Created
01/15/08

I remember seeing somewhere on the Internet (EDIT - I saw it here) that it is possible to produce light from a power transistor junction. As I am all for abusing electronics, I set out to replicate this behavior. I chose for my victim a 2N3055 power NPN in TO-3 case. To get at the junction inside I cut open the case with a dremel and cutting wheel:

I connected it to my power supply through an ammeter so I could measure current accurately. In this particular egregious abuse of technology, the base-emitter junction is reverse biased and the collector is left unconnected.

Sure enough, after cutting the lights in the room and fiddling with the dial on my power supply, a very faint light was eventually observed at 11.53V and 20.66 mA. The light was so dim I could not make out the color. After steadily increasing the power, I detected the characteristic stench of hot electronics at 12.39V and 200.4mA. At 13.4V, 440mA the light was still dim but quite obvious.

My power supply maxed out at 13.8V,526.0mA. Pictures were taken after this had drifted to 15.26V,527.1mA, for a power of 8.044 Watts! I could now tell the color of the emitted light; a pale greenish-yellow.

The source of the hot electronics smell was identified to be my meter leads, the sheaths of which melted due to proximity to the hot transistor case. Curiously, no damage to the transistor was apparent once it cooled!

This was kind of neat, but somehow I just don't think it will bring about a revolution in lighting technology.

I also took some pictures of my notebook, but my camera at the time wasn't so great so they aren't terribly high-quality pictures.