Cursed circuits #6: reverse avalanche oscillator
Last year, I published an article titled “It’s hard to build an oscillator”:
The title alluded to the fact that there’s no shortage of oscillator circuits on the internet, but many of them use unusual parts, need weird supply voltages, or don’t work well (if at all).
But sometimes, bad rises to an art form. Here’s probably the most puzzling bad oscillator you can assemble today with the parts you have at hand:
Reverse avalanche oscillator. Other small NPN transistors should also work.
At first blush, nothing here makes sense. The transistor is upside down and its base terminal is not connected. And yet, the circuit works: hook it to a supply of about 14-20 V and watch the LED blink.
If you connect an oscilloscope to the terminals of the capacitor, you’ll see that the cap is repeatedly charging to about 10 V, then rapidly dumping some of the charge, all the way down to 9.1 V:
Capacitor charge state with a 14 V supply (5.8 Hz oscillation). By author.
It’s no mystery that the capacitor must be charging via the 1 kΩ resistor from the positive supply rail. It’s also clear that the energy is dumped into the LED via the upside-down NPN transistor. But why?
To answer this, you’ll need a basic grasp of semiconductor junctions. If you’re rusty on how they work, this earlier article should jog your memory:
As a quick recap, a conventional diode consists of a p-n junction formed from two distinct types of semiconducting materials. At the boundary between these materials forms a non-conductive region known as the depletion layer. In forward bias — when a small positive voltage is applied to the p -side in relation to the n -side — the depletion layer is disrupted and charge carriers can cross.
In reverse bias, in contrast, the depletion region remains notionally impassable. That said, if the applied reverse voltage is high enough, the electrostatic field grows so strong that it can accelerate charges with sufficient force to knock electrons in the depletion region int…