[sdiy] Unusual transistor failures

[Steve Lenham]

In response to previous suggestions, I got round to chopping the top off 
one of the faulty transistors. The transistor die is intact but the base 
and emitter bond wires (actually thick strips - these are high current 
devices) have both completely detached themselves from the die and 
"curled up" away from it. They also have that golden orange colour that 
one associates with overheating (though to be fair I don't know what 
colour they were originally).

Certainly looks like a thermal problem. New transistors have arrived, 
and I'll double-check the foldback overcurrent protection as well as 
checking for shorts elsewhere in the unit.

Thanks for the suggestions.

Cheers,

Steve L.

Benden Sound Technology
www.bendensound.co.uk

On 11/09/2012 10:33, Steve Lenham wrote:
> Hi,
>
> I'm restoring a unit with a large linear +5V supply using a pretty
> standard uA723 circuit. There are two parallelled pass transistors rated
> at 60V, 25A and 200W each.
>
> The +5V rail was dead (no volts) but there was plenty of base drive to
> the pass transistors. Removing them for a test revealed that they were
> both open-circuit in all directions.
>
> The unregulated side is all OK - correct volts, no blown fuses, etc.
> Installing a temporary pass transistor restores a perfect +5V output,
> indicating not much wrong with the 723 and drive (though I will check
> more carefully). The unit has not been previously tinkered with -
> everything looks original.
>
> I can obviously repair this, but in my experience that is a most unusual
> failure mode for bipolar transistors - I would expect them to go short
> B-E or C-E under fault conditions, blowing at least the fuse and
> possibly other stuff too. I feel a bit uncomfortable not knowing what
> went on.
>
> Has anyone else ever encountered a similar failure mode, or can anyone
> suggest an explanation for the why the transistors died this way with
> everything else seemingly untouched?