Hi all, Oh yes it does.....I've seen it and in fact we used to look at
silicon Planar (Fairchild TM, I think) emitter junctions for poor doping and
masking flaws using that in the early 1960s.

You do need a few milliamps of current and It starts off as a red or orange
glow round the emitter periphery when view under about 100x met. microscope.
I think it "whitens up" as you increase the current.  I still have a feeling
that there may be some truth in what Andy says the hot electrons from the
avalanche at a few milliamps could be  forced quite deep into the device and
some could easily be trapped by the collector. The thing is that most of the
breakdown current is carried at the surface just under the SiO2 passivation.
The field does not really aid electrons going deeper to the collector. It
does inject hot electrons into "traps" the oxide and reduce the gain of the
transistor as well. This is a similar effect to UVEPROMs I think.

Motto never let BJTs run into eb avalanche breakdown ....and that includes
RF devices being driven hard. Nat Sokal of Class-E fame has written papers
on that.

Rik's effect ......I have seen almost negative impedance characteristics on
a curve tracer. I can not remember the condition but avalanche is very like
gas discharge in some ways, and there is a well known sawtooth osc you can
make with a neon and and a capacitor. Some devices have very (spacially)
uneven breakdown and areas flicker on and off as they get hot. At the edge
of breakdown there can be very unstable "spots" round a circular emitter
which light up first. "Finger" emitters are more complicated as there is the
radius of curvature effect on the field as well.

All good fun it brings back memories from 40 years ago.....in the Post
Ofiice Research Submerged Repeater Transistor group.

Cheers de Alan G3NYK
[email protected]