On 2012-05-23, Dave wrote:

I don't *think* the noise is from high frequency feedback, as it is a
low-pitch buzzing, which makes me think of a capacitor being overloaded.
And (REALLY hate to ask this) How would I *decouple* the supply rail? Power
is taken from a wall-wart and run through an LM317T before feeding into the
circuit. Sorry, I really am making this up as I go along. Don't mean to be
painfully ignorant, just am not anything like an EE. The input impedance,
as I calulate it, matches prety well with the 1.5KOhms of the microphone,
and I *thought* the output impedance was similarly matched to the 8-Ohm
speaker.

I thought about the portable radio path, but decided against it for the sake
of simplicity. And it worked fine untill I added Q4.

you seem to have a simulator there (looks a bit like lt-spice)
replace V2 with a more realistic model of your power supply
(internal resitance, ripple)

Your circuit appears to have heaps of gain, especially at high
frequencies, (HF like CB radio - add a few microvolts of that
to your simulated input)

Put some resistors in series with emitter capacitors to make
the AC gain more predictable... do you really need ten million times
amplification?

If you just want to build an intercom and don't want to learn design
by doing try this one, it looks sane to me.

http://www.zen22142.zen.co.uk/Circui...io/doorint.htm

--
š‚šƒ 100% natural