Every mad scientist’s lair needs a Geiger counter. After all, if that UFO crashes on the back patio, you might need to know if it is hot. [Tanner_Tech] shows you how to build a cheap one that will get the job done.
You do need a Geiger tube, but a quick search of a popular auction site shows plenty of Russian surplus for a few bucks. The other thing you need is a source of high voltage (about 400V), which is the heart of the circuit using a 555-based DC to DC converter. You can see a video of the device working, below.
The DC to DC converter needs a transformer that [Tanner] swiped out of an alarm clock. A piezo transducer (stolen from a junk microwave) gives you the characteristic click. If you prefer solid state over hollow state, there’s an open source project that uses a PIN diode as a sensor. Or you could add an Arduino and some LEDs.
Much nicer, transformer-less solution (with similar cost and components count) – http://robodoupe.cz/2016/pozor-radiace-9/ (warning – written in czech)
That uses the tube that is 10 times more sensitive too, but they cost 3 or 4 times as much so I’m not so sure about the cost being similar.
I was talking about the high voltage source. With the tube, I agree.
But anyway, I haven’t noticed until now, that this a work of a young hacker. Good job, I wouldn’t make it, when I was 15.
Can someone translate this czech article in English. It seems very valuable and automated translation is barely readable.
For a while about five or six years ago, there was some talk on the pronuclear energy blogs over putting together a modern version of the Gilbert U-238 Atomic Energy Lab –
https://en.wikipedia.org/wiki/Gilbert_U-238_Atomic_Energy_Laboratory
– that had been sold in the Fifties. One of the sticking points was that everyone felt the new kit should have a Geiger counter but finding one at a reasonable cost was impossible. Something like this would fit the bill.
G’damn this kid is like 12.
Nice to see hacking back in the hands of the younger generation.
I think I was 13 when the CK722 dropped from $75 to $12. That’s when I built my Geiger counter.
How do you test or calibrate it?
Use a banana for scale…
It counts…
Potassium has a large natural amount of radioactive isotype..
Bananas have a lot of potassium.
All true, but my in own tests a banana would never lift the CPM above the natural background. Potassium salt has a lot more potassium than a banana, and when you use a pancake tube you will be able to increase your event rate.
Have a look on ebay for “Geiger Counter Check Source – Uranium Glaze Pottery” they sell a celebrated sample.
Calibrated, LOL, but if ti works I guess it could become celebrated.
Aliexpress has a bunch of “massage rods” or pens or whatever (no idea WTH is that, really), those contain thorium and they come with a certificate saying how much radiation they emit.
They might end up being confiscated by the postal service tho.
I have a project very similar to this one. I made a geiger counter out of some wire.. Oh yeah and that geiger tube thing.
I cant help but think a high value resistor should be placed across that piezo transducer to provide a DC path for the GM tube would improve performance? I may be wrong
I do nuclear stuff all day every day. And EE before that. An Am source should have been far more active than he shows here, even though it mostly makes alpha particles that won’t even penetrate the tube wall (some that don’t hit somthing and make a gamma that will get through and so on). Should be hundreds counts/second off the average smoke alarm source (or more). One possible reason…a piezo is electrically a capacitor. He’s probably just hearing when the voltage across it due to a zillion counts makes it arc over internally. It should have a cap to ground.
Another possible reason is geiger tubes, when run at the right voltage, are not proportional counters. It’s not just that radiation ionizes the gas, it’s an avalanche type device, where the initial discharge makes more ions via both the moving electrons and ions and the UV light the initial discharge creates. The wiki link is reasonable here (nothing is perfect unless you have a really long attention span): https://en.wikipedia.org/wiki/Geiger%E2%80%93M%C3%BCller_tube
FWIW, for my battery handhelds, I use a CCFL inverter, with a fast diode and cap, and an LM395 with a pulse stretching cap and resistor to run a headphone driver – it’s more reliable and draws less power (not to mention, simpler and cheaper).
Most CCFLs are quite “stiff” in their output, and to get a precise high output voltage, all you have to do is get the input right (hint, LM 317). At these current levels, it might as well be brick wall regulated.
Of course, if you want fancy (and no, I don’t sell these anymore, Fukushima cleaned out the ebay stock of good Russian NOS pancake detectors): http://www.coultersmithing.com/forums/viewtopic.php?f=52&t=467&start=0
FWIW, a lot of the older (and it seems all the Russian) tubes use neon and an alcohol to quench the discharge (else it’d just count once and be stuck on). These “wear out” as one molecule of alcohol (at least) is destroyed each pulse – they last a few billion max. Newer counters need closer to 800v and are halogen-quenched and last “forever”. Just an FYI.
Good point yeah, smoke alarm source should be at least a low buzz.
Crap on this can’t edit. I meant the piezo should have an R to ground, it’s already a capacitor.
So Danavixen and dcfusor2015 are the same person. Got it.
nope, I spotted the error first :P
Actually, we cross posted, but if you want the win on that, it’s fine with me.
And my original longer post is still “Awaiting moderation” I suppose because I put in a few URLs and that’s not popular around here or something…(can’t blame them, I do computer security too).
Nice project, but these tubes are 188 – 235 cps/R/h and the more expensive ones are 22 – 29 cps/mR/hr, notice the R vs mR value. So am I right in saying the more typical kits are 10 times more sensitive than this design?
You are correct sir. The small interaction volume of these tubes makes them less sensitive than the larger SBM-20 / 20U tubes used in the kits.
That is handy to know, because if you are looking at radiation in terms of what damage it can do, to you, then you need to know what the event count through a given volume of space is?
I’m impressed. Of course kids being kids – you most likely will be seeing this power supply being hooked up to a pumpkin or a doorknob. Still, 15 years old, I was still buying TTL off of Radio Shack’s wall at 15.
Yes, you want to know the flux per square area. But it’s much more complex than just that. Check out some of the definitions, the modern unit is Sieverts. (Wiki again, if nothing else). Note that things these tubes don’t count at all can still give you quite the nasty dose. (I’ll refrain from more links as they seem to hold the post if you put them in)
Also, it’s worse if fast. You get a lot of stuff during a life and it’s not very important. Get a lifetime’s worth in a couple seconds, you’re probably dead. XKCD has a well researched chart on all this.
Rule of thumb: Alpha radiation (what most naturally radioactive decay mostly makes) – harmless on the outside, your skin stops it in the dead outer layer. Inside you – no dead skin layer to stop it – extremely bad news. Hold it in your hand, fine, just don’t inhale or eat the substance. Look up the 4n, 4n+1 etc natural decay chains. Worse yet of course is either pure U or Pu, or many nuclear waste isotopes (strontium and cesium) which your body tends to sequester right in your bones, kind of worst case.
Beta rays (electrons) a little worse – they can get through skin, but usually a couple mm of aluminum stops most of them.
Most nuclear waste, being neutron-heavy, decays via betas (a neutron becomes an electron and proton, E-MC^2 favors this).
Gammas…complex absorption pattern (see some reference on biological and radiation). There’s a reason they use the energy they use in hospitals and dentists offices, vs lower or higher energy.
Neutrons – along with a lot of gammas, these geiger tubes simply don’t count them at all. The tubes will count some of the gammas, as they knock electrons off the metals in the tubes…but not all, and not all energies. The calibration is complex.
To put it in perspective…you get hit by a lot of much more energetic cosmic ray bursts/second than most other background. A reasonably sensitive NaI:Tl counter here shows me 100-500 per second (varies with time of day) on about 16 sq in scintillator, since it can time-resolve most of the stuff in a burst if the original cosmic ray hit air (or my solar panels) well above me. So freaking out about a few counts from the “hot rocks” where I live in SWVA (it’s actually quieter inside my house and lab) is pointless. That’s of course why they go down in mines for stuff where a big background count messes up what they are looking for.
Forget bananas. Even pure potassium nitrate counts “at all” but isn’t loud compared to a smoke detector source, an anti static brush, or a thoriated lamp mantle (or, more available, a thoriated TIG welding rod). We use new old stock lamp mantles or a box of welding rods as secondary standards to calibrate our “just counters” vs other things we use to calibrate a gamma spectrometer.
I’m guessing the source is the real key – and your purpose for building needs to be explored. If you are trying to detect a leak from the local Nuclear Plant, I’m guessing that it would be specific to the fuel they use. If you are trying to detect a bomb going off in the atmosphere, I would guess that broad spectrum radiation would occur from that. The first likely might kill you if you could actually detect it, the second would be more handy for detecting particles in food and water, or you would die from the initial particle emissions. Maybe slowly, maybe quickly, depending on the dose.
Nice one! If you are looking for a Cheap Simple DIY Geiger Counter that also ticks some current buzzwords (IoT ESP8266 MQTT) with even lower component count have a look at this project: https://hackaday.io/project/12933-esp8266-geiger-counter. (sorry for shamelessly plugging my stuff.)
geiger tube don’t last forever, and all the Russian tubes are more than 30 years old
That’s a pity indeed. But hey, there seems to be something brewing again between those two, so maybe we’ll be getting new SBM20s again too :)
So nice !
How did he know he needed 4.65Hz (I think that’s right) for driving the MOSFET? And how does the transformer work just off of switched DC (No negative component to the signal is neccessary)?
It’s all okay as long as you don’t tell the transformer it’s working off switched DC… otherwise it will pull out it’s union contract.
Other transformers to be careful not to tell they’re working off switched DC, your cars ignition coil(s), your laptops screen inverter, your motherboard voltage convertors… … …