parametric sweep of fusor parameters

Data from actual runs of fusors goes here, we can discuss it elesewhere in other sub forums I will create as needed -- let me know.
Forum rules
Put data and fusor information from actual runs here. We'd like to know how well you are doing, and how you did it in some detail here. We can discuss elsewhere, this is for real reports from actual experiments only, or at least, mainly.

parametric sweep of fusor parameters

Postby Doug Coulter » Sun Aug 05, 2012 4:54 pm

I have too many computers! To easily get this youtube link, I'm posting it here in this thread before I'm really ready to do the rest of the thread, but what I've done is take a sweep of the fusor vs "you name it" and collected some darned good data that shows me some things we all knew already - and some things most of us never had a clue about. This vid just shows my setup, there is a LOT more to come with real data, 4-d plots and movies of me twirling them around, the whole shebang. This should put to rest a bunch of the dumb newbie questions - and yes, when I'm done I'll also post a teaser over on fusor.net.

http://www.youtube.com/watch?v=bLU9AZZroy0


Now for the good stuff. Here's a vid of the resulting data. Volts, milliamps, and Q (x,y,z) vs pressure (color) are shown in a video that lets me twirl the axes so you can see all the features of the data, concatenated from all those runs (I will make the files into a zip so you can download them yourself and use plotdat, already up here someplace, to do it - or you can just look at the text yourself).
This is the setup for this movie:
plotdat.png
plot parameters for the video showing the results of the para sweep. Quite a lot more is possible and informative, that's why plotdat has presets...


Now for the cool part:
http://www.youtube.com/watch?v=WJe0YBAXwPw
Note, I mis-spoke in the video, but not in reality. Q is computed here as neutonrs/minute over power, not mere current, as you can see clearly in the setup above. Oops.


More to come - Now that I have all this data, there's more than one very interesting way to show it, this is just one of them.

It's pretty obvious that at higher voltages and lower pressures, Q is highest - I even had a couple of rogue points I edited out when the thing went unlit, then lit back up again, which showed Q's (on this arbitrary scale, not the real deal quite) on the order of 100's - that's because at 1 sample/second, I missed the current the power supply was actually putting out - so I edited out a few points that made me look too good to be true, because, hey, they weren't true. You have to know the limitations of your gear!

I've known most of this all along, but got a lotta crap trying to explain it to the beginners. Here it is now, in a form a child could understand. No kidding - more pressure or more current just don't get you an advantage in Q, and usually not in raw neutron output either. I will take another set of runs at lower pressures using an ion source to keep the fusor "lit" at the lower pressure, where I fully expect to see even higher Q's on my arbitrary scale. But the fact that max Q happens at far less than max output - or input is something no one had mentioned (or measured) so far - so this is "news" or should be to most fusor people. With an ion source, I can sweep more of the space - lower voltages and lower pressures at any voltage, to explore whether gas pressure has an effect (of course it does) and in which direction (collisions with neutrals, space charge limits, and all that stuff). Stay tuned!
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: parametric sweep of fusor parameters

Postby Doug Coulter » Sun Aug 05, 2012 6:32 pm

Here's a picture of another plot of the same data, and oh, by the way, the data itself in case you want to find plotdat and play yourself (it's up here under PC software - homebrew).
FusorOutput.png
Raw neutron output in CPM from our hornyak. 1k cpm is about 1 million neuts/second


Sweep.log
The concatenated data from the runs, raw
(217.54 KiB) Downloaded 347 times
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: parametric sweep of fusor parameters

Postby Doug Coulter » Sun Aug 05, 2012 8:03 pm

Here's a vid of that raw plot. Same data, but neutrons not normalized. Peak output is around 4 million neutrons/second here.
http://www.youtube.com/watch?v=8vpCt5kqzYo


This shows that yes, output does go up with power input, especially at higher voltages. But as the previous vid shows - not as fast at the input power. So, tuning for max smoke isn't tuning for best Q. No big surprise there for anyone who has built drag racing engines.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: parametric sweep of fusor parameters

Postby Doug Coulter » Mon Aug 06, 2012 10:51 am

I also took screenshots of the runs in 2-d. I normally look at these plots while running to tune things, so I figured I'd save them, as they show things vs time, which I've not yet done on the 4d plots (though I can, it's just a matter of changing what I map onto some axis, probably color). Here they are so you can see how each run evolved. Note, all the a/d inputs are scaled to volts, not the real units of what is being measured. This is for quality control during the acquisition - the real numbers are easy to derive later, but not if I ran an a/d input offscale!

The keys for these plots - the points/lines are the counters, red is geiger, green is hornyak.
For the a/d plots, the red line is voltage, the green is current, and the blue is pressure (logged so it doesn't look like it moves much). The last line is a voltage reference that allows my software to compensate full scale changes due to the USB 5v supply changing (which changes the a/d full scale range). As usual, click any picture to get a bigger version to view.

25kv20maRun.png
25kv run

30kv20maRun.png
30kv run

35kv20ma3e-2mbar.png
35 kv

40kv20ma2.5-2mbar.png
40kv


There were two 45kv runs, one of which I forgot to get the log file for, but showing both is instructive as to run to run variance anyway.
45kv20ma2.5e-2mbarXLog.png
Not in the 4d dataset, oops

45kv20ma2.2e-2mbar.png
This one made it in

50kv20maNoIon.png
Last run, mostly a little over 50kv, more like 51
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: parametric sweep of fusor parameters

Postby Doug Coulter » Mon Aug 06, 2012 11:25 am

Q versus time. This could be an indicator of how much deuterium is in the Pd coating of my chamber walls. I tried that based on a theory of Richard Hull's. His idea is that you can then get more ions created at the right place in the field. Mine is that we are seeing beam-on-target strikes of fast D's on the tank walls, probably due to charge exchange collisions in the middle. I *think* my interpretation is the correct one, as both Tyler Christianson and I have noted a non uniform neutron output vs where the detector is - in his case, with a bubble detector, and in my case with a small hornyak button I can move around. This implies either beaming of the neutrons - no theoretical justification has been found for that one - or neutrons being produced at points near or at the wall - neutrons created at the center focus would not show the effect we've both seen, but honestly, the jury is out on this one for the moment. There is no question that coating the walls with a holder of D improves things, however it does that, and also that I should have used something that holds D tighter, like Ti, rather than Pd. Adding the Pd also ramps up the X ray output due to it's higher Z than most of what's in stainless steel, which is a negative in my case - I don't want to be seriously exposed to extra X rays I don't have to.

http://www.youtube.com/watch?v=qgczy36997Y


Somehow, my tools - either "recordmyscreen" or handbrake - messed up the sync with the audio - the audio talks about where the mouse cursor was a long time before, keep that in mind, sorry - no idea why this one messed up. And no idea where that hum is coming from unless someone playfully added it in the software I'm using - the setup has no hum when used with other software!
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA


Return to Run Data -- just the facts please.

Who is online

Users browsing this forum: No registered users and 1 guest