Space Detector Part 2

  27 Dec 2012


(Read Part 1 of this writeup)

High Voltage Power Supply

Geiger tubes require a relatively high voltage, 360-420V specific for the SI-180G. It turns out HV supplies for Geiger circuits are pretty simplistic. My first resource was the DIYGeigerCounter, which is based on a Nuts and Volts article by Tom Napier. I had a lot of trouble with the 555 approach for some reason, and I wanted to use a microcontroller anyway for USB communication, so I replaced the 555 with my favorite micro at the moment, a Teensy 2.0 and used a PWM output to drive the power transistor. I hooked up the tube and got this:

Bad Tubes! Bad!

Tube Trouble

Not what I expected. It seemed like the tube was ionizing once and then continuing to oscillate, rather than provide nice clean individual spikes for every particle interaction. To the forums! The friendly folks on the GeigerCounterEnthusiasts message board helped me diagnose a borked tube. My best guess is that the "quench gas" (used to quench the cascade effect caused by particle ionization) had reacted with the walls of the tube over time (as suggested by forum user mirroromatic), meaning a particle would trigger a cascade in the tube that would continue until the voltage was dropped. The result is that the tube basically turns into an oscillator, giving the waveform seen above.

I confirmed that one of the tubes was working fine, but the other two were not. I got on the email-horn to Boxtec in Switzerland and explained my predicament. They immediately promised to re-test all their tubes, and they were even gracious enough to send me three new, known-good tubes, including overnight shipping so that I could complete my project. Thank you Christoph and Boxtec!

The PCB

I had my new tubes the following Monday, and started laying out a circuit board for the circuit, including three high-voltage supplies, three detection circuits, and two NAND gates to handle the coincidence detection:

PCB

It took about an hour and a half to mill it out on our spiffy Shurline mill, immediately after which I realized I had made a rookie mistake: I had routed everything on the top layer. Oops. My deadline was looming so I decided to go ahead and solder everything from the top rather than mill another board. In the end, it took longer to put it together than if I had just redone the board, but it worked. One question: do I really need three separate HV supplies, or can I get away with one? I didn't have time to test, so hopefully someone will let me know.

At the last minute, I added a four-digit LED display to count the number of Muons detected. I used a MAX7221 LED driver because it has a really easy Arduino library, and I freehanded the circuit on the back of some protoboard. In retrospect, it would have been easer to design and mill such an easy board. Oh well. I also added neat LED on/off button and used one of the remaining PWM pins to do a little "breathing" effect while it's off.

Testing