Page 1 of 1

Well, I liked seeing this

PostPosted: Fri Sep 15, 2017 10:57 pm
by Doug Coulter
Some will even understand what I managed to measure and somewhat prove here too.
fusortiming7.png
Years were destroyed in the making of this picture of data

fusortimingspread.png
But the puppies did survive

Re: Well, I liked seeing this

PostPosted: Fri Sep 22, 2017 4:42 am
by csnyder
I was hoping someone else would show their ignorance and ask you to expand on the traces a bit. The more cryptic you are the more interesting it probably is. I follow that you have been looking at phase shifts and transit times but for me there is just not enough info to say ah ha.
As always, thanks for sharing.
Charles

Re: Well, I liked seeing this

PostPosted: Fri Sep 22, 2017 11:04 am
by Doug Coulter
A close guess, which some Newtonian math about acceleration and time to distance and felt-force will show agrees with one of the extant theories. (the deuterons are nowhere near relativistic)
But I'm not saying much more yet. I'm kinda waiting to:
1. collect more data in a parameter sweep - check that that math checks across a decent range (it's scary-perfect feed-forward but only one data set so far)
2. do something with that data - make things happen that will greatly enhance fusion Q. I have a plan, but it's going to be a lot easier to accomplish without trying to build stuff multi-octave response.

The point of this data is to save having to iterate endlessly over a few parameters.

Then, when it replicates, I can announce what I've been itching to for some time now, and skip the being laughed at part. "Cash talks, BS walks" - as it should in science (and other things).

"Nearly everything anyone has said about Farnsworth fusors is wrong, but almost everyone investigating something electrostatic/non-thermalized is at least partly right..." [me]
No one saw the entire picture...many sins of selective omission, and plain wishful thinking. But almost everyone saw at least one part of the whole. There was one missing, I think I got that one, and synthesized the rest.

After you know (if in fact I actually do) it's a head-slapper. Working on testing those assumptions now. That 99% perspiration thing.

More info soon, for some value of "soon". This took too long and also longer than I expected. I *think* it'll go faster now. Working on a "conceptually trivial but actually hard" sensor right now to get more-accurate triggers for the timing...sub uS accuracy on tens of kV signals with 100 meg loading is no walk in the park (waving a hand a foot away doubles or halves the response, and shifts phase all over). Consistency is easy if the waveform is known, but it changes with conditions here. Imagine making a 10,000:1 scope probe with great frequency compensation...passive of course (unless you know of a voltage follower device that can hack these numbers...eg unobtainable magic).

You can see in the neutron case whether you had one neutron or a burst, with this detector and preamp. Even when there is a range of timings in the current setup as shown in the second pic, there's one that is "the best" even just waiting for chance to produce that timing...we won't be waiting for chance once we know what we want.

Re: Well, I liked seeing this

PostPosted: Sat Sep 23, 2017 10:21 am
by Doug Coulter
A handy tool, if you've got perl and Gtk3.pm - I can post the source for this too.
Screenshot at 2017-09-19 18-54-21.png

Computes speed vs volts for various ions, or if you know time and distance, what gradient and total volts they traversed assuming the gradient was constant (and yes, I know that's a heck of an assumption - but those bottom numbers are "real" and were taken with D and with 50kv and there was no fudging). Almost scary how that came out.

Re: Well, I liked seeing this

PostPosted: Sat Sep 23, 2017 1:07 pm
by Donovan Ready
Doug, I'd love to see the code. Thanks!

Re: Well, I liked seeing this

PostPosted: Sat Sep 23, 2017 1:52 pm
by Doug Coulter
It's here now: viewtopic.php?f=60&t=1053

About 5 lines of meat, the rest for gui junk.

Re: Well, I liked seeing this

PostPosted: Mon Sep 25, 2017 1:16 pm
by johnf
Doug
100mk/s seems abit slow
I would have though nearly ten times this (back of fag packet calc) --maybe gaussian dristribution of velocities
strange that neutron comes at end of pulse --although thinking about it d2 -accel --flight -- crash into someting = xtime which is set by chamber dimensions and accel voltage and damage cascade

Re: Well, I liked seeing this

PostPosted: Mon Sep 25, 2017 2:51 pm
by Doug Coulter
if you're looking at the top part, that's from Fredrik Terman's Radio Engineers handbook, which gives speed vs total potential of 5.97e7 * sqrt(E) cm/sec, where E is in volts (real volts like we use on voltmeters).
Seems to check out with transit time effects in radio tubes etc? You can put the charge/mass ratio vs electron in the square root for heavier stuff (1/1831 for proton - according to wikipedia and others, our old familiar numbers have changed a bit over the years)

Maybe you were faked out by my bad GUI design (it just grew, honest) - for the top lines it's 102 volts or thereabouts...there's more than one volts box. :oops:

The bottom part I put together from the stuff I linked in the code...it was a pita to be sure. FWIW, it works out to 50kv when I measure time, distance, and put 50kv on the thing.
Which seems high for the volts, as I'd expect various things (scattering) to reduce the effective potential in there. On the other hand, this is really low pressure compared to what we normally ran- fairly deep into molecular flow (billiard balls) everywhere but near the focus.

EMI is just a cap coupled signal from the main grid and is due to the supply having some series impedance (it's a scope probe near the HV to the main grid). Faraday_F is more or less when the ion source goes positive and starts repelling deuterons which hit a wire in the tank nearby. There is some geometric error there, and it probably varies depending on the main grid field. I'm working on that one now. Supposedly, the EMI signal would be "highest" when the deuterons are just entering the outside of the grid? (induction)

I haven't really verified the top line stuff yet, other than it's accurately reversible. It was just a start to working out the time to distance stuff, which is really the main thing I wanted to know what kind of drive speeds and feeds to use on a main grid waveform to bunch the incoming deuterons and then yank the electrons in after them. I hope it's close, as those kinds of speeds and feeds are fairly easy to make compared to DC<>Daylight, though it'll take two stepup transformers to handle the slow and fast ones (over 60x different F0).
Having the pulsed ion source is what made the measurement possible. Not perfect, but I wanted to get in the right octave before building things and burning them up.

I triggered on the neutrons, and selected for "a burst that overloaded the phototube/ZnS recovery time". The events of creating and then accelerating the ions naturally would occur first since they happen mostly on the other side of the tank from where the neutrons happen. We see some before the main arrival, and some after, but the peak is pretty much when they seem to be hitting the focus - it's that gaussian you mention, once you "get" what was measured. Sure, we get some on the way in, and on the way out but most where they are densest and head-on.

Happy late birthday!