oscillation in Ne5532 simple preamp

[DiggyFresh]

Hi folks! 


i'm toying with a very simple pre-amp circuit with a Ne5532,  the circuit can be found here:

http://schematicdiagrams.net/images/0/electret-microphone-pre-amp.jpg
from this page: http://schematicdiagrams.net/small-electret-microphone-pre-amplifier.html


I use it to drive a line input. i replaced R5(100k) with a 10k in series with a 250k pot for gain control (in negative feedback) (as described in article)


the issue i'm having is when i turn the gain pot a little from minimum, it adds a pitched tone that varies in frequency with the gain control pot.
I have very limited knowledge in opamp design, i guess this is due to oscillation (?)


Any way i could improve stability ??   i dont need frequencies above 10k-15k, maybe i could add a simple filter?

It showed up during testing with a bunch of aligator clips tho, will it be more stable once wired and shielded in the box?
or maybe try with a tl072?


any help appreciated , thanks

John

[R.G.]

It's almost certainly an issue with one of:
- power supply decoupling; there should be a 10uF and a 0.1uF *ceramic* cap as near the +V and -V pins as you can get them, both electrically and physically.
- a wiring issue, where the wires don't go exactly where you think they do
- a soldering or bad contacts issue.

[DiggyFresh]

Thanks R.G,


I cured the problem swapping the ne5532 with a tl072, the "oscillation" went away and the gain pot now functions normally.
i guess maybe because of the higher bandwidth of the ne5532,  my 9volts battery is already decoupled like you said with 100uf and 0.1uf but not ceramic.

what i don't get is i used a circuit initialy using the ne5532....   


Is there a "cure"  to be able to use the ne5532?   i wanted to use it because of the low noise, but may not make a difference in my application...


Something like an rc network somewhere to get rid of the higher frequency or something?


thanks!

[Lurco]

Perhaps an additional cap across R7?

[PRR]

{as Lurco says} R7 *really* needs to be bypassed to ground with several uFd cap.

It's high-gain (more with your mod). If output is anywhere near input, it is sure to howl.

The basic topology is non-optimal. Non-inverting is quieter. For 2-pin electret bandwidths, one '5532 stage can do your gain of 250 (if you really need that much).



(Let me know if that image does not show.... trying something new.)

[DiggyFresh]

Thanks alot guys!  that's the kind of info i needed.  I will try to add a big cap across R7 to see if it helps!


@PRR  thanks! I wish i had found a schematic like yours first, i will try it also.  When you say for "electret bandwidth" , what kind of bandwidth do you mean?
  i found the first circuit online but i'm not using it for an electret microphone. I was looking for a simple, lowest possible noise pre-amp to replace
  a lm386 output stage in a toy voice changer to get a better, more linear line level output. (a black blob chip, wich i don't know the output impedance..).  My first circuit works, and is an improvement but if there's a better circuit for my application i will try it.  The  voice changer chip is a pretty Lo-fi effect and must have a small bandwidth but i would like to have a circuit wich can get the most out of it.


any ideas realy apreciated, i'm brand new to these small opamp preamp circuit.