Subject: Re: Thermal runaway: please settle this dispute!
Organization: Nobody nowhere
X-Newsreader: Microsoft Outlook Express 6.00.2720.3000
Date: Thu, 2 Jan 2003 14:47:43 -0000
NNTP-Posting-Date: Thu, 02 Jan 2003 14:47:53 GMT
"Paul Burridge" wrote in message
> Hi all,
> I ended up having a bit of a disagreement with a friend of mine over
> the holiday concerning thermal runaway in a BJT.
> I maintain that an increase in temperature across the emitter diode
> leads to a fall in that junction's barrier height, making the
> so-called 'energy hill' easier to overcome for the forward biased
> current flowing through it. The causes increased current flow, giving
> rise to further heating of the emitter diode, thereby lowering the
> barrier potential still further and so on until the device is
> destroyed. The fall in barrier height is around 2mV per degree C., if
> I recall correctly.
> My friend, OTOH, maintains that it's nothing to do with the
> emitter/base junction at all, but solely arises due to an increase in
> temperature at the collector, causing the same cycle of
> heating/increased current until the device is destroyed.
> So what is the panel's view on the subject? I always believed I was
> right, but this chap is regarded as pretty eminent in electronics so I
> can't simply dismiss his views out of hand.
I think your friend is right. The way I understood thermal runaway,
especially with Germanium transistors. Is that the hotter the transistor
gets the more current it conducts, and so gets even hotter. The cycle
continues until the transistor destroys itself. Which is why in some
amplifiers they employ a negative temp 'NTC' thermistor in the output stages
to reduce the current when the heatsink gets hot, to try and stop the
amplifier destroying itself with thermal runaway.
I did experience this myself once with my dads music centre (which had
Germanium output transistors). I was playing it on 10 on the volume. And it
suddenly decided that it had enough and let the 'magic smoke' out.
Fortunately for me it only burned the Collector resistors, I caught it in
time, and hit the off switch straight away. :-)