Geiger Clicker radiation detector

Geiger Clicker radiation detector

After a couple weeks and two revisions, I was able to make a working Geiger clicker. The following is a detailed description on how it was made.


My first prototype/build was using a si-3bg geiger muller tube (shown below, at the top of the image), but I later found out that these are not very sensitive tubes, and will not detect background radiation as well as the SBM-20(bottom of the image) which I ultimately ended up using for this project.
CORR2556The SBM-20 is more expensive than the aforementioned, but is well worth the money, as it is highly sensitive. I purchased mine on ebay for around $15. The SBM-20 needs a minimum of 400 volts to work, while the sensitivity increases as the voltage increases up to 450v. Don’t give it more than this or you could risk killing your tube!

Many people use a flyback transformer setup to achieve the required voltage, but I have had horrible luck with this and decided I would be better off using the tried and true boost converter setup that I use for all my HV tube stuff.



Above you will see the high voltage generation part of the circuit. For L1, I used a coilcraft 100uH coil. The HV cap was something that I bought years ago on ebay (I got a ton of ’em). Just make sure your cap isn’t too big, as the circuit only needs a small cap.  The rest of the parts are as shown. There are some useful tips on building the boost converter from this circuit in my article, A Nixie Clock.

The driver for the IRF830 in the previous schematic can be seen below. The frequency is determined by R1, R2, and C1, as well as pin 5, the control voltage that sets the 2/3 point on the threshold pin 6. You don’t have to understand that, just understand that as the transistor conducts, the frequency of the FET_DRIVER, pin 3 raises, which lowers the output from the boost converter above.


The feedback for the above circuit comes from the below schematic. If you want to understand it better, I suggest reading Flyback Convertors for Dummies.


The resistors above were chosen to make the output on the HV cap anywhere from 200-800 volts (not that it would actually go that high). I’ve measured over 500 volts from this exact circuit though, but I didn’t want to push it.

Just a side note on measuring the voltage from this unit – when measuring low current high voltage, you must take into consideration the loading of the circuit from your meter (usually 10 megaohm – sometimes less!) Not only will it load the circuit and bring down the voltage (so you won’t know the actual voltage your circuit is producing), but if measured after a resistor, it will form a voltage divider and change your reading as well.


Take this very circuit for instance. If I were to measure the voltage at the GM tube anode, then to my 10M internal resistance meter, it would look like the below circuit:


Since the GM Tube will (usually) not conduct, the circuit simply looks like two 10M resistors in series. When  measured with the multimeter, it appears as being half of the voltage it actually is! Just keep this in mind. Also, not every meter is 10M. If you’re using a cheap multimeter from Harbor Freight (a.k.a. China), I’ve seen them as low as 800k internal resistance!!

The 100k-220k resistor on the cathode of the GM tube is what provides us with the pulse_out, which is connected to the below circuit to amplify the signal.


This amplifies the signal that would otherwise be very weak. I then chose to hook the bigger pulse, big_pulse, to the below circuit.


This is a 555 monostable that takes the pulse and stretches it out, making it easily audible using a piezo element or speaker and visible using an LED.


And finally, in the above, you can see the click acting on a Piezo and LED (chosen to be amber for  effect).



Above is the full schematic. You may click on it for the large version. Below you will find the board layout. I must note that this was the second revision of this circuit. The first did not utilize a ground plane and the output from the boost converter was too close to the pulse detection part of the circuit. The below circuit works very well. (click for larger version)


Please feel free to post any questions or criticisms!