The connections of the tube are: cathode/aluminum foil/copper tube/BNC jacket, and anode/copper fitted connector/soldered to central BNC connector. Everything is nice and tight with no mechanical play at all.
As a source of EMI I used a high voltage arc in air; it's produced between 2 copper electrodes about 1cm apart, fed by a NST nominally 3.2kV/30mA (but it doesn't get anywhere close to that). The NST is specced for 220V primary, so I also use a upconverting transformer 110V-to-220V/500W. Using a kill-a-watt type device the total power used by this chain is 52VA, so allowing for losses in the 2 transformers I will assume that no more than 25W is fed into the arc. That's why I said above "doesn't get anywhere close" - but good enough for this test.
Setup:
- tube + Faraday cage held in a fixed position above bench in a lab clamp, with the other end supported by a metal ruler which is exactly 15cm. (because the lab clamp isn't strong enough). It's connected to the MCA then to acquisition computer.
- arc is set 30cm away from the tube between electrodes as described above, supported by a flower pot for insulation.
- neutron source placed in relevant experiments at predetermined position, 15cm away from the tube. The source I have posted about before, built from a 5mCi 210-Po source + beryllium. Calculated current activity of the source is 3.68mCi.
- moderator built around the detector tube in certain parts of the experiment, consists of a 4x5x20inch block of HDPE and 12 quarts of mineral oil.
Counted neutrons 8 times, 10 minutes each; recorded all data including pulse and baseline waveforms.
Experiment truth table (and quick results for the impatient):
Code: Select all
test EMI(HV arc) source moderator | counts in 10 minutes
N1 0 0 0 | 3
N2 1 0 0 | 3
N3 0 1 0 | 5
N4 1 1 0 | 5
N5 0 0 1 | 4
N6 1 0 1 | 2
N7 0 1 1 | 29
N8 1 1 1 | 27
I think the numbers speak for themselves. I did not notice any abnormality of the baseline when the electric arc was turned on. The SNR remained excellent even in the presence of interference.
I'm attaching the full datasets to this post.
The next post will contain screenshots of the final result for each of the 8 datapoints above. Post #3 will contain pictures of the experimental setup.