These spectra were created at NPL to allow gamma spectrometrists to assess the performance of their own spectrum analysis software on real spectra representing known amounts of radionuclides. The process of preparation of the radioactive sources and the acquisition of the spectra are described in...
Standard Gamma-Ray spectra for the
comparison of Spectral Analysis Software, S.A. Woods, J.D. Hemingway, N.E. Bowles, J.L.
Makepeace,(1997) NPL report CIRM 2, available from them.
This book also carries the results of a study in which a number of commercial spectrum analysis programs were used to analyse the standard spectra. These analyses were performed by the manufacturers themselves and, separately, by the expert appointed by NPL, John Hemingway. John also provided an Appendix in which he discusses the general features and usability of the programs.
These spectra were measured on a Canberra Genie-ESP system were provided in 4k and 8k sizes.
I have transferred the spectra to two ORTEC spectrum formats, .SPC and .CHN. These are available for download here. At this time, there is no other information available for download.
- Set 1 spectra were designed to assess peak detection, nuclide identification, activity measurement, ability to predict escape peaks (?!), ability to measure overlapping peaks, ability to cope with differing peak widths and the estimation of uncertainties.
The detector was an 18% n-type HPGe detector and the sources measured at 550 mm from the detector face. The calibration source was a mixed nuclide source similar to QCYK. 'Sample' sources were A) a mixture of 60Co, 85Sr and 137Cs and B) 125Sb, 137Cs, 134Cs, 154Eu and 155Eu.
- Set 2 spectra have the same aims as Set 1, but with the additional complication of true coincidence summing. Sources were measured on the same detector system but placed on the detector face. Calibration material used for Set 1 apectra was modified added bu adding 54Mn and 65Zn. Sample sources were the same as for Set 1, but at different concentration. These spectra are also accompanied by a background spectrum.
These spectra are to some extent irrelevant, because there are no calibration spectra supplied that will allow TCS to be corrected. Nevertheless, they do allow one to get a feel for the effect of TCS.
- Set 3 spectra were measured is such a way that a high degree of scattering was introduced. The detector in this case was an 11% n-type detector mounted within an un-graded lead enclosure. The sources used for Set 2 were used but measured at 150 mm from the detector face. Again, background, calibration and sample spectra are all provided. These spectra contain a pulser peak to allow random summing to be corrected.
Beware: When the results reported within the report referred to above are looked at in detail, it is apparent that neither the manufacturers not the NPL expert have taken account of random summing. Nor have the other creators of the report noticed, so the discussion of Set 3 results should be taken with pinch of salt.
I am seeking permission from NPL to host the original spectra on this website. In the meantime, they should be available from:
Centre for lonising Radiation Metrology
National Physical Laboratory