Weird Power Transister Sound.

[centar]

During testing of a power amplifier design and running the output into a resistive load at high power in a quiet room I noticed I could hear the music source.

Upon inspection to see where this sound was coming from it was found to be emanating from a pair of Mj2955 transistors bolted directly to an heavy aluminum heatsink (no insulation, just heatsink grease)

the sound seemed to be comming from within the transistors as you could muffle it a bit by placing a finger on the transistor's round metal cover.

Has anyone else experienced this? Is this common?

I have been an electronic technician for 25 yrs and never seen (heard) of something like this...

Can't even begin to understand the mechanism behind the phenomena, It is solid state right? Just a piece of silicon...

Can anyone shed some light on this?

 

[Andrew Eckhardt]

I always assumed it was a thermal effect, but clearly one way or another, the die changes shape with the signal. Clipping power stages sound particularly nasty.

 

[Tekko]

Likely piezoelectric effects in the die or air pockets around the bond wires, allowing them to vibrate ever so slightly.

 

[Andrew Eckhardt]

Where's a semiconductor physicist when you need one?

 

[Conrad Hoffman]

I've heard this, especially with high voltage amplifiers (1kV). Frankly, none of the proposed explanations are very comforting. Die size changing? That doesn't sound like a good thing. Bond wires vibrating? Seems like a recipe for failure. Piezoelectric effects? Hopefully not in silicon, though possible where capacitors are included in ICs. I would have said simple electrostatic forces between the case and sink, but in this case there's no insulator. Anyway, I'm clueless- some would say that's not new news.

 

[centar]

I had thought piezoelectric but that would mean that transistor would be microphonic and I have never seen that...
Die wires vibrating? I think that would fail within mins...
Thermal is interesting but I find it hard to believe that it could respond fast enough to create sound...
It defiantly has nothing to do with the mounting/heatsink as I played with that and it had no effect...
I'm fairly sure I have heard it in the past and never really bothered with it, but this time because it was a "design" I was most interested as it "sounds" like a possible failure mechanism for the output devices...I mean something is moving, right?
It appears that "SOLID STATE" isn't so solid after all... ;-)

 

[jez]

This is quite a common phenomenon. I have heard this many, many times when doing sine wave tests on amplifiers, valved as well as solid state....
Amongst other possible causes is one called magnetostriction...
Also just simple electromagnetic effects between conductors in the same way as a loudspeaker works.