Fusor remote - the big one

This is somewhat of an admission of failure. You can't easily pigeon-hole everything, and most real projects use commercial software, homebrew, and hardware all at once. So, for you makers out there (including me) - this is where to put whole projects that don't fit well in the other forums.

Re: Fusor remote - the big one

Postby Bob Reite » Wed Sep 07, 2016 7:37 pm

When I do mine I think I'll be using a good old fashioned mechanical relay in series with the SS relay. I would also have the option of just turning off the power to the shed, which you may not have with your setup.
The more reactive the materials, the more spectacular the failures.
The testing isn't over until the prototype is destroyed.
User avatar
Bob Reite
 
Posts: 142
Joined: Wed Nov 11, 2015 1:02 pm
Location: Wilkes Barre / Scranton PA

Re: Fusor remote - the big one

Postby Doug Coulter » Wed Sep 07, 2016 8:58 pm

Bob, that was my thought at first too. I realized that in this case, the only thing that can make real neutron output runs on 240v, and either SSR opening on command (and that only if all else fails, I have interlock and other control) should be good enough. There's not much room in the box, and not much time on the calendar for this, so...it is what it is. I may add a mechanical one at some point, dunno. That's fairly skinny telco wire going over there (it s a long run and shows around 6 ohms round trip) and a lot of coil current might cause other issues.

In fact, Bill and I put the thing on the fusor today for its very first time. One wiring error, and one small misunderstanding of the control requirements for the SL2KW later, and it all appears to be working correctly, complete with feedbacvk to the hardwire panel. Geiger counter and voltage sensing are tested and working on the computer side of things as well. It got late, and we were tired, so stopped for today.

Heck, for the manual over-ride no-computer panel I now have two of those red flap covered "Arm" switches and a "pickle button" just like a fighter plane. Semi-cool. And oh yes, full round trip feedback to leds about the state things are really in, not just how the switches are set.

Pressure gage testing next. (re)Calibration of a few things - is this 5v reference the same as the old one... (this is a lot of hoofing back and forth so I'm trying to do it when Bill around to be at the other end with an FRS radio), some camera mounting (they work but we want them pointed correctly etc). Then gas in and out solenoids, with the fervent hope that catch diodes at the solenoid end of the cable will keep kickback noises out of our box well enough so the computers don't crash. I'm using rather large (14 amp, 50v) logic level gate fets to switch those, with a big R in the gate (100k) so they won't switch too fast, I hope. But I'll definitely put fast catch diodes at the far (solenoid) end of the gas control wiring.

Then we can actually run in "fusor.net" (eg a mere few million neutrons/second) mode to refine the rest - and do the "other" realtime data plotting at the remote op position (SQL replication being a prerequisite I've not set up while experimenting with basic operaion - I've made a mess and wiped that database a zillion times and replication would get real busy for a long time duping that - back to "zero" - if I'd set it up earlier).

Since there's just one real tech guy doing this (me), I have to think "operations research" level and minimize the motion, or I'll never get it done on time. There's a plan, it's just not the most exciting thing I could be talking about.

Doubtless there will be some backing and forthing with EMI created by the fusor itself. Fingers crossed, we're taking it into account in the design. Whether we did enough? We'll be finding out.

But, marathon day, lots of good things happened, so far, so good.
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: Fusor remote - the big one

Postby Doug Coulter » Thu Sep 08, 2016 2:54 pm

Taking a break from all that wiring and grunting on the messy floor over there - this isn't the same as my nice spacious well-lit workbench that all this got built on.
But, I have some neat pix to share from testing out two of the three cameras; I'm not sure at this point if I need the third overview and servoe'd camera, as it's going to be a heck of an emi risk with that long flat cable, and I don't have a couple meters of 1" diameter braid to put over that. I have been using the 3rd one to just peek at the power supplies' front panels while I test the remote control. Don't even need good focus for that part.

It's slightly weird sitting here at the op position and hearing some very high fidelity, perfect imaging audio from "over there", fans, pumps and so on. Almost as good as a picture.
But not as good as these pictures:
im_0000_20160908_191343.jpg
4 channels of 2.5 ghz sample rate DSO

I remember the days when we wouldn't convert data to human-readable ascii to save bits over our low bandwidth modem. Now you just real time stream HD video from a scope screen, as it's easier and cheaper. Wow.
im_0000_20160908_193731.jpg
Tank innards, seen via top viewport (all under lead).

This cam really takes a beating despite the half inch thick leaded glass it's looking through. Sparkles from X rays and neutron hits on the sensor when the thing is running. RIght now, I just have a 500w "stab in" bakeout heater for light.

Now back to it. I have gas control solenoids to wire up and test.
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: Fusor remote - the big one

Postby Doug Coulter » Sat Sep 17, 2016 6:16 pm

Rrealtime plotting from SQL replica coming along, far from perfect, but...
FWIW, I'm doing the dev on an odroid xu4, which has two advantages over the big machine - it draws next to no power so I can work on it after sundown...and it's very pi-like so the habits transfer easily.
Clearly, I have an issue with the ion volts measurment, which was showing up weird on the analog meter too, earlier. OK, maybe I either need to find and fix the ptv supply or just make another voltage divider.
You can also see in these a difference in both start time and noise floor between the two arduinos I am using. The one doing pressure is the newer, better structured code that takes 10x the number of samples and does some fancier DSP on them - it's so much better that it's still nice and quiet after exponentiating the input.

There are for certain NOT our usual operating conditions. What I'm doing here is using way too much gas to load the various supplies to their current limits so I can calibrate these plots to the analog meters when things are stable.
You can see (with some interest) that even under total current limit conditions on the main HV, that turning the ion source on and off (obvious despite the noise) changes the voltage it current limits at.
Screenshot-1.png
Before I added ion current.

Shows my dinky GUI for this too.
Screenshot.png
After ion current added.

Ion current and voltage are measured at the supply, not after the ballast R, so I'm suspecting that crazy volts plot is instrumentation error.
Screenshot-2.png
Pressure, nice and smooth - and BTW, accurate.

Well, accurate is relative in this case. That's what the pfeiffer display said too, but the dox on it say it reads high on hydrogen, which is what was in there (well, it was deuterium, but...).
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: Fusor remote - the big one

Postby Doug Coulter » Fri Sep 23, 2016 11:49 am

Here's a screen shot from the end of the run in the videos:
ControlTest.png
What I see during a run.

The counters are scaled in CPS, which for the hornyak neutron detector works out to 100cps = 6 million neutrons/second (I may change the scaling to counts/minute as that's what I think in).
I'm streaming two videos live from two raspberry pi's that are recording and slaved to a third pi, which controls 2 arduino unos that actually take the data plotted here - that pi talks to the operating position and also runs mysql for the data, which is real time replicated here "just in case". Times are in tenths of second since start of run. At tne end of the run I turned off the high voltage and ran over to put the silver, then the indium on the geiger counter, but silver decays very quickly, and indium is harder to activate - you can tell the difference between the background count rate and when I did those things - barely, at the end of the plot for counters. No worries, we first of all can trust the hornyak, and secondly, if the gonzo mode works, the indium at least will be plenty hot after a run.

Note on pressure. The documentation for the Pfeiffer PKR-251 gauge I'm using says it reads high on hydrogen (or presumably, deuterium). But we're in that middle zone where no one knows what to call this vacuum, and in it's middle zone where it's not sure whether the ion gage or the pirani should take over. So it's a factor of something a little less than 2 high on pressure is the best estimate. Fusors are so touchy that in fact they are far more sensitive to pressure than any gauge you can buy, and after you know the number once...you can count on the fusor itself more than anything else. Too much gas and it becomes a pretty good zener diode and your supply current limits. Not enough and it won't light off at all, and the range is way under 2::1 - even with an ion source. We are in this funny place between "viscous flow" and "molecular flow" which vacuum math doesn't handle well - even when the particles aren't charged. It only gets more complex than that. This is the 3 body gravitational problem only with 10^20 something bodies and two "polarities" of gravity. We have both molecular and viscous flow at the same time in different places, as well as whatever you call those things with ions that can "see" each other even if they're not bumping into one another via Coulomb forces.

I think I learned an important lesson here. When you live in "nowhere" and only have 1 megabit/second top speed, one doesn't upload near-gigabyte videos often. I should have edited out a few snips to comment on here, and will in the future.
Video times are UTC via NTP.

Scope: https://youtu.be/9_UBf5TTTug


Key - yellow trace is the neutron detector. Anything > 2 major divisions is a real neutron. We are triggering on this.
Blue trace is a faraday probe in the tank, about 3" from the main action, and opposite the ion grid. When it goes big-negative, the main HV is on, and we note this isn't exactly a neutral plasma, is it.
Green trace is EMI picked up by a probe a few inches from a wire on the ion grid (south of the ballast R). When it jumps positive, it means the ion grid stopped drawing current for a bit.

Now I can proceed to try "gonzo mode" from a safe distance, or nearly. Almost all the kinks worked out of the remote control now. Minor changes, then we can go places and change the world (I hope, at all, and for the better).

Grid - fusor innards. This was a lot shorter as the video compression was easier to do, I suppose - factor of ten or so fewer bytes, same time (cams are started in sync).
https://youtu.be/lqxArGT1s5c

Yes, I got the grid a little warm by the end. I was playing around with the tradeoff between running higher pressure and the main grid alone, and running less but with an ion source. The data is good as far as I can tell, for my setup. This has all been calibrated the best I could manage, and the plots (click to enlarge) seem to be "on the money".

Key: ion grid in lower right, main grid upper middle.
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: Fusor remote - the big one

Postby Doug Coulter » Sun Sep 25, 2016 4:51 pm

Couldn't resist posting the results of a short test run today (I'm working on the plotting code and video compositing, more soon). Who says perl has to look like line noise or be slow?
Sure, almost any language can be misused - they all hand you enough rope to shoot yourself in the foot...and yes, I'll be posting the source up here when it's ready, or close enough.
Though a lot depends on my exact setup here, it's pretty good as an example to modify for your own.
I changed the scaling to CPM, since that's what we all think in, and this is why the counts look quantized - we are actually taking counts at 10Hz and have to multiply by 600 to get to CPM.
I changed to points on the geiger and hornyak plots so the odd line going back to zero (with short counting intervals, you get those) wouldn't hide the stuff behind it, and added an averaged neutron count as a line.
SunRun.png
Broke 3.8 million neutrons/second here. In plain old Farnsworth mode. Could be worse. The machine seems to like its new robot overlords.
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: Fusor remote - the big one

Postby Bob Reite » Mon Sep 26, 2016 7:35 pm

What was the power input on that run? You ought to calculate your current Q. I'll bet your current "Farnsworth mode" machine will be at least second place on the fusor.net "Q contest".
The more reactive the materials, the more spectacular the failures.
The testing isn't over until the prototype is destroyed.
User avatar
Bob Reite
 
Posts: 142
Joined: Wed Nov 11, 2015 1:02 pm
Location: Wilkes Barre / Scranton PA

Re: Fusor remote - the big one

Postby Doug Coulter » Tue Sep 27, 2016 6:08 pm

Q is for certain "very not bad" for this setup, and it was decent before. I've not yet written the parametric 4 (or more) D plotting software, as I'm in a real time-crunch just now on the patent process.
But of course, we could easily win the Q battle vs the Farnsworth/Hirsch/Meeks crowd and it still stink for anything practical. We don't need 2x, we need mllions X - and I think we had it earlier.

This is regularly breaking 4 million neuts/second with ~ 400w in, FWIW. That doesn't count the ion source or the vacuum pumps and instrumentation. Call it 8 microwatts, or by my (I think better) measurement, more like 10 or 12 (free neutrons not only have kinetic energy, but give another 2 megavolts or so on decay or capture). So, factor or 10+ million or so is what we would rather have here, and real close to what I think I got back then that almost killed me. This is all dry-run for that, getting the remote stuff reliable and safe in the old mode at least. And this time, with remote replication of the database in case the fusor-local computers should crash as they did last time. If I had that data, we'd have saved a couple years.

Depending one what Bill wants, the expensive legal team tells us there is another, cheaper way to put this in public domain I missed on first reading of the "America Invents" "reform". If he decides to go for that (saving ~~ $20k and a lot of stress on me) - I can get details on the good mode (which this was not) out a lot quicker.

Basically, there's a journal that patent examiners read - pay per publish, but it's comparatively peanuts - we can use to release the info. It has to be written to the old patent standards, eg, not obvious to a practitioner of the art without this info, but replicable without undue experimentation with it. Once it's there, it constitutes the same thing that "prior art" used to.

It's not just up to me as I'm not the only one with money and time in this - but for the record, I don't care either way. I feel like if monetization wants to happen, our skills and experience will be incredibly valuable to those who want to "go big" with it - saving them quite a lot of money, (== time) some of which would obviously come our way. More or less the same situation either way, in other words, but faster and cheaper. You can't really keep science a secret anyway. Especially not if there's an existence proof out there. Remember, the Russians took a lot less time to get a nuke than our somewhat vain generals and scientists predicted - after our use in Japan. They knew it was going to be worth the try, no risk there. Applies to lots of things.

Having just had surgery today (large cancer removed from face, or so we hope), I'll be down for a little bit, but back to it really soon. We're right on the cusp of the re-wiring for replicating. Exciting times! When I can move around again without opening the wound and get some food in me...I'll be back on it with maximum effort.
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: Fusor remote - the big one

Postby Donovan Ready » Wed Sep 28, 2016 10:40 am

Damn. Take care of yourself first, Doug.
Donovan Ready
 
Posts: 239
Joined: Thu Apr 17, 2014 1:22 pm
Location: Austin, Texas

Re: Fusor remote - the big one

Postby Doug Coulter » Wed Sep 28, 2016 1:52 pm

Aw, it'll buff right out. Right?
100_3115.JPG
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

PreviousNext

Return to Combined projects

Who is online

Users browsing this forum: No registered users and 1 guest

cron