A No-Solder, Scrap-Bin Geiger Counter For $15

Scenario: your little three-hour boat tour runs into a storm, and you’re shipwrecked on a tropic island paradise. You’re pretty sure your new home was once a nuclear test site, but you have no way to check. Only your scrap bin, camera bag, and hot glue gun survived the wreck. Can you put together a Geiger-Müller counter from scrap and save the day?

Probably not, unless your scrap bin is unusually well stocked and contains a surplus Russian SI-3BG miniature Geiger tube, the heart of [GH]’s desert island build. These tubes need around 400 volts across them for incident beta particles or gamma rays to start the ionization avalanche that lets it produce an output pulse. [GH]’s build uses the flash power supply of a disposable 35mm camera to generate the high voltage needed, but you could try using a CCFL inverter, say. The output of the tube tickles the base of a small signal transistor and makes a click in an earbud for every pulse detected.

You’ll no doubt notice the gallons of hot glue, alligator clips, and electrical tape used in the build, apparently in lieu of soldering. While we doubt the long-term robustness of this technique, far be it from us to cast stones – [GH] shows us what you can accomplish even when you find yourself without the most basic of tools.

30 thoughts on “A No-Solder, Scrap-Bin Geiger Counter For $15

    1. Also, only 3 hours steaming/sailing from civilization??? better off lighting a fire…

      Jack Sparrow: No! Not good! Stop! Not Good! What are you doing? You’ve burned the food, the shade. The rum.
      Elizabeth Swann: Yes, the rum is gone.
      Jack Sparrow: Why is the rum gone?
      Elizabeth Swann: One, because it’s a vile drink that turns even the most respectable men into complete scoundrels. Two, that signal is over 1,000 feet high. The entire Royal Navy is out looking for me. Do you really think there is even the slightest chance they won’t see it?
      Jack Sparrow: But why is the rum gone?

    2. The others won’t be inclined to eat the Professor when the food runs out. Gilligan and the skipper will be the first to go. And you’ll have a nice high voltage source to dispatch the others before they get you.

  1. One of William Gibson’s books mentions kids making stun guns out of disposable cameras. That sounded plausible to me.

    On the subject of radiation I think I’m paraphrasing Korf when I say that a fork and a spoon, properly arranged will count (meaning detect radiation). The large round alpha sensitive detectors we used in school here are mica window, metal cylinders with a sharp pin in the middle (they have low pressure gas inside, typically a noble gas and a trace of halogen), and it’s the small volume near the point of the pin the amplification happens. But if you have a metal container and a pin, or even just a fork and a spoon you can arrange them to have a Geiger plateau. The active volume may be a bit small and the voltage needed maybe a bit high but it will detect radiation with plain 1 atm air.

    1. It’s fairly easy to build something that will click in the presence of radiation. It’s only a tiny bit harder to make it reliable and fairly-well calibrated, but almost no one does this.

      He’s putting 600V into an SI-3B tube, which has a plateau spec of 380-460 volts. This is sufficiently above the plateau of the tube that most (some? all?) of the clicks will be due to runaway discharge, and each click will be reducing the life of the tube. He should bring the voltage down to the plateau range.

      All tubes have a plateau, but they vary somewhat so that you can’t rely on the spec’d voltage. Power supplies should be tuned to put the tube into the plateau. Place a sample in front of the tube and slowly ramp up the voltage while counting pulses per minute. You will start with none, then a few, then a fixed PPM over a large range of voltage (the plateau), then a lot as the tube goes into runaway.

      Set the voltage to the middle of the plateau, and use a calibrated source if you have one.

      At 600 volts, he’s running his tube in the “runaway” range.

      The tube power supply should be DC and fairly well regulated (high and low voltages will give you more or fewer counts, as noted. The output of a camera flash is most definitely *not* DC, and he’s not smoothing the output with the big camera capacitor.

      And finally, I don’t like to harp on low-risk safety aspects, but note that he’s discharging 600 volts through a B-E junction of a 2n2222, which is spec’d for a maximum of 40 volts. With no current limiting resistor, this is almost guaranteed to punch through that junction.

      You can be killed by 9-volts if it gets through the skin – there’s a famous case of someone killing themselves by measuring the conductivity between opposite thumbs by using pins and a voltmeter – and the earpiece is worryingly close to the blood supply through the eardrum, if 600 volts punches through the transistor and gets into the earpiece.

      A camera discharge inverter packs a lot of power, while a regular geiger supply uses a C-W inverter with almost no current, and also doesn’t have as much of a direct connection to the headphones (multiple transistors, current limiting on the base, or transformer isolation of the output signal).

      I hate to be the guy who tears down someone else’s project, but there’s a lot wrong with this and a chance it could knock you down in the right circumstances.

      (And yes, I have built geiger counters, and helped people in Japan with their designs after Fukishima.)

      C-W inverters aren’t that hard, they’re much safer, and they use much less power. Smooth it to DC, use the right voltage for your tube, and put an amplifier stage or transformer between the tube and your ear.

      1. I agree there are better ways to create the DC voltage needed by a Geiger Mueller tube, however this was to represent an off he cuff junk box build. Personally I have to consider that “famous case of someone killing themselves by measuring the conductivity between opposite thumbs by using pins and a voltmeter” as infamous. Not that I’m saying that’s improbable, but I have yet to read an account of the incident that I could consider reliable

      2. Death by 9v aside, that’s a good analysis PWalsh. In addition to using 600V to run a 400V tube, he should have a limiting resistor in series with the power supply. Most of these are in the 5 to 10 MOhm range and without which the tube won’t quench. Might also explain why he is not smoothing the output, the voltage would build up and dump into the tube. With the details as given this circuit should not work. Maybe the voltage isn’t as high as he thinks, or the plateau extends much further than documented and the gap in the pulses from the free running oscillator is enough to quench the tube.

  2. Spiderwort as been touted as ä radiation detector, at least around the time of Three Mile Island. No soldering needed.

    Around the same time, Popular Electronics had a project about a solid state radiation counter, I thought of the same sort as a real Geiger counter. But I can’t remember what was used as a sensor. That might be more findable than an actual Geiger tube.


    1. You can use just about any diode as a radiation detector, keeping in mind if the covering will block Alpha and/or Beta. PIN diodes are popular, and there are many circuits online that use these.

      You can also use camera CCD arrays as detectors – radiation can flip a bit in the image array, and be seen as a white pixel.

      I cut the tops off of power transistors and use the bare B-E junction as detectors. This requires a more subtle circuit, but it works very well for Alphas and other high energy particles.

      You can also use a scintillation plastic and a photomultiplier. PTFE – the stuff that water bottles are made of – is one of the better plastics to use for this purpose.

      The big issue is “aperture”, which is the window through which the particle has to go to be registered. A metal can 2N2222 transistor with the top cut off can be used as a detector (I’ve tried it), but the junction is so tiny that the aperture is microscopic and you’d need a lot of radiation to have any reasonable chance of detection.

      Usually people use something with a big surface area such as a BPW21R photodiode.

          1. It seems that PET alone has a peak in the UV range (not ideal for most PMTs), the prefered material is a blend of PET and polyethylene naphtalate, which seems to be an additive for some heat resistant blends of PET…
            I’d personally suspect the dies used for coloring the plastic bottles will also play a very significant role in scintillation properties.

            Anyway this is a neat tip, thanks for that ;-)

  3. The other problem is that the tube used isn’t quite sensitive, it’s specified for 300R/hr (3Sv/hr). You would hear some clicks in the Fukushima reactor, but it’s useless for detecting radioactive food. In fact, with these doses, you’d probably start feeling the copper taste in your mouth before some clicks registers (that’s how Chernobyl bio-robots described their experience on that famous reactor rooftop).
    There are much better tubes – and more expensive, of course.

  4. The best bet is a sealed to light scintilator and solid state flash detector, no high voltage to design around and the simplest PIC will work and could last for a decade on a small lithium cell. Failing that there is the no-battery accurate dosimeters called the Kearney fallout meter. It just needs a comb and hair or a roll of tape to charge it, the Kearney meter was in a DIY lab book from the 50s which I found in a book sale as a kid, it also had making a cloud(Wilson) chamber, flower pot carbon arc furnace, making lab glassware, and many other other HAD’esque lab tool making hacks.

      1. There’s actually an Android app that turns your phone into a relatively decent geiger counter. The trick is to cover the lenses with a duct tape, so I’d say that the phone lenses aren’t a big issue. Of course that limits your detection abilities to hard beta and soft gama, but typical russian tubes are about the same. For alpha, you need a decent pancake probe but that’s around $100 at least.

    1. If you have one in your junkbox, sure, but I would not start from scratch with a crappy tube like that, unless you are looking for something small. For example my combo UV flashlight/geiger counter for flea marketing would be OK with that tube as it is assumed that you would “probe” anything that looks interesting with the geiger counter end of the flashlight. That tube would not be sensitive enough to walk down the aisles with and be alerted to something feet away on a shelf.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.