7 Comments
User's avatar
Zeljko's avatar

I built a lot of them in 80's. Loved this circuit. RIZ BC107, metal can, worked the best.

What about zinc oxide negative resistance oscillators? Hidden gem for after nuclear war.

And forgotten UJT transistors.

lcamtuf's avatar

Hah, had no idea about zinc: http://www.sparkbangbuzz.com/els/zincosc-el.htm

Yeah, UJT, neon lamps, and tunnel diodes all go onto the pile of stuff that almost no one has at hand. It cracks me up that tunnel diodes are still often talked about in textbooks when the only affordable supply is basically stolen Cold War surplus from former USSR republics.

Zeljko's avatar

Tunnel diode should work the same, just with the res and cap. Never had anyone at my disposal.

lcamtuf's avatar

Tunnel diode has a different curve, so IIRC, the minimal circuit is usually this:

http://www.seekic.com/uploadfile/ic-circuit/s2009719213449657.gif

But yeah, it's more or less a relic of a bygone era.

Zeljko's avatar

Ah tempus fugit...

2N2646 ujt...

I was filing the top of the BC107, 2N3055, to get a miraculous 0.5V solar cell...

Wyrd Smythe's avatar

I saw a YouTube video about this. Isn’t it also potentially destructive for the transistor?

lcamtuf's avatar

In itself, reverse breakdown isn't. But because it can cause uncontrolled current at a high voltage, it's pretty easy to exceed the power rating and fry it. In this case, because you have short pulses (< 10 ms) with low duty cycle (< 5%), it's probably fine both for the LED and the transistor, but it's obviously not a good design.