electret capsule interface

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

ambi_doug

Active member
Joined
May 21, 2005
Messages
26
after discussion with my father, i've revised the circuit for my electret capsules thus:
elect-caps-pre1.gif

apologies for crappy photoshop schem.

the only thing with this is that the capsule will be operating with a +15V bias, and i'm not sure if this is going to cause a problem. i figure that it's worth trying though - what with the price of these capsules being almost nothing (80p - £1 ish).
 
I'm almost sure that max volatge for this capsule is 10v...do a search for ecm8000 mods,because there is a great circuit by Marik
 
it's not an ecm8k capsule.... but i can see that they're probably nearly all the same.

is there a simple way i could divide my +V supply? would a simple potential divider work here?
 
a resistor and a zener diode would do the trick...

also take a look at :

http://homerecording.com/bbs/showthread.php?t=111051&highlight=ecm8000

a treasure of information...

BTW many electrets i've seen work between 1,5 and 10v...seems almost universal...
 
does anyone know what'd happen running a capsule on 15V?
i don't have any zeners to hand...
and indeed may have already destroyed this capsule due to overheating. that case connection isn't easy :(
 
There are no FETs I know of, even the teeny ones designed for capsule service like the Toshiba 2SK266, that have less than 15V breakdown. Moreover for a JFET such breakdown, if the current is limited (which in your proposed circuit it is) is not destructive.

Also, in the circuit you are typically running at a considerably lower voltage across the part.

I just read the thread from the other forum and it sounds like the typical capsule FETs are quite robust. It wasn't quite clear though that the two contacts and the case are always distinct different connections, but maybe some one could shed light on this. Clearly this is required for your circuit to be configured as shown.
 
Further thoughts: it might be well if the 100k bias R's were larger and the PNPs matched for beta. With my model of the 2N5087 the bipolars are pulling about 6.8mA, resulting in excessive base current from a noise perspective. That setup is also yanking the outputs down to only about 2V above ground, which is o.k. but sacrificing some potential dynamic range.

Using a 2N4119 as the hypothetical FET, that part has 12V across it, assuming a floating gate potential of 300mV. If there is a high meg R there across the capsule keeping the gate at d.c. ground (probable) the voltage across goes up a bit.
 
well, powering it up as shown wasn't exactly successful - but that's hardly surprising.... 1st attempt and all.

a number of mods have been made, and now it works quite nicely.
(edit: appears i didn't damage the capsule with heat :grin: )

yes, it's a little noisy.

yes, the cheap capsules sound a bit cheap.

but it's now in an operational state, where i can start to prove the other aspects of the system (my ambisonic project).

i'll post a revised schem later, but i feel like doing a little mech work now - working out ho to mount the capsules in the frame i made.
 
Congratulations on making the first step! :guinness:

Back when I was wrassling with noise and the electret-to-phantom circuit I ended up just using a resistor in place of the capsule so I could minimize noise in the circuit. Yup, the Panasonic capsules are noisy for ambient recording, but there's nothing like being able to actually get sound for cheap.
 
i'm not even using panasonic capsules.
got some projects unlimited ones.
link

but yes, i'm really quite happy that i'm getting audio through something i built! :grin: :grin:
 
That's the way to go, absolutely---get some stuff built and get some experience with it. Then analyze and revise.

I'm still puzzled as to the capsule connections though. I looked at some of the projects unlimited datasheets and they don't say anything (so far) about connections if any to the shell of the capsules. Does anyone know if there are three-wire capsules (i.e., ground, FET source, FET drain)?
 

Latest posts

Back
Top