cicastol wrote: ↑
19 May 2019, 06:09
yes,unfortunatly now is a little bit too late to see the lower 134Cs peak with a NaI(Tl) scintillator, your's current estimated resolution of 6,1% is a little optimistic, i noted that is common a compression of the spectra that tend to show higher than real resolution, you need to do energy calibration across multiple peak,well suited is Th232 for this purpose.
Right now I am in no position to tell if the resolution I am reading is optimistic so I can't really answer that, I will leave it to Steven, or Luuk.
I have a Th232 source (two lantern mantles, a classic) and yes, they are great for multiple points calibration. I would say for this particular spectrum a calibration with a Cs137 check source is still pretty good, since basically that's what you are mainly detecting in the sample and peaks looks right on the money, off by no more than 1-1.5 keV.
In my experience Cs137-based calibration starts to go off at higher energy for obvious reasons, and in my case it's typically 15-25 keV off at 1460 keV (K40).
Anyway, I repeated the test on the soil sample. I more than doubled the duration for both the background measurement and the soil measurement.
The soil measuremenet alone lasted 18 hours, by far the longest I did so far. This is still unshielded with background subtraction.
I think that's a good improvement on the last one as far as "smoothness" is concerned.
Also the very weak K40 peak I got the first time around disappeared, so maybe that was coming from the background (or maybe it was just noise incidentally located in K40 region?). This time you can spot something, even weaker, centered a little bit south than 1400 keV. It's a bit more evident in the energy per bin view, but still so weak I didn't call it this time, even because I wouldn't know how to identify that.
As expected Cs134 peak at 604 keV again didn't show up. In the early stages of each measurement I saw something there (although PRA showed no gaussian correlation at all), but then the slope of Cs137 buried it completely. In the log scale view there's maybe a hint that something is there (actually not a peak, only a change of slope) but this is probably me wanting to see something where there isn't. So I reminded myself the words of Richard Feynman "the first principle is that you must not fool yourself and you are the easiest person to fool" and finally I left it alone.
I post both counts per bin and energy per bin spectra.
I also post both spectra without background subtraction, so the magnitude of Cs137 and Cs134 peaks at a few mm distance are comparable with the usual background peaks. You can see something around 600 keV there, but it's Bi214 peak at 609 keV+ Tl208 at 583 at keV, both from the background.
In order to know what you are looking at just read the description in each image.
As soon as I can I will also post some quantitative data.