| Author |
 Topic  |
|
|
ullix
Germany
1171 Posts |
 Posted - 07/01/2017 : 09:57:34
|
The calibration is meant to establish a relationship between the count rate in CPM and the dose rate in µSv/h. The GQ GMC counters have 3 calibration points, which would allow to accommodate some non-linearity in the relationship to take care of count rate saturation effects. However, currently all 3 points establish the same slope, hence effectively only a single calibration point is used:
Device Calibration in GMC300E+:
Calibration Point 1: 60 CPM = 0.39 µSv/h (0.0065 µSv/h / CPM)
Calibration Point 2: 240 CPM = 1.56 µSv/h (0.0065 µSv/h / CPM)
Calibration Point 3: 1000 CPM = 6.50 µSv/h (0.0065 µSv/h / CPM)
GeigerLog uses the calibration of 0.0065 µSv/h / CPM.
Unfortunately, there is no official statement about what this actually means. For what situations is it applicable? What type of radioactivity? What beta, gamma energies? What count rates?
I was unable to find specifications for the Geiger counter tube M4011, currently used in GQ counters. However, I established that the SBM20 tube, an old Russian Geiger tube, is similar to the M4011 at least in some aspects, and can even be used instead of the M4011 in the GMC300E+ counter, see http://www.gqelectronicsllc.com/forum/topic.asp?TOPIC_ID=4571
And for the SBM20 one does find specifications, like here: http://www.gstube.com/data/2398/
where it says:
Gamma Sensitivity Ra226 (cps/mR/hr) : 29
Gamma Sensitivity Co60 (cps/mR/hr) : 22
Co60 is a beta and gamma emitter; Ra226 (plus its decay chain) is an alpha, beta and gamma emitter. However, both are typically packaged such that only gamma can escape the package, and so we now assume pure gamma emission. With that we can equate mR with mRem, and with 1 mRem = 10µSv, we get:
Ra226: 29 * 60 / 10 = 174 CPM / (µSv/h); invers: #8594; 0.0058 (µSv/h) / CPM
Co60: 22 * 60 / 10 = 132 CPM / (µSv/h); invers: #8594; 0.0076 (µSv/h) / CPM
Average of the two: 0.0067 (µSv/h) / CPM
GQ’s calibration: 0.0065 (µSv/h) / CPM
GQ’s calibration is close enough to the average of the two, and with nothing better at hand we’d say that this is the same, and that this is the base for GQ's calibration factor.
Looking at the gamma spectra attached we see that Co60 is above 1 MeV, while Ra226 is mostly below 0.5 MeV. At least the SBM20 tube, according to the specs, is 32% more sensitive to the lower energy gammas. Perhaps because the higher energy gammas of Co60 have a lower absorption and hence a better chance to pass through the tube without generating a count.
The consequence is that the calibration is ONLY applicable for gamma radiation (and only approximately given the energy dependence), but NOT for beta radiation, for which both tubes are also sensitive!
We simply do not know what the calibration factor is for beta!
Since the case of the counter is basically transparent to gammas, it does not matter to the calibration whether we make the backplate of the counter more permeable by drilling holes, or taking the backplate off completely – especially considering the hand waving we have applied to come up with the gamma calibration.
And when we take it off and get significantly higher count rates with beta emitters, it also does not matter because the calibration, when applied to beta, is wrong in the first place!
Can anyone come up with a defined beta source which allows calibration?
Image Insert:
20993 bytes
Image Insert:
40538 bytes |
|
| Reply #1
ChrisLX200
United Kingdom
35 Posts |
Posted - 07/01/2017 : 15:09:56
|
I think your problem here would be that you can only have one set calibration data loaded at any one time, and we don't know how those three data points are used by the firmware to correct the count (or do we?). I don't know enough about how these devices work but 3 points don't really define a non-linear curve very well. Trying to correct for beta will also mess up the gamma estimate. Realistically I guess you can just use a thin sheet of aluminium (better - a shaped aluminium case) to confirm proportion of beta contribution to the total count and leave it at that. Personally I'm coming round to thinking of this device as a simple alert monitor rather than a quantitative measuring device (for anything other than gamma) but maybe that's being a bit harsh...
|
GMC-300E-plus |
Edited by - ChrisLX200 on 07/01/2017 15:11:26 |
 |
|
| Reply #2
ullix
Germany
1171 Posts |
Posted - 07/02/2017 : 01:52:11
|
Yes, I think you are a bit harsh on this nifty little device ;-) With all the critique I have too, this is actually a capable and robust system, at an attractive price, allowing to do things which you can't do with other, more expensive systems. What it lacks is specification; GQ is not exactly forthcoming with that. The 300E+ is actually the best value, in my view.
The calibration issue is a principle one, which you face with every Geiger counter, irrespective of make and price. It is the physics, which determines how the tube responds, and for different scenarios you may use printed correction tables and/or graphs to translate from count to dose. If you always have the same scenario you could fix that calibration into the firmware.
I could easily put a multitude of complex calibrations into my GeigerLog program. But at present I don't know what they should be. Therefore all I do is using GQ's calibration.
Blocking all beta is possible, but needs several mm of aluminum (3 ...5 mm, depending on source). But due to Bremsstrahlung a sandwich of several mm plastic followed by aluminum is better. (see my Potty Training) However, this leaves you with 10...30 fold reduced count rates for the typical sources used in this hobby environment, which are mostly weak gamma emitters (K, Th, U, Ra, and their decay chains).
The calibration with 3 points is actually addressing a problem different from the translation of counts to dose! The reason for multiple points is the saturation effect at high doses. After a count has registered, the tube has a certain dead time (typically within 50 ... 200µs) before another count can be registered. Beyond (my guess) CPS=2000 (CPM=120000) a new count will occur during the dead time of the previous count and therefore not being registered. This can be corrected to a degree. Correction is a better word than calibration for this case; the "3 point calibration" wording is confounding the issue. See the picture below from here: h**p://www.imagesco.com/geiger/spec.html
If this picture were valid for the GQ counters too, you would put "correction" points at CPS=1000, 3000, and 9000, and get a good representation of the curve. These "correction" points are wayyyyyy higher than GQ's CPM=60, 240 and 1000, equal to CPS=1, 4, and 17!
Since GQ is using a linear "calibration" it is ignoring this saturation issue. I would argue this is justified as hopefully nobody using the GQ counters is working with sources delivering a years worth (20mSv; occupational safety limit in Germany) of dose in a single day!
Image Insert:
12241 bytes |
|
|
| |
Topic |
|
|
|
Calibration of the Geiger Counter
