Geiger tube power supply hack using disposable camera flash unit
Just in… We just received a small supply of these disposable camera flash boards, now available while the stock lasts….
I bought some surplus Soviet geiger tubes off ebay for about $20 each, and wanted something to do with them. My aim is to build an Arduino shield that includes a GM tube, but for right now I just want to get it working.
The tubes are about 1 cm in diameter, by maybe 15 cm long, and come packed in very old tissue paper in very old boxes.
There are lots of schematics for DIY power supplies on line, and many are really snazzy. But I’m a hacker at heart, and had a couple old salvaged disposable camera flash units I’d saved because I had dangerous ideas regarding the capacitors. Well, I still have dangerous ideas but in the meantime Kirk and I hacked a couple of the boards up to serve as sub-$3 high voltage power supplies.
Here’s a schematic for the boards pre-mod.
Here are two shots of the top and bottom with annotations so you can follow what I’m describing.
For a step by step guide to mod the disposable camera flash click here
First off, you’re going to be working with high voltage. That big electrolytic cap has a *lot* of energy stored in it. The two long leads discharge the flash, and have the full flash voltage across them. Test the board with an AA battery and make sure you can hear it charging when you push the charge button. Carefully short the discharge electrodes and you’ll get a nice flash. Yay. Now cut them off because they’re just dangerous. Short the big electrolytic cap’s leads to each other with a screwdriver just to make sure it’s discharged and then cut the cap off the board.
Remove the xenon lamp and the pulse transformer. Note where the black wire connecting the xenon lamp connects into the board when you remove it, because the next thing you do is desolder the lead of the remaining cap, that’s closest to the xenon bulb, and shift it over to where the black wire was. The cap is now going from high voltage to ground: it’s doing the same thing the electrolytic cap did, only with vastly less oomph so you don’t fry your geiger tube.
Now you desolder the piece of metal that acted as the charging switch, and where it was, you solder in some wires that go to a potentiometer in series with a power off/on switch.
And finally you add a 1 megohm resistor from the high voltage section, that’ll go to your geiger tube, and between the resistor and the tube, you put, in series, a small high-voltage cap, like a few nanofarads, and a 20:1 or thereabouts resistive divider to ground.
You can vary the output voltage both by varying the input voltage, and by varying the potentiometer, which controls the base of the transistor. The pot is nice but the transistor/transformer couple are somewhat tuned and you can quickly get into an area of enormous current draw if you mess about much with that. I’m using a 10K pot but it’s only actually at like 100 ohms.
Geiger tubes like somewhere between 400 and 1200 volts. The ones I have are specified at 900-1200 volts, but are working fine at 800 volts.
Gammascout.com has a nice diagram of sensitivity vs. voltage, that makes it clear why you want to work in the midrange of the tube’s sensitivity, here:
Right now I’m detecting the geiger counter activity using an oscilloscope. That’s not very realistic for portability, so you’d want to use a one-shot audio amplifier similar to what commercial geiger counters use.
Here’s a neat design using a 555 timer and a transistor to drive a speaker:
This page has a couple other suggestions for pure-transistor-based audio outputs:
The version we put into a Victoreen CDV700 survey meter (which originally was supplied with only a meter) is similar to that second one, and fits nicely on a 2x2cm board.
This is the audio section I’ve used in the CDV700:
Here’s a video of the geiger tube actually running, that I have to post on YouTube because of its size.