So, been thinking off and on about the leak in Japan like everyone else. One thing they really are lacking over there is a simple, cheap civil-defense level radiation meter.. or more specifically, a radiation ALARM. What's needed is a tool which can be pointed at a plate of food, a bottle of water- and within a minute or so, alerts the user if there's detected radiation. Contamination detection, particularly for foodstuffs, is the key when dealing with nuclear accidents and fallout.
I've been mulling over several different possible configurations, pulling out a few obscure Physics textbooks from my college days and making a ton of online queries that undoubtedly put me on a few "no fly" lists and might warrant a courtesy call from a UN compliance team.. and I think we might be able to come up with something- but a bit of input from a good EE is needed.
Geiger-Meuller tubes are expensive and fragile.. though extremely accurate and sensitive. I want to avoid a GM tube, and I think we can reasonably.
My first thoughts were "photosensors" simply blocked from actual light... CdTe and/or Silicon photocells. Energetic radiation knocks electrons around in the Junction areas, produces a current, amplify and presto. I'd think any PN device has to to be at least SOMEWHAT sensitive to ionization. This tends to bear out, I guess- but it seems that due to many reasons, only a very small select group of Photodiodes actually works as a sensor, and they are very expensive and need cryogenic temps ideally. My idea is correct, it's just that junction noise drowns out the "data".
Then I remembered the Electroscope. If charged, ionizing radiation causes an electroscope to discharge at a faster rate... electron transport by ions created by the radiation interacting with the air. Old smoke detectors worked like this; a small amount of radioactive material is placed near charged plates, if air "ionization" increases, it's not because the radiation increased, it's because there's ionizable particles (smoke) in the air. An increase in ionization is then detected and sounds the alarm.
I'll come back to this shortly.. I intend to see if I can obtain an older model detector and do a teardown. If a smoke sensor is a ionization sensor, then the design below becomes even easier. If I can find a functional one, I may just see if I can set it off using a Thorium-doped camping lantern mantle. Until then, my design question boils down to this:
I think we can build an ION CHAMBER detector ourselves: Take a metal can and cover the open end with several layers of screen (allowing alpha.. otherwise thin metal, probably foil, would work- but would start detecting at Beta levels only probably. Alpha s stopped even by paper) or thin paper. Through an insulator, extend a conductor (a wire, not touching the can, or even a screen or other large-surface-area conductor) into the can.
Now charge can relative to ground... and connect the base of a sensitive transistor to the conductor. Amplify.
Ionizing radiation should ionize air within the can, like any other air. The key is that we charged the can and have a nearby open-air conductor leading to an amplifier. The Ions are going to be drawn in/out (polarity) by the charge field.. producing an amplifiable signal.
Think of it as a vacuum tube without the vacuum or the filament.
Accurate? probably not. Sensitive? not overly. Able to detect the reasonable levels we need to be afraid of in a spill scenario? I think it might be possible... and one HECK of a lot cheaper than a GM tube.. which essentially is nothing more than a beefed up version of this.
We then need to count the decay voltage pulses over time to determine radiation levels. This is a function of sensor sensitivity.. more sensitive, the shorter the sample time required to get meaningful data. Luckily for us, counting spikes and that sort of thing is exactly what a hardware interrupt is for....
Thoughts?