Comments on electret mic amp

[MacGyverS2000]

Been rummaging through my junk drawer as of late and decided to play with electret mics for an upcoming project. I'd appreciate any "10,000-foot level" feedback you guys might have on the circuit, be it the actual connections or just the values. This is by no means a high-quality/low-noise piece, just enough to get a reasonable reproduction of the sound sent to a filter bank.

Thanks!

FYI: The mic I'm using has an impedance of 2.2kohm with a standard operating voltage of 2.5V (10V max). I believe I set up the voltage divider correctly (too many sleepless nights of designing may have let a bad number slip by). The 100k is in there for negative feedback, but the emitter resistor should do some of the same (think I should kill off the 100k?). The 1uF caps should be large enough to let the bass through (but again, I could have been sleeping at the wheel).

+12V o---|----------------------|
         |                      |
         |                     1k
         |                      |
         |          |-----------|----1uF------o
        12k         |           |                Output
         |         100k     ____|             o
         |          |     |/                  |
     o---|----1uF---|-----| 2N2222          -----
In                        |\____             ---
     o                          |             -
     |                        82ohm
   -----                        |
    ---                       -----
     -                         ---
                                -

 

[MacGyverS2000]

Oh, I should also mention the output is going to an AGC circuit, so that should pump up the gain enough to have it swinging between the rails.

 

[logbook]

Don't delete the 100K resistor or you won't have any bias! This is a very small signal circuit, so the bias point, although not ideal for a Class A amplifier, will probably be ok.

1µF at the input is ok because the input impedance of the transistor stage is of the order of kilohms. At the output, however, the impedance may be lower (tens of ohms?) and the coupling time constant would then be too small.

 

[MacGyverS2000]

I hate to make the output cap much larger, but I suppose I'll need to... something around 100uF should do the trick, don't you think? <edit> I had originally typed that I didn't think I could get what I needed for a fair price and still keep it surface mount. If I switch to an electrolytic for that one, Panasonic has some 25V, 100uF for about $0.18/pc in 1k units, and they're still SMD (roughly 7mm square and 8mm tall... not tiny, but I'll have no problem fitting it in).

I haven't analyzed them, but I've come across several circuits with similar characteristics... I wonder how they operate in the real world:

http://www.brainz90.karoo.net/mic-amp.htm

This circuit looks nice, not overly complicated, but there are inconsistencies (for example, the supply is shown as 6-30V, but most electret mics won't take more than about 10V):

http://www.zen22142.zen.co.uk/Circuits/Audio/ecmmic.html

I'm still liking my original circuit for low-cost/complexity reasons, though. You just can't beat a sub-nickel NPN, a few sub-penny resistors, and a couple of sub-nickel caps for price... that 100uF cap practically doubles the price of the whole circuit!

If you think any other values are out of range, though, please let me know. My next step is to put together some Sallen-Key filters (LP/HP/BP) after the AGC... I hate messing with the values since I can't separate Q and gain (especially on the BP).

 

[MacGyverS2000]

Hmmm, what about bumping up the 100k to 200k, and the 1k in the collector to 2k? It's not the holy grail, but it should give me some more headroom with the output cap value. Of course, all of my previous values were based upon a Beta of 100... most of the 2N2222's I've used are closer to 200+.

 

[microhenry]

the 1uF output caps shouldn't be a problem if the input Z
of what you're driving is high enough, say above 4kohm