A Touch Screen Geiger Counter Without A Geiger Tube


We’re assuming [Toumal] was desperately bored one day, because in the depths of the Internet he found some really cool components to build a solid state Geiger counter.

The Arduino and touchscreen are rather standard fare [Toumal] picked up on eBay for about $30. What really sets this project apart from all the other geiger counter builds we’ve seen is the solid state geiger counter [Toumal] used. This device uses a specially-made photodiode made by First Sensor to detect gamma emissions from 5 to 1000 keV.

[Toumal] put all the software for his Arduino touch screen radiation detector up on github. To be honest, we’re really impressed with the rad sensor [Toumal] used for this project, so if you ever decide to pick one of those up, he’s got your back with an Arduino library for it.

33 thoughts on “A Touch Screen Geiger Counter Without A Geiger Tube

    1. Well, yeah. Except I wrote the whole UI and improved the sensor and display library a bit.
      There’s not much you “do” with a radiation sensor output. And this sensor has an accelerometer output so I’d say it’s a bit more involved than a geiger tube shield.

      But yeah, as I said on the page, it’s not gonna set the world on fire.

        1. Alpha is impossible with this detector, neutrons could be detected by “activation” with a suitable medium, but the detection efficiency is horrible (you need a butload of neutrons to detect anything at all) and response time is…long…
          Beta could be done indirectly by letting them produce braking radiation (Bremsstrahlung) with a suitable target (thin, but very dense metal)…

          If you want a universal counter (not spectrum measurement), plastic scintillators are the way to go ;-)

    1. Thanks, I was about to point that out, with much less accuracy.

      Just because the curve goes wider, doesn’t mean the detector is necessarily much good out there. It’s on a log-log scale. Still, if you knew what material you had (and it’s gamma emission spectra), you could scale the count-rate by the insensitivity factor to correct somewhat to the true count rate. “somewhat” because you’ll also overestimate counts due to background, so you should only do it after subtracting the background counts. (you probably don’t want to “build in” such a factor.)

  1. How do you test that it’s working without going into a dangerous environment. Would you need to do something like remove the source from a smoke detector, or are there more benign ways of getting a (non-dangerous) amount of gamma radiation?

    1. Thoriated lantern mantles are widely available and generate a bit of radiation, alpha if I recall correctly. It doesn’t penetrate far. I’ve used one to verify a geiger counter was working after part replacement, but only got a clear count by removing the counter’s outer case and pressing the mantle right up against the detector tube.

      1. Alphas are difficult do detect (impossible with this detector), but the thorium decay chain has plenty of isotopes that emit beta* and gamma radiation which this sensor should detect…

        *probably will not detect betas themselves but rather the braking radiation they produce when they impact the sensor wall ;-)

      2. Verifying it reads “something” doesn’t mean it is working correctly.

        About thirty years ago in California, some idiot walked into an office with an uncalibrated Geiger counter, turned it on and “measured” high radiation levels around some computer monitors. This resulted in a walk-out followed by all sorts of panic and lawsuits. Later, some engineers from the company that made the monitors got their hands on the device and checked it (it had been rented from a rent-all store) – it was not calibrated and gave high readings on anything. After calibrating it, they used it and another unit to re-test the monitors and showed no radiation from them. However, there was a higher background radiation than expected in the room. It turned out the paint on the walls was contaminated!

  2. Looking at the technical documents section at http://www.radiation-watch.org the detectors (not the sensors) seems extremely hackable. To begin with they are open hardware licensed under CC-by. :-)

    They mention among other things that a copper filter will give a more linear response and that the detectors (obviously) will saturate when the cps (counts per second) reaches the sampling rate, which at a sampling rate of 44.1 kHz will happen at a dose rate of 10 mSv/h (according to Wikipedia an *acute* dose of about 5 Sv can be fatal).

    1. maybe its because i never learned japanse, but the documentation on this stuff seems quite unlinked on the page. Do you have an further reference ( I just bought one), and it could save me time to figure things out myself.

    1. Beware I’m not read in on radiation, only the very basics of radiation detectors. If I’m not all wrong, the saturation point should be around 10 mSv/h (when connected to a smartphone sampling at 44.1 kHz).

      Strangely enough I have not seen the sensitivity mentioned on either of the Radiation Watch site nor in the sensor data sheet, but the Dutch Metrology Institute certificate (http://www.radiation-watch.co.uk/VSL_cert_3320408.pdf) for the Type 1 detector (the FRISK mint box diy detector) mentions a sensitivity between 0.07 and 0.09 μSv/h/cpm in their test results.

      1. That’s why I wouldn’t recommend their smartphone models. I mean it’s probably gonna be enough for most applications but using a hardware interrupt like I do in my fork of their library is removing that limit, so you’re left with the actual sensor saturation limit.

      2. It sure is. And 10mSv/h wont be enough to measure fallout after a very large accident (or nuclear bomb, God forbid). We wont get there, but inside a reactor containment the radiation level can get in the order of 10 Sv/h and up after a severe accident.

        After reading in a bit this week on the radiation meters on the market I stumbled upon an Atomic Energy Commission informational film in the SL-1 reactor accident on YouTube. One of the lessons learned was that there was a need for radiation survey meters which could go to ≥5000 R/h (no typo, that’s five *thousand* Röntgen, or about 50 Sv/h), but that is for use inside the confinement and with a burst reactor pressure vessel (to put it in perspective, a dose of about 500 R or 5 Sv is usually considered lethal). :-S

        Side note: A possibly grave approximation is that 1 Sv = 100 rem ≈ 100 Röntgen (I say possibly grave, as the definition of 1 Röntgen has varied quite a lot).

  3. You can use an ordinary PIN diode, I’ve built two so far, and have one outdoors and graphing on an old desktop computer.
    The second one is here on the work bench which is used to check samples. I use a large TO-3 decapped power transistor to detect alphas.

    And yes there is a daily peak at ~ noon, with the lowest readings a bit before dawn. What I’ve noticed is most of the background radiation comes from dust blowing around on dry windy days, and the readings get quite low during and just after rains.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

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