Geiger counter A/D conversion for radiation level crowdsourcing


[Akiba] and the crew at Tokyo Hakerspace are still hard at work trying to help out their fellow countrymen after the recent earthquake, tsunami, and ongoing nuclear crisis in Japan. You may remember the group as they are behind the Kimono Lantern project we featured last week. This time around, their efforts are focused on getting usable information out to those who need it.

With all of the talk about nuclear fallout, they wanted to see what sort of measurements they could get in Tokyo, however they could not locate a Geiger counter anywhere nearby. Luckily, they were eventually able to source two old counters from the Reuseum in Idaho. One is being lent out to individuals in order to check if their home’s radiation levels are safe, but it was decided that the other would reside outdoors in order to collect radiation readings from the air.

[Akiba] wanted to put the results from the external Geiger counter up on Pachube, however these old units are all analog. He figured that a quick and dirty way to do analog to digital conversion would be to monitor the chirps coming off the counter’s speaker. This was done by wiring up an Arduino to the speaker leads, and keeping track of each time the speaker was activated. This resulted in an accurate digital radiation reading, matching that of the counter’s analog display. The Arduinio wirelessly sends the information to another Arduino stationed inside his apartment, which then uploads the data to Pachube.

A walkthrough of his conversion as well as the source code for both the Arduino counter and the Pachube uploader are available on his site, in case anyone else in the Tokyo area has a Geiger counter handy and wishes to do the same.

17 thoughts on “Geiger counter A/D conversion for radiation level crowdsourcing

  1. The calibration of these counters is for the meter reading of clicks per minute.

    This is distinct and has nothing to do with the actual sensing of clicks. A standard geiger tube will have a fixed capture aperture (meaning – X% of the decays are counted) and is largely invariant to minor variations in tube voltage. (Capture ratio is 10% for 6993 tubes, the pancake version shown will have a different value.)

    Clicks per minute won’t change and will have a fixed accuracy basically forever. If you integrate over an accurate timebase, then the output will be accurate.

    Since the expected usage is monitoring fallout from a far away source, 1/d effects will be minimal (ie – the difference between 20 miles away and 21 miles away is small).

    Just count clicks per minute and compare against background – it will be accurate.

  2. For a more accurate reading, run a 2pf capacitor from the center lead of the GM tube. Wire the cap directly into a 74HC04 and also have a 1 meg resistor to ground.

    This will result in a very clean 20 us 4 volt pulse which can be used by a micro for logging.

    The speaker outputs long clicks – long enough that two counts which come close together can be masked by the length of the pulse.

    Also, the speaker pulse on my units had a lot of ringing, which could potentially cause false counts.

    The 2pf cap should be a ceramic with 2KV rating (these are common).

  3. This build really wouldn’t instill confident in myself. In normal operation the meter reading is what the user will use to determine the safety of current conditions, not the speaker output. Yes it would have been more effort to determine what amount of current corresponded to the meter scale readings. There was no mention if the Arduino’s readout matched the Geiger counter’s readout. Finally when is the last time the counter has been regulated? I would consider all that a moot point if this was for this individual’s personal use. However this the readout is presented for public consumption, and that should demand the highest possible level of quality, IMO. Anyway a nice project,an E for effort. When the survivors in Japan take care of all the victims, and return the new normal life, they should take actions on the lessons learned. Chances are nuclear power will always be used in Japan. The hackers, and other interested parties willing to help should build a network of radiation monitoring based on tested,proven instruments regularly calibrated. Shop built is OK if they meet standards. That should be an effort anyone interest should be taking up, that live near nuclear facilities

  4. I think Doug is making some valid points.

    The builders should get the meter regulated, or at least post the date at which the unit was last regulated.

    Also, measuring the current is probably not worth the effort, so ignore my last post as it’s really a lot of work for little benefit.

    And I agree that accuracy is key. A unit which is inaccurate by 5% or more is completely useless in these circumstances. Even more so because the results will be publicly available – that 5% could mean the difference between a comfortable margin of safety and mass panic.

  5. The meter and the digital CPM correlate quite well. There are small differences but that’s because the VU meter uses an analog integrator and decay circuit, where the digital circuit counts the exact number of events.

    A 5% inaccuracy or even 100% inaccuracy will not cause a panic at the levels the geiger counter is currently reading. They’re far below the known safe dosage levels.

    Also, if people here see an abnormal reading, the first thing they will do is check some other geiger counter on Ustream or Pachube to verify it.

    The last calibration of the geiger counter is unknown. However the readings taken indoors range around 0.11 to 0.15 uSv/Hr. That matches the official gov’t readings of around 0.12 uSv/Hr so its unlikely the counter is very far off.

    There is already a collaborative geiger project going on at SEEED studio. Here’s the link if you’re interested:

  6. Doh!

    Of course when is the last time the counter has been regulated, was to be calibrated.

    Oklan warrior’s, comment come through while I was composing mine.The value’s of component’s that make up electronic circuitry don’t last forever.Practically every instrument that use Electronic circuitry needs be be calibrated, if it’s readings are going to be used to make decisions. A simple or old uncalibrated Geiger counter will indicate a rise in background radiation. During a nuclear event we are interest when radiation levels are nearing or have reached hazardous level. Has been so long since I built a Geiger counter, I can’t the theory of operation. other than that radiation entering the GM tube causes an electron tube to conduct. If indeed the circuit count clicks, and the meter displays clicks per the circuit could be calibrated in the home shop remember I know where I can get at least one old Civil Defense Geiger counter for free. Hopefully, and there is no corrosion the battery(ies) have been remove. I plan to get it, and if it’s still functional, I may decide to spend the money for calibration someday.

  7. For those of you on the west coast of the US my research group at UC Berkeley is monitoring the air and rainwater for radiation from the incident in Japan using much more sensitive instruments (high purity germanium detectors). We are able to measure iodine, cesium, and tellurium most likely from the reactor in japan. The levels are very low and dose to the public is insignificant. We post our latest results roughly once per day.

  8. A CDV-700 Is a marvelous detector used by the civil defense in the 60s. They are quite sensitive and come with a test patch of some kind of radioactive isotope on the side. They use 4 D cells and have a real Geiger tube sealed in stainless that spins open to reveal vents around the tube. Because of the Japan panic in the US, they are selling for around $200-300 on ebay. Do NOT be fooled into getting other CDV models as they are for extreme radiation and are not sensitive enough. The 700 comes with headphones to compliment the analog meter. Normal background radiation is about 14 clicks per minute in our area. (N. CA) I have noticed no change. This unit will detect ionizing smoke detector radiation.

  9. There is a calibration issue: If the tube is receiving too little or too much voltage it’s count rate in clicks-per-minute will not reflect reality.

    As for Ludlum meters, schematics are available on their web-site, and it’s pretty easy to find the actual pulse into the sound generator. The model I have even has a terminal for it.

  10. Seeing as XKCD doesn’t take suggestions. Can someone make a XKCD style comic with Islamic fundamentalists trying to catch/net radioactive particles out of the ocean near Japan with an actual net?

  11. Watching the geiger counter rate against time and compairing to the background level seems a good way to go.

    Measuring the radiation level in a volume of captured rain water is also good idea. I recommend keeping the detector a constant distance from the water sample though.

    It would also be possible to capture samples of airborne dust using a vacumn cleaner then test if their is any increase in radiation from its air filtering components. Be aware that the levels of aireborne dust vary with weather & traffic etc

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.