FAQ: Operation - An art for lesser fusors

[Richard Hull]

This is being written after many years of data submitted by myself and others here. It was prompted by discussions mostly over the last year and more recently at HEAS 2012.

Fusor IV has been a great fusor and has done very well in recent years reaching its highest recorded output over two years ago - 1.6 million n/s!

The above being said, fusor IV was, is, and always has been somewhat "leaky". The lowest pressure it ever attained with the diff pump was about 10e-5 torr. Now, 8 years after its build, it struggles to hit 10e-4 torr. Valved completely off, the chamber leaks at the rate of about 1 micron every 10-15 seconds! Still it works, but only as a practiced black art. (this should give newbs some encouragement). At the same time it gives the sticklers and vacuumists here, heart palpitations.

I could never really pinpoint the leak using acetone. After one take-apart in 2007 and some welding at suspect points, the system was little better for the two week effort. I just left it as it was and learned to cope with special operational skills learned by doing.

I have posted on how it takes a couple of days of running for about 3 hours each day before I can hit the "Mega-mark". Conditioning is the order of the day. Once conditioned, in subsequent days the fusor can compete with any mega-mark fusor up to its normal max operation of 45kv applied.

Jon Rosentsteil brought some data from his fusor with his numbers to HEAS 2012. Ultimately, it seems that two fusors can be run at the same voltage and currently but have vastly different neutron numbers based solely on the D2 pressure (how much fuel is in the chamber). Sort of obvious it seems.

examples:

Run #4 from the second conditioning night, October 2, where I could not force the fusor to accept pressures of over 5.4 microns and hit peak voltage of 41kv.
All that follows are the best recorded runs each day.

5.4 microns.... 41.3kv.... @ 10ma ......102,000 n/s

October 3rd the third conditioning night

6.3 microns.....41kv.....@13.2ma......208,000 n/s

October 4th the fourth and final conditioning night

7.8 microns.....40.8kv....@12ma......291,000 n/s

On Friday night October 5th, with about 20 early arrivals present for HEAS, I ran the fusor and the best we could do was........

10.1 microns.....42.1kv....@13.6ma......528,000 n/s

On the Saturday October 6th, the day of HEAS conference, the best run after about 2 hours of operation and several runs, constantly raising the pressure and conditioning was........

15.8 microns......45.1kv (yikes!).....@9.5ma......1,485,000 n/s

So, you see, at much the same voltages and currents, the fusor shows a vast difference based solely on pressure of admitted, flowing D2. For fusor IV, the breakover point for fabulous operation seems to be between 10 and 12 microns.

Now, whether I am merely over powering a leak with D2 or if the higher pressures are demanded on limited 42kv operational maximum is an open question.

The lesson seems to be that it is all about pressure, pressure, pressure! Until you get the pressure up you will not be doing a lot of fusion unless you have a WHOLE LOT of voltage!

About 6 years ago, at one of the HEAS events, Carl Willis noted I was struggling to hit the half mega mark. He noted that my pressure was down at 5-8 microns. His suggestion was to throttle open the chamber valve a bit more and pour in tons of D2.
It worked! I hit the mega mark soon after and have never looked back. The secret is fuel, fuel, fuel, especially if you have a leaky system. It turns out I was way too stingy with the deuterium. Remember pressure in a working fusor equates to fuel.

If you can't climb way up on the cross sectional curve in voltage, you can make up for it a bit by blowing in more fuel.

The advantage of having a fabulously sealed vacuum system is that conditioning takes far less time. Conditioning seems to be the act of burying a lot of D2 in the chamber walls to enhance the process by adding a bit of target fusion to the IEC and other possible fusing processes present in the fusor.

So sealed like a professional or not so well sealed, a fusor can be made to work and work well.

Richard Hull

[Chris Bradley]

For those who are looking for a detailed technical explanation of why a higher flow rate would be required for 'a leakier system', my FAQ may help explain this, and also how to accomplish it: viewtopic.php?f=6&t=3224#p12734

[Doug Coulter]

There could be other issues with a leaky system not generally kept under vacuum. That would be water, which (doggone polar molecule) sticks to everything, oxidizes tungsten, and generally is a mess. It could be mainly that you're getting rid of water when conditioning - I see that here with a new Pfeiffer mass spectrometer all the time.

My recent Q plots show that actually, I get the most Q (but not the most neutrons) at a somewhat lower pressure than my max here. Now, since my gas gage is known to read a bit off on pure D, you can't compare my numbers directly, quite (see the PKR 251 data sheet for some of the error). My Q is in utterly arbitrary units (neutron detector cpm/watts). The particular detector I used was calibrated at HEAS 2009 and reads roughly 1k cpm for 1meg n/s. It's a hornyak (eljen) and very stable, so I think it's still reading correctly.

But it shows Q peaks at relatively lower pressure in my leak free system that's kept under e-8 mbar when not in use. If I open the door to do things, yes, then I get to "condition" for a goodly while, and I have to work up voltage pretty gently (by reducing gas pressure so I don't current limit and do other harm like melting my ballast). Takes a few days of pumping and baking or fake runs to get good again after an opening of the system to shop air. I can cut this to hours if I fill the system with dry nitrogen or argon before I open the door, and work fast.

Every now and again - not common, but it happens - I hit great numbers in an unstable mode while conditioning. More work to do to sleuth out what and why on that one, and to reproduce it more reliably.

Plot of Q:
http://www.youtube.com/watch?v=WJe0YBAX ... ature=plcp
at roughly 1:120, note the exponential rise in Q vs voltage in this video. at about 1:36, note that my highest Q is NOT at the current maximum at all, nor at highest pressure. This was taken with a very well "broken in" fusor, well outgassed, and in fact, just before I tore it down to replace some of the HV feedthrough glass.

Plot of neutron output - the exact same data displayed differently.
http://www.youtube.com/watch?v=8vpCt5kq ... ature=plcp

Same data. Unfortunately, my audio is "off" by some seconds from the video, so I'm in general talking about what the plot looked like seconds ago. Gotta fix that record my screen tool! In this one, right at the start, you can see the GUI I built to control the plotting. For you software guys, this is perl code, eating text data files produced by my PIC based data aq device, one sample per second (one dot), using glade to create the GUI and control Gnuplot on linux.

There can be flaws in this data, prime among them is something that happens too quick for my a/d to respond to say, a current spike, but the counters never drop a count (I've fixed my EMI issues very thoroughly, however). This is why I admittedly edited out a few "too good to be true" points for both plots.

Moral of the story - different strokes...

Yes, you can run with a leak. Shop air actually is one of the better stabilizers in tests here - a tiny bit of water is good if you want stability, say 1/2% or thereabouts. However, it cuts down both Q and total output as well. We tried other gases as deliberate contaminants, and water is the most effective (in whatever sense) one. Argon, neon, nitrogen - took more of those to produce any noticeable result, but none helped Q. Water did help stability, but not Q or output.

So, yes, you can have a bit of a leaky system and do fine, but not top rate. I've not found that heavy hydrocarbons have a ton of effect (my system is oil free re the pumps, but sometimes I lube an O ring with vacuum pump oil and some of that creeps in according to the mass spec). Again, tiny amounts there, however.

Baking out your system from the inside - a dummy or conditioning run, is about the best way to get the crud out. A "fake" fusor run of a few minutes is more effective in my system than an hour of baking with stab-in quartz heaters FWIW.

[David D Speck MD]

Doug,

Have you tried short wave UV irradiation as a means of speeding cleanup of the chamber?

I have some short wave mercury vapor arc tubes that I would consider using as adjuncts to help cleanup. I think I've heard of other trying them in non-fusor applications.

Dave

[Dennis P Brown]

If memory serves, hard (short wave) UV can break the water bond. So it will clean a vacuum surface (while the walls are under moderate vacuum!) Just remember that if you have a quartz window, short UV will pass through it and that is deadly for eyes. Use proper safety glasses. UV can blind.