I use a two in NaI crystal to collect gamma specs fairly often. I use a Canberra PMT 2007 base for a preamp, a 2025 spec amplifier, and a 8706 ADC. I gather the data from the ADC with a home brew microcontroller board that I programmed to bin the peaks as they come. I usually run at 512 channels but can go to 16k.
Many of spectra are short, but often I do long counts of low activity samples. I noticed huge shifts and loss of resolution over these long counts because of what I assume is temperature changes. My lab is in my home, and generally the air temperature only changes by a few degrees. My counting cave is a foot from an exterior wall and maybe that's a problem. I cant really change that but maybe I can insulate.
Has anyone tried to address temperature changes over long counts where it is impractical to tweak parameters? I wonder if I can bring temperature into my microcontroller and make adjustments automatically.
Thanks for any ideas.
Jim K
