Decoupling circuit, capacitors?

[sanpei]

Hello everyone!
This is my first post here and I would glad if someone can help!

I have a circuit where a component introduces spikes on the regulated +5V rail, making some amplificated electet microphones become crazy.

I think I should add a decoupling circuit in front of the Op-Amps of the microphones, but how to do it? Is only a capacitor ok, and what capacitance?

Thanks for the answers!

 

[Nigel Goodwin]

You need to post your FULL circuit, and layout pictures as well - we can't make suggestions without knowing exactly what you've done wrong.

 

[sanpei]

Thanks for answering. Here it is, sorry for the drawing but I hope you can understand it.

 

[MikeMl]

Are there any decoupling capacitors at all????

 

[sanpei]

No, there are not. I was asking if it's necessary to add them and how (capacitance, where, etc).

 

[ramancini8]

Go to TI's web and look for Practical Analog Design Seminar under the Technical Documents section. Chapter 3, signal integrity, contains a discussion on generated noise and decoupling that you should read. If you want an answer without knowledge put two .1UF ceramic caps on each digital IC, and filter the power lines to the analog circuits with RC filters comprised of 51 ohms from Vcc in series with .1UF ceramic cap to ground, and OP amp power taken from the junction of the RC..

 

[audioguru]

I think I should add a decoupling circuit in front of the Op-Amps of the microphones, but how to do it? Is only a capacitor ok, and what capacitance?

You showed a simple block diagram instead of a detailed schematic so we are just guessing that there is no filter between the noisy 5V and the resistor that powers each electret mic. You also do not show the details of how the opamps are biased directly from the noisy +5V.