waveform of
the amplified supply ripple.  The op-amp circuit is nothing
special. I have both stages set up with 1:100 gain using
470K for feedback and 4k7 on inputs.  The photo diode is biased
with a 100K resistor to V+ with the other end of the diode
tied to V-.  The junction of the diode and the 100K is coupled
to the non-inverting input of the first stage with a .01uf
cap.

I would probably just give up and use an "IR receiver in a
can" but I feel compelled to know whay mistake I have
made here.  Even though there is no way I am going to pile
all of those parts on a board for each sensor, I still want
to learn form this...

Any tips will be greatly appreciated!  This is going to form
the basis of an IR bumper for my PIC'Bot.


Dan