Neutron rad safety

[Doug Coulter]

I suspect that JohnF has better detectors available than most do, but yeah, a tandem beam on target can really crank them out. Heck, I think maybe even I do at this point. And that can be a problem as we get more fusor builders here, because it becomes hard to compare our numbers when one guy has exquisite sensitivity and the next one doesn't. I tend to suspect that certain pretty amazing activation/transmutation results on fusor.net are related more to the fact that one guy has cryo-HPGE and everyone else has NaI at best....so the HPGE guy can see things that are in the noise for the others, for example.

For that reason, I've spent some time championing the idea of one that is very cheap, impossible to cheat (as far as I know) and very repeatable from place to place. Might be off a little due to details, but not by integer factors. The standard I'm proposing is activation of silver (or In or Mg, working on those to cross calibrate them now) which is moderately sensitive and fairly easy to put together for low bucks. I'm using some .010" foil I bought from a jewelry supply house at close to the spot price of silver and an HDPE rod 4" diameter with a cut to place the silver in, one end placed against the fusor. We are using a 2" sq chunk, and reading it on a 2" diameter window pancake geiger counter. We use 10 second counts scaled by 6 to get cpm and watch it decay for a few of those. Our data aq records the time between fusor shutdown and start of counting, as the geiger is running right along the whole time, and during fusor on, it counts that. We let it count one background while we move the silver from the "oven" to the counter. If (when) any of our other gear doesn't scale with this, we know it's the other stuff for sure. EMI doesn't activate silver but makes counters count sometimes. Bursting (which is what we are working with as things improve) can overload-blind the counters, but not the silver. And you can reuse the silver almost right away because of the super short half life, where say In you have to wait most of a day to reuse or you're counting some previous activation too. Looks like the progression as things get hotter is Ag, then In, then Au. We've tried the gold, but that stuff ain't cheap and it's pretty numb too. In will probably be the next standard, and we are cross calibrating that with silver on every run to get a good number for the relative sensitivity in our particular oven (the resonances aren't the same).

I guess I should put up a thread with details on this? It's a dirt simple thing to put together, and though going from that to absolute numbers might be tough (haven't done it) to me, that doesn't matter much till we get a lot closer to gain and can boil that cup of tea without counting the input power waste. For us, knowing "better or worse and by about how much" and whether our other gear is acting up is what we need, and what this provides.

A good 5 min timed run here gets us to about 1536 cpm on silver, a lousy one about half that. Richard's numbers from 2009 would say we are at about 1 million neuts/second with the lower number, or if you use his numbers from 2010, about twice that (and yes, I have a problem with that -- he didn't justify the new numbers and reports factor of two change with no backup on that, and no changes in equipment other than measuring gear). He ran over 20 min and got 470 cpm in 2009 (I was there to see it), and claimed that was 500k n/sec. I don't think the runtime matters as much as linear scaling, or it doesn't seem to here due to the short half life of the activated silver. We've had two new BTI's differ by 3x on the same run here! Therefore we don't make claims based on those, despite the temptation to use the "better" numbers they sometimes give, or use those to cross calibrate against. They're just not all that good, certainly expensive, and tend to fail young.

[Joe Jarski]

I like the idea of a simple standardized detector design that everyone could build to compare their results on a level playing field. There are so many variables otherwise that any comparison of different fusor designs is darn near impossible to evaluate.

[Doug Coulter]

Yes, I like the idea too. I'll post a thread on what I'm doing on metrology, and link to it from here when I do. I think I have a decent proposal, but of course y'all chime in if you think of anything you like better.