|
|
| The completed counter | It's a tight fit in there! |
|---|
This project was based on a geiger counter kit sold by Oatley Electronics in Australia. As far as I can tell from their website, they no longer sell this kit.
The original circuit utilised a simple pre-built module to generate approximately 500V from a 3V battery supply to drive the geiger tube. The output of the tube drove an 8 ohm speaker via a FET, and drove an open collector switched output.
A disadvantage of the inverter module used is that it draws over 200ma from the 3V battery.
The tube is a thin-walled model of Russian origin, it is sensitive to beta and gamma radiation between 20KeV and 1.2MeV.
Obviously a geiger counter that simply clicks is of limited use, so my primary aim was to provide a counter and display to the basic circuit. A secondary aim was to try and fit the whole mess into a plastic case I picked up at a local surplus dealer. This case was probably intended to house a DVM or some similar small instrument.
I used a PIC 16F84 to count the pulses and drive an LCD display. The PIC input can't detect the pulses directly, as they are too fast and of insufficient amplitude. I used a CD4093 CMOS Schmitt trigger to condition the pulses and drive the PIC input line. There was no particular reason for choosing the CD4093, except that I happened to have one lying around. Any CMOS Schmitt trigger would probably do the job.
The Schmitt trigger gives a strong 5V pulse which is used to drive the RB0/INT line on the PIC, which generates an interrupt.
The output of the PIC drives a Matrix Orbital 8 character by 2 line LCD display with an LED backlight. I would have liked to use a 16 character or larger LCD display, but the 8 character was the only one that would fit in the case I was using.
The whole PIC plus LCD circuit draws 75ma at 5 volts, and 60ma of that is the LCD backlight. I could improve battery life by making the backlight switchable, or if I really wanted to get clever, I could use a photodiode to detect ambient light and have the PIC turn on backlighting only when needed. However battery life is around 6 hours with the backlight on, which has been quite adequate in normal usage.
The orginal speaker was replaced with a low-profile 8 ohm Fuji speaker. This was the best compromise I could find between size and output. Larger speakers are more efficient, giving a much louder sound for the same input than smaller models. I tried a number of tiny speakers, but the Fuji was best because although it's a full 4cm diameter, it's very thin [only 4mm]. Although not as loud as the original speaker, it is the loudest I could find that would fit in the space available.
The geiger tube is housed in an external PVC tube with holes drilled in it so that the housing doesn't attenuate the beta radiation too much. The tube housing attaches to the main unit by means of a BNC connector. A co-ax extension can be used if the tube is to be remotely mounted.
The 5V supply for the logic and LCD is provided by a stack of low-profile 450mAh NiCd cells, [this is the square black object in the lower left of the picture - it's rather hard to make out against the black case]. The tube supply is provided by two alkaline AA cells. This is one aspect of the circuit that I really don't like. Having two battery supplies is inconvenient, especially as I couldn't fit the AA cells into the battery compartment of the case, so the case has to be opened to change them. An enhancement I'd like to make is to generate the high voltage supply from the 5V so I can have just one rechargeable battery. In order to have reasonable battery life, I'll have to do it more efficiently than the present inverter, but I'll probably still have to use a higher capacity rechargeable battery.
The PIC software counts the interrupts and updates the display every 15 seconds with the total count for the previous minute. A running average of the counts for the previous five minutes is also displayed every minute.
A possible future enhancement is to add an RS-232 interface so that data can be logged on a PC. This would require addition of a level-shifter such as a MAX232. It would only be practical to do this once the system is running from a single power supply, as datalogging would best be done running from external power rather than batteries.
This little project is truly international. The tube drive circuit is Australian [assembled in New Zealand], the tube is Russian, the speaker and nicad batteries are Japanese, the high-voltage inverter came from Hong Kong, the PIC and LCD display module are from the US, and the finished unit was assembled in California.