Well, I was testing my new pi camera setup - looking good BTW, and while I was doing that, replicated something (maybe - haven't proved it wasn't emi, but emi is a little doubtful here) I'd seen only once before, in May, which was ridiculous neutron output from the ion source grid alone. This is a very crappy grid, around 1" diameter, two loops of tantalum wire, used out in the main tank for when we want to run at pressures lower than the "main" grid will "light off" and stay lit. Due to PxD in Paschen's law, the little guy out in the big space can "fire" at lower gas pressures and lower voltages. While I was getting setup, I had it cranked to its max ~40kv and it was "lighting off" intermittently. So you can see what I was doing, here's a picture of the setup in the bake-out heater lights. Yes, the one the camera can see kind of blinds it.
FWIW, the neutron detector we log is a hornyak, totally shielded with internal preamp (coax to counter at 5v levels) positioned about 1" in from the end of the sidearm at the lower right, eg, much closer to the main grid than the ion grid. Assuming inverse square law, it should count lower for a given neutron production in the ion grid, since it's a lot further away...but we see what we see (also in X rays, from a separate sensor, behind lead and lead glass, with loud output (have to attenuate the 40v pulse at the counter/logger) and it correlates with the neutron counter (capture gammas?). At any rate, the only thing with power applied was the ion source. Since our plasmas are net-negative, it shows some main power on that graph, but that's just because the main grid is acting as a faraday probe (else the later Q plots would show infinite Q for this part, since we don't log the ion power at this point - we will later on). So that loud neutron and X ray output that ends around 100 seconds is "all ion grid" if you can believe that - I'm going to have to deliberately replicate it and see some signals on a scope free of EMI, as this is "too good to be true" - but you never know.
Ah, but we're really here to see the pi cam in action. It took a 2.5 diopter lens from reading glasses (we old farts know about this one) to get focus, could use a little more, but I've lost my higher power ones for the moment. At any rate, today's mission was to get a couple decent pix with the pi cam and with the cam in the shield, which I managed to do, so let's look at one. This is earlier in the main-power run, before things heated up in there very much.
After things heated up some - we generally show a somewhat reduced neutron output after some running as the D is driven out of the stainless steel fusor tank walls (that's our guess why, anyway - charge-exchange and beam on target fusion at the tank wall when it has D in it). So, here it is, hot:
As usual, you can click the pix to get a larger version, though I did scale it down from 2500-something to 1280 horizontal rez to save space. At this low light, it doesn't really lose much of anything but wasted internet bits. The sparkles in these are because though I have an extra lens and an extra near-ir filter, the pi camera still gets hit with X rays (inside a lead box to keep them out of the room). Kind of a nice backup if you know what you're looking at/for.
Of course, I am logging things while we do this, as in the screen shot above. Here's the full log plotting vs time. After the run, I put the silver on the geiger (pancake) counter, it's a foil the same size as the sensor. You can see where it took me awhile to think about doing that and get it done - we see some background at my normal levels while I'm making the move.
On the screenshots - this is very raw data, unprocesed as can be. The plot window marked A/D Inputs has blue as pressure (already log compressed by the pk251 sensor), purple is a 2.49v reference diode we calibrate the rest against (later), red is the main HV voltage, green current. I started at 9.8 current limit on the big guy, but went up to about 15 ma later, as you can see from that plot (and the neutron output). While it doesn't show on this plot, there was a small negative voltage picked up by the main grid near the 0-200 second range, or the later Q plots I'm going to show would have blown up mathematically. The initial low volts that you can see on this were because the pressure was far too high for either the current limit or any real neutron output (you can see the main HV was pretty low before I got the deuterium pressure down).
And now for some data mining...since I set a limit on this board of N pictures/post, I'll do that as a reply to this one. I wrote a perl program to drive gnuplot from our log files, and it can do all kinds of pretty neat things, like use that 2.49v reference we log to really know what the a/d full scale range was - super important since we have to re-exponentiate the pressure sensor output, and it can do math, like compute the Q (or really, anything you want to type in), point by point....nice tool, if you understand its limits. For example, we get neutron counts even when nothing but the detector is on - cosmic rays. Nope, no infinite Q, we know darn well those are not good data....we still show them because we know what we're looking at.