Silicon LEDs ?

[Phil Allison]

** See ABSE for pics of a silicon NPN transistor chip emitting light.

No trickery involved.

The type is a 2N3439 ( high voltage NPN in TO5 pack)




...... Phil

 

[Dave Platt]

** See ABSE for pics of a silicon NPN transistor chip emitting light.

No trickery involved.

The type is a 2N3439 ( high voltage NPN in TO5 pack)

I believe that's quite plausible.

Some years ago, Bob Pease noted a somewhat paradoxical phenomenon
involving silicon transistors. If you forward-bias one of the two
junctions (either BE or BC, I forget which) and generate a few mA of
current flow, and then read the voltage across the other junction,
you'll measure a nonzero voltage. Where does this come from?

According to Pease, it occurs because the forward-biased junction is
acting as an LED, and the junction you're measuring is picking up the
photons and acting as a photodiode.

I believe that almost all of the light emitted is in the infrared, as
a consequence of the low bandgap voltage of silicon. Possibly this is
changed by the doping characteristic of the silicon in question?

--
Dave Platt <dplatt@radagast.org> AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

 

[Lostgallifreyan]

dplatt@radagast.org (Dave Platt) wrote in news:ngv1s5-k2f.ln1@radagast.org:

According to Pease, it occurs because the forward-biased junction is
acting as an LED, and the junction you're measuring is picking up the
photons and acting as a photodiode.

I remember reading a long time back that all transistors could be
phototransistors to some extent, hence light sealing capsules for them being
used unless you wanted that to happen. If it is also true that like many
transducers they work both eays, then few transistors would be incapable of
making some light at some wavelength. I don't know if it is a generally
reversible thing though.

One thing I did find was that an phototransistor and a couple of button cells
and a green LED made a great IR sniffer (with part of a dark IR LED's plastic
as filter). Given the high gain that allowed me to avoid any other
components, even unoptimised semiconductors might show light output for
electrical input if you go looking for it. I've wondered if solar panels glow
slightly if you put power into them in reverse, but haven't felt like trying
it out on mine...