Home arrow Blog arrow Misc arrow Geiger Experiments - Power Supply
Geiger Experiments - Power Supply | Print |
Written by Akiba   
Monday, 21 March 2011

Things are calming down somewhat here in Tokyo and the local foodbank said they have enough volunteers for the time being. So I figured I would get to work on helping out with the collaborative geiger counter project being hosted at SEEED studio . Tokyo Hackerspace has 10 SBM-20 geiger tubes on their way and we need to get to work designing the circuit schematic and PCBs for the geiger counters. We have a status update on all the projects that are going on currently due to the past events and you can see it here:


Anyways, the first order of business is generating the 500V DC needed to run the geiger tubes. This is one of the main challenges since it's not easy generating those types of DC voltages. Luckily, the geiger tube does not consume a lot of current. The 500V is just used to set up an electric field strong enough to generate an avalanche process. The quick background is that a photon ionizes a molecule into positive and negative charged components. Under normal circumstances, they would just recombine immediately. However in a strong electric field, the charged components separate and move towards their respective sides of the electric field. If the electric field is strong enough, ie: the voltage is high enough, then the charged particles ionize other particles in the gas inside the tube. This becomes an avalanche effect and when all the charged particles hit the walls of the electric field, a voltage pulse can be detected. This becomes an "event" and a geiger counter counts the number of events per minute. 

So the first order of business is to generate a voltage high enough for the geiger tube to operate properly. I did a few experiments with some components I had lying around and came up with some fairly interesting results.The first experiment I did was using a standard power supply transformer. It converts a 100V AC sine wave to a 3V AC sine wave. The effective ratio is 100:3 (33:1). I ran it in reverse where I drove the 3V side of the transformer with an AC square wave. The square wave was generated with two pins of an Arduino, each switching on and off digitally to generate two signals with opposite phase. The frequency was approximately 1 kHz. I had power from my lab supply fed into the center tap of the transformer. The output of the transformer was then fed into a bridge diode (Vishay, DF06M) and the +/- outputs of the bridge diode were measured on my oscilloscope.The following waveforms are raw waveforms. I didn't include the smoothing capacitor on the output of the bridge diode since I don't have any 500V caps at the moment. I'll be buying high voltage caps shortly and should be able to get better waveforms.

A word of warning. These are high voltages and are potentially dangerous. You should take proper safety precautions when working with high voltage and also should protect your equipment. For my oscilloscope, I was using my probe in 10x mode (divide by 10) which allowed me to stay within the oscilloscope's max voltage rating of 120V. 

Here is the schematic of the first circuit. 

Here is the actual setup. It's ugly, but then I just wanted to get an idea of how difficult it would be to generate 500V:

Here's the first waveform. This one had my bench supply at 9V. The circuit consumed 30 mA at no load, most likely due to losses in the transformer windings. The ouput voltage was approximately 220V. The spikes were due to the sharp transitions of the square wave:

This waveform was generated using the same circuit, a bench supply of 12V, consuming 40 mA,and outputting 300V:

I then tried a different approach. A lower cost method would be to use EL wire inverters. These devices already have all the circuitry to take a DC supply and generate an AC output. From there, the output can be fed into a bridge diode for rectification and then smoothing cap. Here is the circuit for the EL inverter:

Whereas the first circuit had both transformer outputs alternating, the EL inverter only had one "hot"wire at high voltage. The other wire was neutral and had no voltage on it. When I plugged it into the bridge diode, I got a nice surprise. There was a large DC offset with the AC signal riding on top of it. The circuit wasn't under load so I don't know if the DC offset would still be as large under the load of a geiger tube. However if it is, and we were to smooth the AC signal with a capacitor, then the EL inverter would be a much better option to generate the 500V supply. 

Here's waveform of the EL inverter using a 5V input from the bench supply. It consumed about 50 mA. The output was at about 470V. You could also tweak the voltage to get a higher output, but I'm trying to stay within standard AC adapter voltages. That's why I'm using 5, 9, and 12V. Here is the waveform for the 5V input:

I'm still not sure why the huge DC offset is there, but if it looks the same under load, then I'll take it. 

I'm going to pick up the rest of the parts for the supply and hopefully the geiger tubes will come in this week. After that, I can try and generate a smooth ~500V supply to test the geiger tube with. 

As for the parts, the diode bridge was from Vishay, part #DF06M. The power supply transformer I used in the first schematic had a 100:3 ratio and cost approximately $7 in Akihabara. I don't know the manufacturer. You can probably get something similar for much less through wholesalers. The EL inverter was from WanYu and has a part number of WY-ELI-IUC-4.5. The last number is the voltage rating. My quote is for ~$1 in quants of 100. It's also nice because it has mounting ears so they can be screwed down to a PCB. I'll be doing more tests with it once I get the geiger tube. If it can successfully generate a 500V DC supply to run the geiger tube, then I'll be sending any extra inverters I have to SEEED studio, Libelium, and other people that are collaborating on this design. I only have a few extras though. Here is a picture of the inverter:

That's about it. This post is mostly for the OSHW collaborative geiger counter project, but hopefully, others may have enjoyed it too :)

Hits: 16184
Comments (2)Add Comment
Power Supply
written by M. Simon, April 01, 2011
The power supply from a flash camera would probably do the job. Get rid of the flash capacitors and replace them with something on the order of .1 uF 1,000 V. Add a trimpot to adjust voltage. You can probably get the used cameras for nothing at a photo developers lab.

Or you could roll your own with a Mouser flash trigger transformer and a suitable circuit.

report abuse
vote down
vote up
Votes: +1
High efficient 5VDC in/ 500VDC out solution by Radgoes
written by aikifredo, May 26, 2011
Hi Akiba.

Checked with Radgoes who also bought old Geigercounter FH40T like me.
He published a very goed low current consuming 500VDC source here:
Think the point is to tune in the primary winding at its resonance frequency.
report abuse
vote down
vote up
Votes: +1

Write comment

  No Comments.

< Prev   Next >