We were very surprised by the crude nature of the device.

Interesting story about SparkFun. When we were developing the LCD-90 we found SparkFun website very helpful with good tutorials. Back in the October 2008 time-frame, we even utilized one of their services, BatchPCB, while developing the LCD-90’s printed circuit board. We exchanged several e-mails with some of the principles of SparkFun regarding the circuit board. About a half-year later SparkFun developed their own Geiger counter so we took a look at it.

Some interesting things we found. The high voltage section has no voltage regulation. That means the higher the radiation level, the lower the high voltage. Very bad design for accurate data but maybe OK for a toy random experiment, but even then it's a bad design.

The SparkFun tutorial detailing some of the features of their Geiger shows this image of the pulse output from the prototype:

The author assumes all the extra pulses following the one large pulse, are coming from the Geiger tube. This is not the case. Good Geiger tubes do not create additional pulses as shown. All the extra pulses are coming from his high voltage generator as it pumps the voltage up after the one large detection pulse from the Geiger tube. This is partly due to the circuit’s total lack of any voltage regulation.

His solution was to add a 4.7uf capacitor to the output, which stretches the output pulse to such a long pulse that it can even light a LED long enough to be seen. Pulses from Geiger tubes are so short in length that an output derived from them, driving a LED, would never been seen.

A major problem with such a scheme is all other legitimate pulses from the Geiger tube, that occur during the long output pulse from his filter, are missed and in fact would just elongate the filtered output pulse, exacerbating the missing of counts. Given a high radiation level, his circuit would probably output just one very long pulse for thousands of detection pulses from the Geiger tube, i.e. at some point, the higher the radiation level, the less the number of counts recorded by his circuit. It would be very hard to tell if your data is good or at least how many counts are missing from your data.

There is another problem with the circuit. He uses the output pulses to trigger interrupts in the ATMega328 micro-proccessor. His ATMega328 code also uses another interrupt to keep track of time. This type of code can miss closely grouped interrupts.

In contrast, Aware Electronics Corp. group of RMs include a highly stable, very high efficiency, fully regulated and very clean high voltage section and generate output pulses that match exactly the dead time of the tube in use (typically on the order of 10 to 50 microseconds).

In addition our USB-MSP and LCD-90 utilize high speed CMOS counters to perform the actual counting and not an interrupt driven scheme as used by the SparkFun gadget.

We have noticed many Johnny Come Lately makers use basically the same schemes for high voltage generation and pulse counting as the SparkFun gadget.

Our conclusion is for anything but the most basic toy-like monitoring of radiation there really is only one choice.

Aware Electronics Corp. Twenty plus years perfecting radiation monitors and software for the professional and public at large.

E-mail
Phone:(800)729-5397 or Phone/Fax: (302)655-3800

To return to the previous page, use your browser's BACK button.

Return to Interface page?