This depends on the kind of radiation you are going to measure. If you dont know the exact composition of your radiation field, there is no chance. If you will measure a known field eg only Tc gammas, you will get calibfactors from the manufacturer.
You would have to know what you measure: for a known radionuclide you can work out the expected dose rate at a particular point using the gamma of that nuclide (gamma*activity/d^2) and then get a calibration factor for your counter for that nuclide. Low photon energies will possibly give a higher than expected reading because of an increase in photoelectric interactions in the housing of the GM tube. If you don't know what radiation field / energies / nuclide you are measuring you will be stuck I think...
Prof Mushtaq Ahmad has provided you with the most accurate way to convert count rate into dose.
Please add to the information provided in the detector-specific monogram the effects of shielding, distance to source and any geometric conditions specific to your measurement.
18 cps/mR/h is the manufacture specified value for the mentioned GM tube. practically you can measure by exposing to a known radioactive source at different distances.
Thank you for the feedback and answers. The strange thing is (at least to me) the same model of detector produce different set of counts at the same background. For example, detector 1 produce an average of 100 cpm while detector 2 (same type and model lnd7121) produce average of 400 cpm.. Can anyone help me to explain why this happen?
The problem maybe complex. As Said Sabbagh wrote, it is importend to use exact the same geometry. A other possibility is, one of your gm-tube is defect.
As the previous advices said, the direct conversion of counts is COMPLETELY dependent upon the type of radiation that you are dealing with. Once THAT is determined, one can either develop a calibration/correlation curve (such as that suggested in Prof Mushtaq Ahmad's advice), or simply use manufacturer's specs for KNOW radiation types. Your question about different responses from 2 tubes of the same type suggests to me that there is SOME difference or manufacturing defect.
A short answer to differing count rates. Either you see dead time effects possibly caused by an electronic change, or your detectors are not really identical, possible cause for different count rates can be entrance window thickness. Or third reason could be a low power supply in the low dose rate detector. My central question is, don´t you use calibration sources to check the operability of your system?
From the graph on page 3, the sensitivity is about 16 counts per second per mR/hr, which converts to 960 counts per minute per mR/hr or 57,600 counts per hour per mR/hr. For other GM tubes, you would need to get the calibration information from the manufacturer, or calibrate it yourself using a known source.
I read Dennis attend and found the Cs-137 calibration of the manufacturer in the cited graph. The entrance window is 0.254 mm steel. If you change your radiation quality, you will get a lot of attenuation changes and deviations in efficiency.
As most of the experts have opined, the conversion of counts to dose is purely dependent on the quality of radiation. You need to use the appropriate conversion graph (pertaining to your GM tube) provided by the manufacturer. Otherwise you need to develop . Regarding the different responses by two similar GM tubes under the same conditions, then ONE MUST BE DEFECTIVE.
Refer manufacturer design manual for conversion factor .As a thumb rule for Sr-90/Y-90 source considering Bmax 2.27 Mev 1Bq/cm2 correspond to dose rate around 0.2 mR/HR.