(as usual, click on any picture to enlarge it for better viewing)
First, our crew:
- Yes, we're smiling - we just did great and set another record!
Who were brave enough to help with the pictures and videos while not in the "safe zone" behind the maximum shielding. Thanks for taking one for the team, fellas!
This is the basic setup. During this short vid, the fusor is going on and off on its own - I have just the threshold gas pressure for it to "light off" without an ion source. We are using the hornyak from Eljen as our data-aq sensor, which as calibrated using Richard Hull's fusor at HEAS 2009, just happens to work out as about 980 cpm/million neuts/second (as agreed by all the others present there with their own calibrated detectors). The clicking/rushing sound heard in this video when the thing is "going" are two other neutron detectors. One (left channel) is a 3He tube about 10" from the shell wall, the other channel is a WWII vintage B10 tube by GE about 30" from the action.
http://youtu.be/rtmccSMwlt8
We ran a bit to heat things up...then pumped back down and started fresh for a longer run with more-pure D. My new Boron Nitride HV feedthrough is still outgassing a bit, but it's working fantastically well, and looking like it is going to last "forever" - so that issue is finally solved once and for all.
Here is the data aq from that first run.
- Early run to bake things out
On the left, the green is the neutron counter, the red is a pancake geiger right in front of the fusor (but behind as much lead as we could manage - that's all scatter X rays during the run). You can tell when I turned off the fusor as the neutrons drop to zero, then there's a short pause while I fetch the silver out of the activation oven and place it over the geiger tube to count the activation. This being a log plot, you can extrapolate back to when the fusor was shut off - so the variable time it takes me to do that job doesn't matter with this kind of data aq system (designed and built here, with help from Joe Jarski). The right hand plot is a few a/d inputs we took during this run, time-synced. The line at 2.5v is a precision voltage reference I use later in data mining to calibrate out any error in a/d full scale range (changes with the computer power supply). The blue line is log pressure - you can see the drop after the run when I pump back down, but without some math, you can't really see the pressure variations during the run (I do that with other software that knows how to handle this raw data and change it into real units). The red line is the HV - topping out at 50kv. The green is the current - about 22 ma max except right at the end when I cranked up the limit a good bit (this plot is linear) - and notice it didn't really help the neutron output - by that time, there's water, H, 3He, and T in the gas in our batch mode gas handling system. Essentially, we've also lost all the D from the tank walls, where some fusion occurs (maybe half when the walls are loaded, and at present, no special D holding material is there - it's just stainless steel). We get about double the net fusion when the walls have D on them, and more in earlier experiments when the walls had Pd and Ti on them. On this first run, you can see considerable instability during the startup - in all the parameters, mostly due to the thing going on and off at the gas threshold pressure, all by itself. You can see the correlation between the neutron counts and the current input (green on both plots). Did pretty well for a warmup run, silver going to something like 1.5k cpm activation (2" sq sample over 2" geiger window).
On John's suggestion, while we let things cool off, we put in about 9e-2 mbar of D to let the ss tank walls absorb fresh D while cooling, then made our "money" run. Actually, we did this twice or so, since we were having some fun doing it.
You can see the pumpdown and then pressure increase for one of those in this shot:
- Crummy run followed by background, pumpdown, then adding extra D. Note pressure plot is a log of the pressure (PKR-251 sensor comes out that way)
Now for the ideal "money run" where we hit I believe almost 7 m n/s.
- The money run. Not shabby for me. 50kv tops.
(we see neutrons at about 16-17kv, well out of the noise, for reference)
You can see where we hit our peak output - just as we hit current limit at about 220-something seconds, before we'd baked the D back out of the tank walls. All downhill from there, but still darn good. I am going to wind some cooling water coil around the tank at this point, so we can keep that D, and coat the wall with Ti again to hold more than stainless steel will - it's obvious we've created some sort of tandem accelerator with our fusors, and much of the fusion takes place at the tank walls if you set up things so it can happen there, or obvious to me - this isn't the first time we've seen this. We also took video during this run, but it's taking its time to upload in HD, so I will add that link when it is finished.
Great runs, great fun to have John here - another truly talented hands-on kinda physicist - to advise and teach me things. Thanks John Futter! Hope you had as much fun as I did exchanging our ideas and some culture (along with some decent beer during our theory sessions).
The other great news from my POV is that it appears that the HV feed-through problems are finally solved for good and all with that tight fitting BN insulator inside thick-wall pyrex, but extending 1.5" past the glass. No more sparking, no more longitudinal arcs around a too-small piece of BN (started small - that stuff ain't cheap!). A little grid material sputters onto the end of the BN, and vice versa - sandpaper after a few hours running, and it's all new again, and it only takes about one wipe of 320 grit to get all pristine again. This REALLY beats replacing cracked pieces of glass/quartz in the FT, believe me.
Here's a video Bill took of the last run.
http://www.youtube.com/watch?v=WdacMDo8-OU
