Quick and dirty Geiger counter       

Many years ago, when I was a young guy, and the Chernobyl nuclear power plant blew up, I wanted to have a means to detect any significant radioactive fallout. I live in the southern hemisphere, so I could count on at least a few days before the dirty cloud might reach me. I knew a military surplus store, and I had seen some small old Geiger tubes there. So I went and bought a JAN5890, the cheapest they had. After all I was a poor student in those years!

I cobbled together a very quick and dirty high voltage supply, using a voltage doubler off the 220V grid, added a small audio amplifier, and that enabled me to hear the ticks and pops and crashes. I counted them with the help of a stopwatch. For the next days and weeks, I regularly measured the background radiation that way - and indeed, roughly one month after the Chernobyl disaster, I was able to detect a slight rise in the background level, which then slowly tapered off over several months.
After that I put my little Geiger tube in storage. A long storage it would be.

Many years later, prompted by certain ugly events that unfortunately are happening in this world, and in the same area of it, although for different reasons, I again saw reason to dig my Geiger tube out of storage, and again cobble together a very basic Geiger counter. Only that this time it was just one step more refined: A regulated DC-DC converter, instead of a plain old voltage doubler!

The circuit is a little self-oscillating flyback converter, using an airgapped ferrite core and a small transistor,  with a voltage regulating circuit wrapped around. It runs off a nominally 12V supply, but it accepts a wide range of input voltages, and it consumes just a few mA. The lack of any loop compensation results in the converter operating with pulse trains separated by pauses, which helps conserve power.

Plain common 1N4007 diodes are used as fast rectifiers. They aren't fast, of course, but flyback converters have the advantage that diode recovery speed is quite uncritical, so these slow diodes work just fine. Two are used in series, because a single one would be running pretty close to its breakdown voltage.

The transformer was entirely wound with #39 wire. Of course it's inefficient to use the same wire size for the low voltage, high voltage and feedback windings, but the output power required is so tiny that this doesn't matter.

Note that the phasing of the windings must be as shown, or the oscillator won't oscillate. Left is left, right is right, no phasing dots are needed...  If you want phasing dots, draw them yourself, all on the same side!

Instead of building a little audio amplifier, a pulse counter, etc, I simply connected the output of the Geiger tube to one channel of the line level input of my PC's sound system. That allows me to listen to the pops in the PC speakers, and record the pulses over any time desired, using any sound recording software. Later the recording can be watched with your favorite sound editing software. With the old but good software I use, it looks like this:

The audio software allows zooming in, to the point of watching the waveform of each pulse, telling double and triple pulses from simple ones, and of course it's easy to manually count the pulses per unit of time. One can also nicely see how the background radiation sometimes seems to come in waves.

The JAN5890 is a very small Geiger tube, so its sensitivity is very low. With the background radiation at my place, it gives only a few counts per minute. Still, by integrating long enough, one can easily detect the radiation of a moderate amount of any potassium salt, including potassium nitrate fertilizer, and sodium-free salt substitute from the supermarket!  And a tiny sample of uranium nitrate, out of an old black-and-white photography lab, drives the tube crazy.

Maybe someday I will build this Geiger counter in a nice way, portable, battery-operated, with a PIC added to count the pulses and present them on a display as total dose, pulses per minute, per second, etc. But being old, tired and lazy, so far I only assembled the circuit on a small protoboard. The high voltage circuitry is insulated by fresh air, still the best and cheapest insulation material we have!

Note that the big yellow high quality polypropylene capacitor is total overkill! A plain cheap ceramic capacitor would suffice, even one of a much smaller value than the 220nF this one has, and that's what I would use in a definitive built of my Geiger counter. But for the test I used what I had lying around on the desk.
If you need to buy a Geiger tube, I suggest that you get a bigger one than mine, to get higher sensitivity. Usually we want them for measuring background radiation, and perhaps to go treasure seeking for any radioactive minerals, or to detect radon in our basements. All these are pretty weak radiation levels, making a large size tube most appropriate. Also you need to determine whether you are happy with just detecting gamma radiation, which is typical of glass Geiger tubes, or if you want beta and perhaps even alpha too. My ceramic 5890 tube detects mostly gamma, but also beta when it's strong enough. Definitely no alpha, which needs a tube with a very thin window. Such tubes are fragile. In most cases just detecting gamma is good enough.

My little high voltage supply can easily be reconfigured for higher or lower voltages, as needed by the tube you get. It's just a simple matter of recalculating the voltage divider of the regulating circuit.
The core type number given in the schematic is an Amidon part number. You can as well use any other double-E ferrite core of comparable dimensions. The exact ferrite type is uncritical, given the use of an air gap.  Surplus electronic equipment these days is full of such small ferrite transformers. Slowly heating them up in an oven, to maybe 120C or so, will soften the glue holding them together, allowing easy disassembly (with thermally insulating gloves!), after which you can unwind the original windings, and make the custom winding required for this circuit. And then re-assemble the core with adhesive tape of suitable thickness between the halves, to create the air gap. To create the 0.1mm total air gap I used, you need to separate the core halves by 0.05mm, of course. Scotch "Magic" tape is what I used. It has that thickness.

Back to homo ludens electronicus.