Are supposed to be hard to deal with, but tonight I kludged up the simplest imaginable circuit and the two I have worked great with it, using the fusor as a test source for neutrons, and a very hot torbermite rock as a source of gammas to make sure I was only counting neutrons.
Here's the circuit I used:
This opamp "shouldn't be fast enough" but it works fine. I get 1-5v pulses on neutrons, roughly 50 uS wide, and about 250mv on hot gamma rocks. I used 10 meg R's in the opamp feedback and positive input for balance. This could easily drive an NPN transistor for a digital count output (use a 10k or so R in series with the base to limit base current). I used my standby CCFL inverter for the HV, about 680v or so, and for that, the nominal 12v input inverter only needs about 3.5 volts. I'm going to fool around with this a little more and see if I can get it going with a 5v opamp next, so maybe I can even make a portable version -- that CCFL can't be drawing too much at that voltage input level, they are fairly efficient, so maybe I can go battery operated. I tried with both short and long BNC cables to the tube -- it just works. The unlabeled R's on the opamp inputs are 10 megs. The only issue I had (built this on a push-in kludge board) was noise from the CCFL, which radiates a fierce amount of 43khz -- but a copper shield fixed that right up -- and is something you usually need for those. I used my old standby, Digi part number
289-1025-ND, which is a JKL BXA-12529 type. I find these are a ton better than the TDK made ones for quality and efficiency. They run nicely all the way down to about 1 Vbe at reduced input/output, and are quite stiff -- you can regulate the output well enough by regulating the input. For now, I'll make a fixed preamp/power box and feed the counts into our multigeiger PIC project for logging during runs.
As you can see, I'd not miss many real neutrons running with a more or less fixed .6 volt theshold (eg a transitor BE junction) as my "comparator". I guess a nit picker would say it would be temperature dependent, but I'd reply that if you think any of these count neutrons accurately enough for that to matter -- I want some of what you're smoking. I see about 300mv max out of this on the hot rock I use for testing; some 20k cpm or more on a geiger from a good distance -- torbermite ore with lots of daughters. That one I keep in a baggie and a pig outdoors when not in use, and wash after touching the baggie (radon daughters build up quick in the pig).
That was not a particularly hot fusor run, and the moderator/tube were about a foot away from the action. So don't despair if you can't find a "better" 3He tube -- you're good to go for fusors with this ancient tech just fine, thanks. In fact, I have to "numb" my 3He tube to use it at all close to the fusor, setting the theshold at about twice what a single neutron hit inside a "tube time" gives, so I only see multiple hits within say 10us of one another -- else it counts at over 10kcps and I can't use that on audio to tell how I'm doing!
I use a 6.5" diameter piece of UMHW HDPE for the moderator on these kinds of things. It's an absolute bear to machine/drill etc, but the best there is otherwise.
I had to use a chainsaw to cut the 3' rod to length, and it took a couple hours with a drill and extension to drill out an 18" by 1.5" hole in that stuff. I now have one very strong arm from running the drill press table up and down -- the only way to cut this is to melt it, and you have to pull the drill real often and clean it....about once per half inch.
A hand saw just skips over this stuff, and my metal bandsaw melts it and grabs. It nearly ruined my chain saw! So much for a precise cut. I could have lathe cut it had I been willing to change chucks, maybe.
With some adjustments, this would probably work well with BF3 tubes too. I'll try it soon.