Hi everyone,
I might get a chance in the following months, to collect a rather special spectrum, but the count-rate will be some kind of high, i will except around 100-300uSv/h for the spectrum collection, which will be only a few minutes possible (<5mins). For this i would have to build a portable spectrometer with a tablet and a sound card and a standard scintillation detector. My 34B57 does give around 50cps at a room background, which possibly is around 0.1uSv/h so i would expect a rough count rate of 150k cps at 300uSv/h. Quite high! Also there will be neutron radiation present.
Given that soundcard-spectrometry circuits usually have pulse lengths of 100us, the maximum count rate would be 10k per seconds, theoretically. This is a huge problem. Maybe it would help to assemble a specific detector, that has a rather small crystal to get a lower count rate. The problem with small crystals will be, that the usually miss high energies more, since the stopping power is lower. I successfully collected spectra with theremino and PRA to around 3k cps
I currently also was thinking about writing a small program on an FPGA, to just sample the original signal, which is much faster (shorter). But that would need something to process later....
Has anyone here made contact with "high rate" spectrometer? Are there any other issues i have not yet noticed? Are there any other tipps on this?
high count rate spectrometry
-
Sesselmann
- Posts: 1398
- Joined: 27 Apr 2015, 11:40
Re: high count rate spectrometry
Jonathan,
High count rates are the enemy of sound card spectrometry, it is much better suited to measuring contamination in mushrooms :)
Your estimates are too high, as you have not allowed for deadtime, with pulses of 100 µs your counting plateu is more likely to be 3000 cps. Generally the minimum deadtime has to be twice the pulse length (preferably a bit more).
You should definitely not rely on the counts from a scintillation detector as a personal safety device, because of the low plateau actual exposure might be a lot higher.
Now with some modifications you could push it a little higher, replace the coupling cap with a smaller value, maybe 100 pF instead of the 1000 pF supplied as standard, this will make the pulse taller, then adjust the pulse length on the trimpot to make it decay faster. You then need to reduce the length of the sample string from 60 to 20 or less (10 either side of 30)
Since the input sample rate is only 48 kHz these changes would compromise resolution, but it should also reduce the dead time and allow 2-3 times more counts.
On the subject of fast counting, I met some guys at the IEEE conferance last time they were in Sydney, they were developing special purpose boards for super fast pulse capture, absolutely insane Gigasamples/s rates. https://www.naluscientific.com/technology/ obviously at that rate you couldn't capture a very long recording.
Steven
High count rates are the enemy of sound card spectrometry, it is much better suited to measuring contamination in mushrooms :)
Your estimates are too high, as you have not allowed for deadtime, with pulses of 100 µs your counting plateu is more likely to be 3000 cps. Generally the minimum deadtime has to be twice the pulse length (preferably a bit more).
You should definitely not rely on the counts from a scintillation detector as a personal safety device, because of the low plateau actual exposure might be a lot higher.
Now with some modifications you could push it a little higher, replace the coupling cap with a smaller value, maybe 100 pF instead of the 1000 pF supplied as standard, this will make the pulse taller, then adjust the pulse length on the trimpot to make it decay faster. You then need to reduce the length of the sample string from 60 to 20 or less (10 either side of 30)
Since the input sample rate is only 48 kHz these changes would compromise resolution, but it should also reduce the dead time and allow 2-3 times more counts.
On the subject of fast counting, I met some guys at the IEEE conferance last time they were in Sydney, they were developing special purpose boards for super fast pulse capture, absolutely insane Gigasamples/s rates. https://www.naluscientific.com/technology/ obviously at that rate you couldn't capture a very long recording.
Steven
