Coil whine and signs of a degrading power supply

[axero]

After posting the following thread:

https://www.eevblog.com/forum/beginners/converting-~320v-dc-down-to-~12v-dc-%28not-much-power-required%29/

I started having a new perspective on coil whine on switched power supplies.

When considering a pure transformer with a primary and a secondary coil, there are types that transform mains AC down to 12V AC and being used to power e.g. low voltage halogen bulbs. They tend to hum at say 50 Hz or 60 Hz, which is the frequency of the alternating current. If the transformer is behind a dimmer, the transformer tends to have a rattling noise that comes from the clipping of the dimmer as it switches the mains voltage on and off rapidly. So the coils on the transformer are moving with the electricity as a result from the magnetic fields involved.

If we go back to a switching power supply, the switching occurs at say 20000Hz or somewhere in that region depending on manufacturer and their design. Then I conclude that the whining noise comes from the transformer that is running under influence of the switching circuitry.

This raises a few questions as to what measures are taken to eliminate coil whine. Perhaps they let the switches operate at rates that are way above human audible spectrum, perhaps at say 60 kHz. But, wouldn't it still give rise to human audible subfrequencies? Could the other way be that they instead let a high voltage capacitor take the hit from the incoming pulse instead of the coil in a flywheel circuit?

Then I came upon a batch of small electronic devices that all share the same specifications where one of them had a high pitched whining noise that the other devices did not have. After a few days of operation, the power supply failed completely.

So what went wrong here and gave rise to that whine that was absent in the other units?

I also have an LED power supply that used to be silent, but after a few years of operation it has started to whine a lot, what has happened here? It is still functioning normally otherwise.

With all that taken into consideration; If I buy an expensive ATX PSU for a PC that is strong and has a Platinum energy rating. If it is suffering from coil whine, why should I accept it and not expect it to fail eventually due to a failing circuitry?

[dmills]

All power supplies can be expected to fail eventually, just like everything else (PC power supplies being mostly consumer junk (Yes, even the "expensive" ones) somewhat more them most).

I am not sure that noise from the inductors (It may actually be from the ceramic caps) is really diagnostic of anything except that the switching frequency is annoyingly low or that something has not been varnish impregnated.

Further you sometimes get inductor noise due to comparatively low frequency load changes, nothing to do with the power supply at all. I have seen this with LED multiplexing at ~4kHz (Really bad choice of frequency!) where the inductor on a 1.2MHz switcher was audible at the led refresh rate.

On a modern switcher I would expect most of the noise to be load changes not sub harmonics of the basic switching frequency.

Regards, Dan.

[Cliff Matthews]

Never heard it on transformers, but I've silenced whine on caps and loose ferrite beads on transistors with silicone. Hearing it makes it seem like things are going to croak..