9v to HV supply for Geiger counter

I'm need a HV supply for a geiger counter.
I was thinking a 9v battery with a 555 oscillator with a wall socket 9v output transformer used backwards.
However i'm not sure how to get the proper output required for the Geiger Muller tube and need some professional input :smiley:
Most tubes on ebay have ratings of around:
Working Voltage 380-460V dc
Working Current 0.015-0.02 mA.
Any help is appreciated.

How about dismantling a throw-away camera and using the flash unit power supply. Runs off around 1.5 volts and gives several hundred volts

This is a schematic from an article I wrote in 1986.
The text is here:-
http://www.doc.mmu.ac.uk/STAFF/A.Wiseman/Acorn/BodyBuild/BB_86/BBC40.txt

Grumpy_Mike:
This is a schematic from an article I wrote in 1986.
The text is here:-
http://www.doc.mmu.ac.uk/STAFF/A.Wiseman/Acorn/BodyBuild/BB_86/BBC40.txt

Very nice!!
What kind of frequency and waveform did you use on bit 7. Is there an optimal one?

I have used the disposable flash method, which worked well, but also this approach, which used the PWM output of the Arduino in my Arduino + Android Projects for the Evil Genius book http://www.duinodroid.com. You can then tweak the voltage by adjusting the duty cycle. I did try to close the loop with a potential divider and an analog input, but even with 10M? the load of the potential divider droppend the output too much.

Note this a rough and ready design, with no optio-isolation from the Arduino, and is fine if you just want to make something that beeps when an event happens and will accept the consequences of potential destruction if you get it wrong.

WARNING !! HV electronics is dangerous - keep one hand in your pocket.

What kind of frequency and waveform did you use on bit 7. Is there an optimal one?

It was a square wave just like you will get from a PWM pin on the arduino.
I built up about 30 of these so it is a reliable circuit.

Many years ago, as an apprentice in the nuclear industry (pre-transistor days) we used to use either 60 or 90volt ( I can't remember which) dry batteries to power the GM tube detectors. It is essential to plot the tube characteristic to find the plateau, as Mike's article shows, so that you get a linear reading. Otherwise you'll be getting bleeps that mean not a lot, especially if your drive voltage is at the upper end or beyond the plateau.

jackrae:
Many years ago, as an apprentice in the nuclear industry (pre-transistor days) we used to use either 60 or 90volt ( I can't remember which) dry batteries to power the GM tube detectors. It is essential to plot the tube characteristic to find the plateau, as Mike's article shows, so that you get a linear reading. Otherwise you'll be getting bleeps that mean not a lot, especially if your drive voltage is at the upper end or beyond the plateau.

http://www.ebay.co.uk/itm/Russian-GEIGER-TUBE-COUNTER-SI-3BG-Oty-1-NOS-Calmps-/190601699514?pt=BI_Security_Fire_Protection&hash=item2c60bf20ba
The above link is what I am looking could you explain in terms of that please?

Also could use a 555 in astable mode for my square wave?
Do I need a stable DC voltage output? (i would think you do to get a nice signal output but just like to be sure)
Could I use a monostable 555 with a speaker on the output instead of my arduino?

The above link is what I am looking could you explain in terms of that please?

You need the data sheet, that will show the characteristics of the tube. All you can gather from that link is that the working voltage is between 380V and 460V so aim for 420V.

Also could use a 555 in astable mode for my square wave?

Yes.

Do I need a stable DC voltage output?

Yes.

Could I use a monostable 555 with a speaker on the output instead of my arduino?

Yes or just connect a transistor to a speaker with a capacitor. BUT what is the point in hearing the click, you need the arduino to count the pulses and then apply a dead time correction to get the true count.

Gentlement... In the Elektor - November 2011 issue - the site -http://www.elektor.com/magazines/2011/november/improved-radiation-meter.1971931.lynkx and a June issue - the site -http://www.elektor.com/magazines/2011/june/measure-gamma-rays-with-a-photodiode.1810862.lynkx. Simply use a photo diode, amplify the signal and use a microcontroller to count the pulses. I think it is worth investigating. For me, that sound like a interesting project to do.

It sound interesting but the comment:-

Small signal levels make this design an interesting challenge.

In the only credible reference is a bit of a caution.

Thank for your expert opinion Grumpy_Mike. The amplifier circuit is the challenge in my opinion. Using JFET transistor and JFET op-amp should be OK to use and proper construction techniques should be fine.

zer044,
You might check this circuit out. (credit to Tom Napier for the original)

It supplies a regulated 200-610V without using a transformer and only draws a few mA.
BTW, I wouldn't suggest the SI-3BG tube. It is one of the least sensitive.
I have a comparison and sources of GM tubes here.
John

Grumpy_Mike:

Do I need a stable DC voltage output?

Yes.

From what i know a square DC wave will induce an AC output that peaks on the falling and rising of the square wave.
Now I see from your circuit that the rectifier diode will remove the AC part but still leave you with spikes instead of a steady direct current, how does your circuit make your output HV a steady direct current?

bHogan:
zer044,
You might check this circuit out. (credit to Tom Napier for the original)

It supplies a regulated 200-610V without using a transformer and only draws a few mA.
BTW, I wouldn't suggest the SI-3BG tube. It is one of the least sensitive.
I have a comparison and sources of GM tubes here.
John

I am a yr 13 student just trying to understand the concepts of electronics, my knowledge only goes as far as A-level physics and have a small budget :frowning:
Thats why I chose a cheap Geiger Muller tube to get started with.
But the link you sent me looks very interesting, definitely will get to it as soon as i can get a basic Geiger counter working.

part but still leave you with spikes instead of a steady direct current, how does your circuit make your output HV a steady direct current?

The 0.22uF capacitor smooths it to give a steady voltage.

will I need to change any of the circuit parts for Grumpy_Mike's circuit to make it give a specific voltage for me or will I be doing that through the pulse width modulation on my square wave?

I cant notice any voltage regulating parts on the HV side of the circuit.
The rectifier diode is rated at 1000V so i'm assuming he just chose one that will be good enough for the upper limit?

So considering i want around 400Vdc should i use the 1N4004 diode or something thats higher?
From what i know the capacitor value of C1 will change the average DC voltage from the smoothed output, so should i change this to suit my needs?
Everything else seems to be current limiting and signal stuff.

The value of capacitor will have no effect on the final voltage. It will affect how much ripple is on the voltage.
You don't regulate HT in the samw way as lower voltages.
I would go with at least a 1KV diode, don't run diodes close to the limit.
The transformer winding ratio will determine the voltage, you can reduce it by taking the pulse width either side of 50%.