Best Practices for Small, Low Activity Samples

[Geoff]

I collected rain spectrum a few weeks ago and I found that the primary radioisotope in the rain was lead-212. I found it kind of curious that such a short lived isotope was the primary source of radiation. We've had a volcano nearby that has been erupting pretty constantly since December with frequent eruptions over 10,000 meters. I reached out to the US Geological Survey's Alaska Volcano Observatory to see if they had analysed any of the ash that was produced by the volcano, and if not, would it be possible to get a sample from them. To my surprise, they agreed to give me a small sample of ash from one of the only eruptions to make land fall. Most of the ash has fallen into the ocean. The sample is several grams of unsorted ash from the March 8th eruption.

I've been trying to collect a decent spectrum, but I have limited equipment. I currently have a 25x25 mm NaI detector that has exceptional sensitivity below 1000 keV, and runs at a very low voltage, around 400-450 volts. I have a 25x1 mm CsI detector, the RAP-47. And I have a homemade 20x120 mm NaI detector. I'm using the GS-1100A with the low voltage modification and a dedicated Windows desktop with external sound card. The problem I've been running into is that PRA stops collecting data occasionally and needs to be resumed. This makes a 48 hour spectrum take over 60 hours. That long of a collection time makes collecting a background spectrum pointless, as background radiation conditions vary pretty widely over that time period, especially with the intermittent rains we've been having. When I did manage to collect a 48 hour spectrum and 48 hour background, the net result had a lot of negative data points.

Does anyone have suggestions for improving my setup to allow for such a sample? I've got my voltage set at around 420 volts, which is the lowest voltage at which I collect meaningful data. It's my understanding that when setting up a detector, you want the voltage set as low as possible to avoid noise. Power to the GS1100-A comes from a different computer than the one being used for the spectrum, and both are on an uninterruptible power supply. My shield around the detector is 10 mm of tungsten metal followed by 20 mm of lead. My sample shield is 1 mm of copper followed by 3 mm of tin, then 30 mm of lead. Since the sample is a fine powder, I tried putting it in a plastic zip bag and wrapped the sample around the detector for my first few tries. Currently the sample is just sitting in the sample shield. Any advice would be greatly appreciated.

[cicastol]

Hi Geoff,
imho it's best to power the GS1100 and take spectra from one computer only, 2 power sources add only noise and possible make ground loops.
UPS generally create a lot of high frequency hum/noise and switching harmonics that propagates all around, so if you can don't use it.
I've found the best voltage supply for the detector need to be carefully selected for best balance between gain-noise and resolution.