Jeff Duntemann's Contrapositive Diary Rotating Header Image

Steampunk Geiger Counter, Part 1

I haven’t really done much in electronics in the last year or two, and I miss it. So when stumbling around in my scanned schematics folder tree, I came across a couple of Geiger counter circuits that I discovered while scanning Carl & Jerry out of Popular Electronics a few years ago. What struck me back then was how little there actually is to a Geiger counter circuit, and, with a Geiger tube in hand, I could have a working counter in a couple of hours or less. (A really ugly clip-lead lashup might take me half an hour.) And although Uncle Louie gave me a Raytheon counter tube when I was 11, I can no longer find it. So up I went to eBay, and discovered to my delight that somebody was selling an Amperex 75NB3 counter tube. This is significant (nay, an omen!) because I’ve been looking for one for a while. It’s the tube called out in a circuit PE published in July 1955, with the cool Ed Valigursky prospector cover. (Scroll down to it.)


The circuit is simple; nay, minimal: Basically, a 375V DC source applied to the center element of a Geiger tube through a current-limiting resistor. Any time an energetic particle passes through the tube, it ionizes some of the gas inside (generally neon with some trace gases to sharpen the pulse by quenching the trail quickly) and for an instant the tube conducts. You can pick off a pulse through a blocking capacitor and hear it with sensitive headphones as a sharp click.

Getting 375V worth of battery is nontrivial these days, but also unnecessary. Note what happens above: The batteries do nothing but charge a couple of capacitors. A circuit I found in the July 1957 issue finesses high-voltage batteries completely by setting up an output transformer as a step-up, and applying interrupted DC to the output (low impedence) winding. The interrupted DC induces high-voltage pulses in the input (high-impedence) windings, and if you capture them in a capacitor, you can power the Geiger tube from a single D cell.


The one glitch is interesting all by itself: You have to pass the pulse through a spark gap. In this circuit it’s an automotive spark plug (remember those?) but it can be anything you can crank down to a thirty-second of an inch or so. In a similar circuit published in The Boys’ Second Book of Radio and Electronics, Alfred Morgan uses two small nails held in binding posts, and lacking a spark plug in the junkbox, that’s probably what I’ll do. The spark gap acts as a crude rectifier, making sure that only the positive excursion of the induced pulse goes to the hot side of the .05 cap. The interrupted DC is generated by repeatedly pressing and releasing a momentary-contact pushbutton switch. The transistor here is a headphone amp, but again, high-impedence headphones will make pulses audible direct from the tube.

So. Is this really a steampunk technology? In other words, could someone with some skill and knowledge have built one of these in 1900? (Again, the circuit does not require a transistor, nor even a vacuum tube to amplify the pulses.) I don’t see why not. You’d need somebody who understood ionizing radiation, but that’s no stretch for a 1900-era mad scientist in brass goggles. Neon gas in a graphite-coated glass tube? Transformers? Headphones? Kid stuff.

While I’m waiting for the counter tube, I’m going to lash up the cap charging circuit and see how it works. The output transformer called out in the circuit used to be present in every single All American Five clunkerjunker tube radio I found on the curb on Garbage Day, but you may have to ask around for used units, or spend (much) more on a new transformer from Antique Electronic Supply.

And to test it? In my box of gas-regulator tubes I have a couple of old units that were “salted” with something like a trace of Lead-210 to make them conduct instantly when power is applied. Such tubes aren’t very radioactive anymore because the salting materials have short half-lives, and it’s been 55 or 60 years since most were made. And hey, if they really are dead, there’s always a few cosmic rays floating around.

I was told recently that bananas are mildly radioactive. (It’s the Potassium-40.) This seems like a stretch to me, but…we’ll find out.


  1. Young Kwark says:

    That range of voltage might be reachable if you stacked the HV inverted outputs of two disposable cameras – the kind with a built in flash.

  2. Tom R. says:

    I am not sure if it is still true, but the mantles used on gas camping lanterns (e.g. Colman lanterns) used to contain a good bit of Thorium which is somewhat radioactive. That might be the easiest test subject.

    I am not sure about the stuff out west but I am only a few miles from a chunk of granite in Georgia called Stone Mountain and the background radiation from it is a bit higher than normal due to the Thorium content too.

    This sounds like a very cool project Jeff and we hope to hear more and see pictures soon!

  3. Masonry walls are typically slightly radioactive. However, to accurately calibrate the thing, assuming you’ve any desire to do so, you’ll need something like one of these:


  4. David says:

    Some smoke detectors would be a good source of radioactive material. Have you heard the story of David Hahn the “Radioactive Boyscout?”

    1. Yes, that was quite a story. I actually thought it was a hoax for awhile, but it does appear to be the real thing.

  5. Jack, K8ZOA says:

    One of the ‘gas traps’ should work as a reliable, repeatable spark gap. Available in a variety of striking voltages from Mouser, DigiKey, etc. Mouser PN 576-SL1411A075A for one example.

    For a 1900 version, a fast relay could be used as a synchronous rectifier. Called a ‘vibrator’ when used in car radios of the 1940’s and ’50’s. One set of contacts to chop the DC on the primary side of the transformer and a second set of contacts on the secondary side to synchronously rectify the HV. Or, of course, there’s always the dynamotor option, although a real power hog.

    Or, depending on GM tube and capacitor leakage, one of the piezoelectric lighters might work. I think they generate several kV as mechanical force is translated to electric charge.


    1. Vibrators! Heh! I have a couple in the box somewhere but didn’t think of them. Overkill, really, though the concept is certainly something they could do in 1900. I also think I have a couple of those little glass spark gap thingies, though I never knew crisply what they were used for; I was thinking shorting out voltage spikes across inductive loads when the field collapses. I’ll dig around downstairs later and see what turns up. I would like an adjustable gap so I can experiment with spacing, and one way to do that is pull the mica out of a mica trimmer (of which I have legion) and rig it so the screw adjustment tweaks the spacing between the two plates.

      I have a lot of used rotary wafer switches, some of them ceramic and quite robust. I was thinking of pulling the ball-clicker mechanism out of one so it turns freely and putting a crank handle on it, so with a couple of spins you break current fifteen or twenty times. Don’t know how well that would work but it’s worth a try, and I’ll report back here.

      1. Jack, K8ZOA says:


        The primary purpose of the gas traps in the communications business is over-voltage protection for voltage induced by lightning strikes. Unlike an MOV, the gas trap has essentially zero shunting capacitance (a pF or less). MOVs can have thousands of pF capacitance which limits their usefulness for high speed or high impedance circuits. I’ve written a bit about MOVs at and should do the same for gas traps, but haven’t gotten around to it yet. My preamplifier kits have gas traps on inputs and MOVs on DC power feeds. for example.

        A vibrator power supply requires arc suppression – or else the vibrator contacts have a very short life. The inductive kick from breaking the primary causes a significant transient voltage. Vibrator supplies commonly used a series RC snubber circuit. (I’ve also written about RC snubbers towards the end of

        73, Jack

        1. I went looking for gas traps today at OEM Parts (our local junk shop) and found nothing. I did get a couple of honking caps: 1.5 MFD @ 630V. That’s three times the capacitance I had in one lump with a rating over 600V. Will be messing with the spark circuit tonight, and will report my results soonest.

  6. […] barstool downstairs in my shop, drowning my pain in milk-jugs full of antique electronic parts. As I mentioned yesterday, having tried and failed to do so in 1963, I’m attempting to build a Geiger counter. […]

  7. Lee Hart says:

    An interesting project! Todd Johnson at Fermilab has built these, and has good sources of radiation for calibration. Talk to him for ideas. You probably have his address so I won’t post it here. Let me know if you need it.

    There’s a clever capacitive voltage multiplier circuit that would be good for this high voltage / low current project. Find an air variable capacitor with no rotational stop (butterfly caps are a good choice). Spin it with a crank or motor. Charge it with a 9v battery through a diode. Discharge it into a second output filter capacitor with another diode.

    At maximum capacitance, the battery charges it to 9v. At minimum capacitance (about 1/10th), conservation of charge Q=CV says the voltage must be 10x or 90v.

    The “diodes” can be any type; silicon, selenium, tubes, etc. They can even be mechanical switches actuated by the same shaft that spins the capacitor.

    The two sections of a butterfly capacitor can be used for two 10:1 multiplers, for 100:1 overall. 3.75v worth of batteries can become 375v.

    There is a variation of this design that requires no battery. It is self-exciting and can theoretically produced infinite voltage (until something arcs over). 🙂

    1. I’ve never heard of that capacitor trick before–does it require substantial capacitance? Most of the butterflies I have are very small and intended for VHF applications.

      Todd reads Contra, so if he wants to jump in I’d be honored to have him.

      One purpose of the project is to see just how primitive a counter I can make and still detect particles. If Mark I does in fact work, it makes a counter plausible in a steampunk technology setting, assuming you have a goggles guy who can create a neon-filled glass tube with a conductive interior coating. After Mark I proves out, I’m going to build Mark II with a two-stage tube audio amplifier so it can drive a speaker. If there’s a Mark III (we’ll see if I get tired of the whole business by then) I’m going to make one I can sling on my hip.

  8. Lee Hart says:

    Let’s see… Assume 4 x 9v batteries = 36v input, so a single 10:1 multiplier is enough for 360v out, and your capacitor is 10-100pF. When fully meshed, the capacitor stores Q = CV = 100pF x 36v = 3.6nC (nanoCoulombs). This would deliver 3.6nA for 1 second.

    Unmeshing the plates cuts the capacitance to 10pF. Conservation of charge means the voltage goes up to 360v. Interestingly, the *energy* stored increases by 10:1. E = 1/2CV^2; you added energy by mechanically pulling the plates apart. So it can still deliver 3.6nA for 1 second even at 360v.

    Cycling it open and closed once per second delivers a trivial current. But if you spin it at (say) 3600 RPM, it cycles 100 times per second, and you get 100 times the current; 100 x 3.6nA = 0.36uA. That may be enough for a Geiger counter.

    Butterfly caps are handy for this because they are balanced, and spin well at high RPM.

    PS: Is this inappropriate for your blog? I haven’t seen any math or technical discussions.

    1. Brilliant! I’d never heard of a mechanism like that, and if my Big Box of Military Surplus Caps still contains a chunky butterfly unit I remember (the box hasn’t been opened since I packed it in the fall of 2002) it would be worth lashing up on the bench just to see it work. Most of the others are little things the size of my thumb.

      Inappropriate? C’mon: You’re one of the best engineers I’ve ever met, and the idea here on Contra is to tell people things they didn’t already know. As far as I’m concerned, the more math and tech the better. What I object to is the sort of slobbering tribal hatred that more or less dominates the blogosphere these days.

      There’s something very George Ewing-ish about this whole project, and in one sense I’m doing it in his memory. If he needed a Geiger counter in a hurry, this is pretty much the way he’d approach it.

  9. Lee Hart says:

    Oops; I meant to say “spin it at 6000 RPM”, not 3600 RPM.

  10. Lee Hart says:

    George? He’d do it with a relay wired as a buzzer, switching the primary of a car ignition coil, and a 1B3 rectifier tube from an old TV set to rectify it! A flashlight battery would run it. 🙂

    I think your audio transformer only delivers ~300v because it has a significant amount of capacitance. In theory, instantly interrupting the current in an inductor makes the voltage head for infinity. In practice, you have a resonant circuit consisting of the winding’s inductance and its capacitance. It “rings” at this frequency, damped by the winding resistance.

    An ignition coil, photoflash transformer, TV flyback etc. is designed to minimize this capacitance. Try one of these types of transformers. You may have to add a new low-voltage winding on it to get the desired primary voltage.

    Also note: These types of transformer are designed to have significant leakage inductance (loose coupling between primary and secondary). This is so the primary won’t have such horrendous voltage spikes as the secondary. The equivalent circuit is an “ideal” transformer with separate inductors in series with the primary and secondary. The primary is then driven with a switch (transistor, tube, whatever) that has a capacitor in parallel with it. This capacitor limits the peak voltage, and the leakage inductor lets the “real” transformer winding voltages spike up much higher.

  11. […] Popular Electronics, July 1957: “Geiger Gun”. Compact gun-shaped hand-held counter counter using a CK1026 GM tube, pushbutton interruptor, and 2N107 transistor audio amp. Article is not online, but there are images of the counter as built in a junction box by someone here. (Scroll down.) Circuit is here. […]

  12. […] worst of it is down in my workshop, where I’ve done nothing ambitious in almost a year. (My steampunk Geiger counter was the sole exception.) Predictably, when something comes to hand and no good place for it is […]

  13. joe says:

    use a lord kelvin water dropper as a charger.

Leave a Reply

Your email address will not be published. Required fields are marked *