viewtopic.php?f=7&t=542&p=4407&hilit=fine+wine#p4407is what we wound up with. This expects a floating power supply for the preamp - 9v battery.
The board had a coupling cap for AC coupled input, but that tiny thing wouldn't have stood off the voltage required for a neutron tube.
FWIW, I tend to run those tubes at a somewhat higher voltage than they normally are run at to get a bigger signal out of the tube,
bordering on getting geiger action - not quite, but I do get some "gas gain", which is, BTW, easier with He3 than BF3, which I never got really happy with.
I have an old B10 tube I love - it's fairly numb but very reliable and with a big signal. With a fusor, once you get a little past go, it turns out
you don't need super sensitivity, even a low quantum efficiency will still count a copious neutron source. What you DO want is rejection of
all things that aren't neutrons - so you will only see a little background from cosmic rays, but nothing from the average hot test source like a thorium lamp mantle or a smoke detector.
My main "I really trust this" detector around here these days is a hornyak button (from Eljen) on a phototube, using this preamp. It's not very sensitive, but it does discriminate well, can't be flooded and stop counting if there are too many neutrons, is very stable over time, and is time-accurate, unlike nearly all other methods.
Anything that requires slowing down the neutrons to detect them introduces significant "time smear" as to when they came out vs when you got the pulse.
This is important to me as I'm doing non DC fusor drive and want to understand the dynamics better. For DC drive it probably doesn't matter much.