Fixing Noisy Power Supply Issues

[Evil Bunny]

"noisy" power supply?

So, there was a high pitched sound coming from a component (a radio, not sure if it was a transmitter or a receiver)...

An experienced technician said that it sounded like a bad power supply... He mentioned something about a bad capacitor. The power supply was changed out and the noise went away. He was right.

This explanation was puzzling to me... Any ideas on what would cause this high pitched noise?

 

[yungman]

It is very common for an unstable power supply to oscillate. I think what you have is the power supply oscillating. People call this noisy, even some switching power supply manufacturers require a minimum load to keep it stable or else it gets "noisy".

 

[Evil Bunny]

Can you explain further? Oscillating? What in a power supply would oscillate and create sound waves?

 

[yungman]

Power supply always has closed loop feedback to control the constant voltage. When you have closed loop, there is always possibility of instability. I see too many of this situation. You'll be surprised the application circuits provided by the IC manufacturers that are unstable. It is more common than you think...than you imagine. I know because I had fixed so so many problems like this. You have to treat it as a closed loop feedback control system.

You should ask people here like Jim Hardy that is very good in this. I just know how to fix the problem and not get into the Laplace transform. I only use Bode plot and it worked so far for me.

 

[Evil Bunny]

Thank you Yungman, but I still don't understand what caused the sound waves...

 

[SteamKing]

Was the radio on when you heard the noise?

 

[Evil Bunny]

Yes it was on.

 

[yungman]

Thank you Yungman, but I still don't understand what caused the sound waves...

Closed loop instability manifested as an audible frequency. You can pull your hair with where the sound, what is the source and why. My experience is all about the closed loop instability of the control system. When the system start oscillating, it is all over the place. If you have the schematic, it is very easy to pin point where is the pole zero compensation components. My way is to draw the bode plot and find a way to have one pole cross over.

Yours shouldn't be even hard to find because it is "singing" consistently. The hard part is when it only conditionally singing and you have to find a way to trigger the oscillation. Look at the controller chip and there should be a small cap mainly for pole zero compensation. that is the place I would start. Again Jim Hardy should be the person in control system and Laplace transform.

 

[fleem]

If the sound was coming from the speaker, then that might mean the voltage from the power supply was not well filtered (and why the tech said it might be a bad electrolytic filter capacitor on the output of the power supply). If the sound was coming from the supply itself, then it might be a coil or transformer with epoxy that got loose, allowing the windings or core parts to vibrate from the AC in the transformer or coil. In either case, the reason its a high frequency and not, for example, line frequency (i.e. 50 or 60Hz), is that a switching supply uses high-frequency so that the coils, transformers and capacitor can be cheaper and smaller.