[sdiy] Switching regulator and microphone noise

[Harry Bissell Jr]

Looks to me like the problem is the 10K connected
directly to the positive supply. This will couple
supply noise DIRECTLY into the electret

You should make a
nice stiff decoupled or regulated supply... maybe
a 1K resistor and a large value cap (in parallel
with a small value cap).  I would seriously think
about a zener shunt regulator. People think of zeners
as being 'noisy' but with decoupling, they are way
better than what your doing now.

A ferrite bead in that line might also help a lot...

H^) harry

--- Bert Schiettecatte <bert at percussa.com> wrote:

> Hi all,
> 
> I'm trying to build a good electret microphone
> amplifier here. I got
> some cheap electret microphones from Digikey (horn
> industrial brand) and
> some LM833D's here. I built a simple circuit first
> shown here:
> 
>
http://www.geocities.com/ferocious_1999/md/micpreamp2.html
> 
> Instead of letting one opamp handle the gain, I've
> used the two opamps
> in the LM833 with 14K, 47K and 2.2K resistors to
> create 4.4x and 7.4x
> gain stages, so in total 32.6x gain. The signal
> coming out of the LM833
> goes into an ADC (0 to 3.3V range). 
> 
> When I connect the circuit to a big lab power supply
> and run it at 3.3V,
> It seems to work fine. When I connect it to a
> switching regulator, I get
> a lot of noise in the background. The microphone
> signal is still there
> and audible, but the noise is a lot compared to the
> lab power supply.
> The switcher's frequency is between 1.2 and 1.8 MHz
> according to the
> datasheet. 
> 
> There will be a CPU in the design as well so I'd
> like to isolate the
> audio circuitry from all the digital stuff. How do I
> get rid of all the
> noise and interference? Decoupling caps seem to
> filter the noise a bit
> but don't remove all of it. 
> 
> By the way, is there a difference in quality or
> noise between LM833D
> (On-Semi) and LM833M (National) ?
> 
> Thanks a lot for all the help!
> Bert
> 
>