In the course of doing the Simple Music Maker I experimented with a few other odds and ends, both as real prototypes on a

breadboard and messing around with the simulator. Here's one of them. I'll blog about a couple of the others, too. These

aren't complete projects, just a bit of tinkering and trying things out, but they may be of interest for all that.


When I was in my teens and first developed an interest in electronics, one of the projects I built back then was a simple

electronic organ. The oscillator was based on an unusual kind of transistor called a unijunction transistor. It was unusual

because it had a single junction, with an emitter, two base connections and no collector. Nowadays, you can't seem to buy

new unijunction transistors, and old ones sell for a lot of money, but you can buy a replacement called a 'programmable

unijunction transistor' so I'm gong to try using one of those. This modern part isn't really a single-junction device at

all, instead it's a 4-layer structure like an SCR with two external resistors needed to define the trigger point (hence the

'programmable' in the name), but it should behave in a similar way and allow me to design an oscillator that will run at

audio frequencies.



The Device


The device I'm going to use is the 2N6027 from On Semi. This is readily available and not too expensive. Here is how it is

arranged to mimic the kind of unijunction device I used back then.



The characteristic of the device that allows for oscillation is a region of negative resistance. This is how it is shown in

an old book for a real unijunction part [1]




Once the trigger point is reached (by the emitter), the voltage decreases as the current increases. This characteristic is

very similar to that of a neon bulb (which can also be the basis of a simple oscillator if combined with a resistor and

capacitor), though the physics is obviously somewhat different.


If we connect the device to an RC circuit, the capacitor will charge to the trigger voltage, discharge until there's no

longer enough current to keep the device active, and repeat the cycle. [In practice, it's slightly more complicated than

that, because the dynamic load-line doesn't meekly follow the curve down, rather the capacitor holds the voltage up and the

dynamic load-line swings out and back again.] 



Practical Experiment


So let's try that on a breadboard.


Here's the circuit and the constructed circuit:




This is the voltage across the capacitor (yellow trace). The frequency is approximately 360Hz:



Here's the voltage across the 47R (blue trace) when it's discharging the capacitor:




Simple Music Maker?


So the next question was: how can I turn this into the simplest possible instrument? I wondered about adding a follower, to

buffer the capacitor voltage, but was there anywhere in the circuit that might drive a 40 Ohm loudspeaker? The timing

circuit was out, but the resistance that I'd used from the cathode to ground was 47R, so why not try there? It's a stupid

idea really - the discharge only lasts for a microsecond - so I had to try it and it actually works; dumping the energy

from the capacitor into the loudspeaker coil at an audio repetition rate sounds a note. It's quiet, but it did mean I could

make a very, very simple electronic organ with just one active device (hurrah!).


I did, briefly, have a go at doing that, with a selection of resistors to give different notes. But I gave up after I got

to five notes - in something that might, possibly, have approximated a pentatonic scale - because it was such a hassle

determining the values. Next up would have been a buffer and an amplifier for the speaker, but I'd gone back to the Music

Box at that stage and didn't get any further, so it remains a curiosity to blog about and not a project. Which suits me



And, since I called this Uni-Tunes, I had better end by saying: That's All Folks!


[1] Handbook of Linear Integrated Electronics for Research. T.D.S. Hamilton. McGraw-Hill 1977.