DIY Geiger Counter HV Circuit Not Working

[BlueApple]

I'm working on building a geiger counter using an old soviet glass tube. I'm using the first circuit at http://techlib.com/science/geiger.html. I built the circuit on a breadboard but using a 1GOhm resistor in series with my multimeter to measure the output I get 250v, but it should be 500v. The current draw is also 10mA, but it should only be 0.5mA according to the website.

Any ideas what the culprit could be?

[iMo]

I've been using the second schematics there (with the 10mH choke), and it works fine. The diodes in the HV multiplier are some high speed 1kV one, not the 1N4007 in my case..

[BlueApple]

I found the fix! I read that page over again and found where it said to swap one of the windings if it generates to low a voltage. I have a new problem though, my geiger tube needs a max of 460 volts, will feeding it with 500v hurt it?

[iMo]

My diodes are BYV-26..
I've been running the russian metal GM tube with a little bit lower voltage than the recommended plateau, afaik.
The 2-3 zeners in series set the voltage. When loading it with a resistor during the measurement it shows a little bit lower voltage than it is.
The higher the voltage and the higher the number of pulses detected the shorter the GM's tube life.

[BlueApple]

I tried hooking my tube up to the power supply then to my multimeter with the gigaohm resistor. Even when I place my test source on the tube the voltage didn't jump. Did I kill the tube buy hooking it up for a few moments to 500v?

[calmtron]

Do you have an anode resistor in series with your geiger tube?

[BlueApple]

Yes, a 4.7M.

I also tried hooking the geiger tube up to the base of an NPN transistor and wired a speaker to it. Is it possible the pulses are to short to turn on the transistor? Do I need to put a capacitor at the base or something?

[SuzyC]

If you can apply >500 V to a 460 Vmax Geiger tube without any breakdown, the tube is NFG.

BTW, you could post your circuit to help others to understand exactly whatya doin.

[BlueApple]

Here is my schematic, excluding the HV generator. (The switch is the geiger tube)

[BlueApple]

I think I discovered part of the problem, the seller sent me a CI-21BG not a CI-11BG (easy mistake to make, right?) but I still think my radioactive red plate would trigger it. Since the wrong tube was sent it could have been an untested/dead unit, as I see they sell those.

[iMo]

The pulses are short, like 50-100us, afaik. The amplitude of the pulses is something like 5-10V, it depends on the ratio of the resistors (a R2/R1 divider) - like the schematics below. I would decouple the amplifier with a capacitor.
Without a radioactive source you have to get several pulses per minute (like 1-5) as the background radiation (it depends where you live).
With a source - like a WWII watch or a WWII meter with radium filled painting on the dial - you may get 100x more pulses per minute, or even more..

[BlueApple]

I got it to work! I used a very sensitive darlington transistor in place of my standard npn.

[BlueApple]

Turns out my MPSA18 was in backwards, killing it. This must be why the current was to low to drive a standard NPN. I'm gonna order a new one.

[BlueApple]

How would I go about reducing the voltage? Just add some lower voltage zener diodes (say 20v) to the chain until it's around 400v? Is there a way I can calculate how many/what voltage I need?

[iMo]

The output voltage is approximately the sum of the zener's voltages in series multiplied by the number of steps in the voltage multiplier (x3 in your case). The actual voltage depends on the load.
The optimal voltage is at the beginning of the "GM tube's plateau", this point (the "threshold") has to be found experimentally (ie. by increasing the voltage slowly till the number of counts starts to be "constant").

PS: As you may see in the sim below the output voltage is rather bumpy with 10nF C4 and 100Meg load (as an example). You may smooth out the output voltage by increasing the C4 (and perhaps the other capacitors in the multiplier too), the average battery current will increase, however.

[iMo]

Below a more realistic scenario - with 2x 120V zeners, giving aprox 720V into a GM tube, where the 720V=(120V+120V)*3, while GM tube is receiving 5 pulses (5 particles detected) per second, each pulse is 100us long.
During the pulse the power supply works into the 4.7M ohm, outside the pulse it works into 1T ohm..

[BlueApple]

With 2x 100v zeners I get 425v, well within the operating range of my tube. The CI-11 tube is still pretty insensitive. I plan to get an LND-712 or a giant russian tube soon.

[TurboTom]

The SI-11BG is a tube designed for high radiation detection. You will typically find it besides a more sensitive tube in soviet era military radiation screening gear.
The LND-712 tube is very good and highly sensitive, also capable of detecting alpha radiation. But since it's a mica windowed tube, it's mechanically quite delicate.

A good, rugged and fairly sensitive tube is the russian/soviet SBM-20. It will detect beta and gamma (x-ray) radiation and it's widely used in contemporary, commercial radiation detection gear. These tubes are easy to find and quite inexpensive.

If you want a really sensitive tube, the SBT-10A may be a good choice for you. It's characterized as a highly sensitive (low energy) beta and gamma detector by the manufacurer, but it will also have some sensitivity for alpha radiation, though you've got to get really close to the source. Due to its huge mica window, it needs to be handled with care. I once designed a geiger counter around one of these tubes and utilized their cellular design with individual anodes to provide a uni-dimensional spacial resolution which proved more tricky than anticipated, but finally worked. Here on the geman language "Geigerzählerforum", I reported on that build... (you may use an automatic translator if required).

To increase the sensitivity of your instrument, you may always parallel several identical tubes. Use individual anode resistors in this case. It may also be beneficial to use capacitive signal coupling at the anode instead of the more traditional cathode-coupled current amplifier. Especially on large tubes, the massive cathode may pick up all kind of noise and cause signal interference.

[BlueApple]

Thanks for the suggestions! I don't think alpha detection is very important to me so I'll probably go will the SBM-20.

[BlueApple]

Have any of you ever used a CTC-5 tube? It looks very similar to the SBM-20 and it costs $15 less.

[TurboTom]

The SBM-20 is more modern and of much more recent production date than the STS-5. The SBM-20 is better in almost all aspects, the most relevant feature being the useful service life. The SBM-20 will last at least 2E10 impulses while the STS-5 is only rated for 1E9 impulses. Here's a russian language book that covers virtually all their radiation measurement devices: https://easyupload.io/pwk40c (will be taken down after 14 days).
You will find the relevant Geiger Müller tubes on pages 26 and 36.

I would always recommend to spend the difference on the more modern SBM-20. 

[BlueApple]

Thanks for the info!

[BlueApple]

I ended up buying a NOS SBM-20U (little cap on positive removed), for a very good price. Thanks for your help!

[BlueApple]

I just got in my SBM-20. It is very very sensitive. It responds very well to low radiation from things like uranium glass and trinitite (my piece of trinitite is very very weak). It goes crazy around my radioactive red fiestaware and radium dial watch!