A new fusor build!

[Doug Coulter]

This isn't what I've described elsewhere as Fusor-II, just another fusor -- the next step to the big changes I plan. This one is being put together on my smaller turbo system to test some different things. The system was originally built to do beam on target work, and someday, it will again. But I wanted a place to test some things without disturbing the main fusor, which is in a very nice state of tune just now.

So, we make another one to try some new things on. As I write this, it's sealed back up, baking, and coming down in pressure about as quick as I can expect, given that I just touched a bunch of things in there -- fingerprints everywhere, new HV feedthrough, grid, electron emitter, faraday probe, secondary grid to influence the main field, and a new screen to protect the window from charged particle bombardment. That's a lot of new surface area that is only now seeing vaccum for the first time, and after about an hour of pumping and baking, it's down to about 4 e-5 mbar.
Before I opened it this morning, it was hitting low e-7 mbar, and that was with a leak I found with the new mass spec. I think I fixed that, but there's just a heck of a lot of new stuff with a lot of surface area in there now to outgas. It's baking at about 70C and warming the room nicely -- should have done this in winter!

Pictures are fun and not only for eye candy. They save words (even for me!)

Here's the thing as it now sits. HV isn't yet hooked up for real, but I have a ballast resistor standing on end there. I have a 60kv 150w supply I plan to use on this, as soon as I make a cable for it.

Overall view

Here's what's on the computer screen right now as I'm baking it out:

Mass spec output

Yup, water, air, dust, fingerprints. Not unexpected at all. I expect that tomorrow it will look like it did yesterday, or close.
Here are the new screen for the window, and one to lay on the bottom of the tank and act as an electrode to both measure and perturb the E fields we will have.

New screens inside

This is stainless steel screen Bill brought over the other day. The electrode one is tied to the fat wire with nichrome wire, in the manner of sewing. I also sewed a few fish beads to the bottom to insulate it from the tank. A piece of pyrex tubing is the main insulation, and the other end of the wire plugs into an octal tube socket I put on an 8 pin feedthrough. Can't do much kV on that one, but that's not what it's for.

Here's what it looks like before I put in the protective screen and seal it up.

Innards before final assembly

I've already run a glow discharge in (mostly) He with an oil burner ignitor to see that things basically are OK. I expect to run real fusion tonight or tomorrow, depending on when the beer overcomes my shop and fabrication energy - and if it's not too late at night -- we run on batteries after sundown, no point wasting them.

If It works, the other neutron detectors are only about 20 ft away as the crow flies, so I should be able to see them get out of background without moving things. One of the things I plan to "test" here is a couple of BF3 tubes we've had but not lashed into anything...

This will be a severe use-test of the new HV feed through I did the other day.

[Doug Coulter]

I already used he shoots, he scores. The one day fusor has produced neutrons! Not many, and there are "issues" but...not bad at all.

This one is hard to take pix of due to the screen in front of the window and it's also not real bright with the wimpy power supply. I'll have to get out the lead (get the lead out?) and shield this sucker, as henny penny says I just got about a chest X ray doing this. It made just enough neutrons to make the detectors downstairs count maybe 10 cps (background is roughly one cps).

Pretty haywired at the moment, I'll fix 'er up a goodly bit so I can do the tests I wanted to do -- after doing the deal with the lead, of course. I'm not a true believer in radiation hormesis.
At least none of my phosphors or scint crystals out in the room lit up real bright . Interestingly, behind that grounded screen there were little lighting patterns across the face of the window, really neat looking. And everything around picked up a static charge, so ... the usual more work needs to be done.

But that has to be some sort of record -- a fusor in one day. Yeah, it does help to have tens of thousands of bucks of working gear "just laying around" and a little experience, I'm not trying to boast -- I'm amazed myself.

Oh, and for Cliff's benefit -- that Gamma research supply I tried was just too wimpy to do this, perhaps it needs a current shunt or something. So, as one might expect, a trusty Spellman PTV-200 was what turned the trick. Spellman rocks!

Now for that beer.

By the way, the camera didn't mess up the color. That's a D and "junk" mix, a lot of water and some hydrocarbons mainly, and that's what that looks like. I've found that some crud makes a fusor more stable, and I need to condition this grid, so that's fine for the moment, and I expect it will get better as things finish outgasing anyway. This grid had a funny poisser in the other tank too -- I may try it without the ceramic rods over the vane elements as well to see what I see doing that. It's currently the most precise grid I've made -- made out of "billet" graphite on good tools.

[Doug Coulter]

'Cause we're going out tonight! There really is no such thing as a one-day fusor, that was sort of a stunt to say the least. In initial testing I forgot how much comes out that window, and picked up an estimated 3 mrem of X rays before I got the dosimeter and then quickly turned things off. Well, that's not acceptable, so on goes the shielding, shown in progress here:

Dressing the lady in situ

We bought a qf-40 to CF-2.75" adapter from Andrew Seltzman, so I could relocate the mass spec head in an out of the way and more useful place (depending on what's planned), down closer to the pump and inside the main tank valve. So now it's here:

New spot for mass spec

The idea in this case is to be able to look at gas compositions when the main tank is above the pressure at which the mass spec can work. By relocating here I can't see tank ions directly anymore with the spectrometer ion source turned off - but I have that going on in the other system anyway. Also, since this system is in a high foot trafiic area, that head hangning out into the aisle was a worry, no more.

So, after some more fiddling, there's still work to do, but it's an improvement.

Needs the tough parts still -- all that filling in the gaps in the lead and something done about the window. New inlet valve installed.

Note that this was a 2.4e-7 mbar before I tore it down. After two hours baking at 150c, and cooling down it was still at 4-5 e-6 mbar. The next morning (now), it's at 4 e-7 mbar. Quite a lot of the mass spectrum shown isn't air -- it is mostly water (as usual) and the water is bigger than the air, so if I have a leak, it's a pretty dinky one - probably the new inlet valve, as I'd forgotten to clean the seal button before putting it together -- no big deal, I can do that easily at any time.

Most of that high mass crap seems to be an artifact of the mass spec itself -- it all goes away after half an hour, but is always there right when I turn it on. This thing must have been used in some dirty process in the past, or something similar. As of half an hour after taking the above picture, it's mostly gone except some stuff in the CO2 range -- normal after you've wiped down some O rings in solvent during assembly.

Note that two hours of fairly stiff baking wasn't even close to enough and that all night pumping after that still wasn't enough. Most people without mass specs have a completely wrong "feel" for just how hard it is to get to hard vacuum even in a leak-free system. Just no idea. A couple days with one of these will change your vacuum technique for life, and you won't need one as much anymore - but till you actually experience just how hard it is to get rid of that adsorbed gas when you're in molecular flow (eg random billard balls of gas atoms leaving, then sticking on the next thing they hit till you get lucky and they go into the pump and are gone) you just can't use your normal intuition on this kind of thing. As a wag at Lesker said, "vacuum doesn't suck". You have to wait for the individual gas molecules to "blow" out of the thing, and the system walls are like flypaper to them, so they keep getting stuck on them, come loose now and then, but more often than not, simply stick to the other tank wall and wait for nanoseconds -- to weeks, before doing it again. Water, being sticky (polar), is the worst of the normal things you run into, but other things you should never even allow in the system if you ever want high vacuum. (Cd, Sulfur...Zinc, Solder...oils...you get the idea).

I laid the ballast resistor I'll be using on the tank so you can see it. The PVC fittings will hold this and protect me some from HV, and will be in turn covered with grounded wire mesh for EMI shielding. This system won't have the big input powers I use on the other ones, so I'm hoping I don't need forced air cooling for the ballast this time. In case anyone is wondering, that gage is way out on the elbow to keep its stray H field away from things. Early on I got some "interesting" photos of ions being guided by that right into it from lng distances away, which couldn't have been doing the gage much good, and was sure messing up my lower energy experiments -- so that's the fix I applied for that problem.

[Doug Coulter]

And here's what that mass spec looks like with no changes other than another half hour of warmup. You need to pay attention to the behavior of your gear, or you'll be fooled a lot.
Evidently, this one picked up something nasty earlier in life, and what ever has diffused to the surfaces overnight gets blasted off and detected every time you turn it on. Once that's over with, you have a chance to see what's really in the tank itself.

Mass spec after run-in time -- limitations of gear.

With air about an order and a half below the water -- it's mostly water, and probably almost no leaks. And all that water was from opening the system for about an hour total, on a not especially damp day. Once I get the window handled so I don't get fried, I'll run some power through there in a glow discharge and really blast the walls for awhile, flush, and it should be a lot closer to normal after that. Nothing works as well as baking from the inside -- heat, UV, hot electrons....much more effective than just some heat from outside.

[William A Washburn]

Doug,

This looks like the size and form-factor I would like to use.
Would you please post the diameter and length of the cylinder
so I can begin looking around town and on the internet for something
about the same size?

Thanks, Bill

[Doug Coulter]

The cylinder is the vacuum standard 6" OD (yes, that's weird compared to plumbing where the ID is usually the spec). I got the pipe at Kurt Lesker -- it matches all the standard flanges for welding into them, and did the welding myself. This seems to be a good size for this sort of playing around. I made this one long, about 2', so as to play with not-fusor type things too -- beam on target was what I had in mind when I did it. A good thing to consider would be a CF flanged 6" T (which has 8" OD flanges). Bill's fusor is built on one of those, works fine, it's versatile and more fusor-right-sized. We just adapted off the side arm for the pump system, and put a SS screen over the sidearm hole so the E field in there is still round (no bulge into the sidearm). That way, at least 3 of the most crucial welds are done for you. If I think of it, I'll try to find a link to pix of that one -- it worked out really nice, but the diffusion pump and 1hp forepump are a little spendy of power to use it much here -- if we keep it, I plan to convert it to one of those pfeiffer mini turbos and a smaller forepump so I can afford to run it. My freezer has trouble coming up with enough ice to keep the diff pump cooling water cool in the 5 gal bucket we use for that -- no indoor plumbing in my lab.

The diff pump and forepump constitute extreme overkill for a system that size, by the way, even after I welded in all the 2.75 CF nipples (which add a lot of surface area) and Tees.

On that system we contrived a quick removeable window and other end for the HV using O rings. Works like a champ, but did take a little custom lathe work on the special pieces. You can be in and out of that system and back to vacuum in way under 10 minutes -- both ends just come right off so you have a straight shot into the bottle you're trying to build a ship into.

[William A Washburn]

Do you think this combo would be good enough to pump down this size of container?

Saw it in sale somewhere.

http://www.pfeiffer-vacuum.com/products/pumping-stations/hicube-tm-turbo-pumping-stations/hicube-eco/onlinecatalog.action?detailPdoId=5161

[Doug Coulter]

It's roughly what I'm using myself. I love the thing. Kind of slow if you put a big tank on it, but rock solid, no oil....moderately programmable. You can add one of their PKR 251 pirani/ion gages to it and it will report what that reads on the UI. I have mine programmed to run the forepump only when the turbo needs it, which lets me leave it on all the time without the noise, power, and wear on the forepump.

That's an ISO (O ring) flange. As much as possible, the rest should be CF (metal gasket) stuff. Too many miles of viton will let water permeate too much.

[William A Washburn]

Doug,

I looked at the Lesker page and the std. 8" tee looks perfect. I have a couple of other questions, though.

I will need two other holes in the straight section of the tee. One near each end. I want to put in one of those small
Pfeiffer mass-specs and an ionization gauge one at each end. I have noticed that you have your mass-spec on an extension tube,
I suspect to keep it away from heat and radiation.

The nipples I have found in the Lesker catalog are too short for this purpose. Did you just weld some extension pipes on to some nipples?

Also I want to use the sidearm for viewing so I will have to put some SS small gauge hardware cloth where the sidearm
connects to the main pipe and weld it into place, bite the bullet and buy a glass closure for the sidearm. One end will be the vacuum exhaust
and the other will be the high-voltage feed-thru. As an aside, since this is 8" in dia. tee can the fuel feed be toward the outside of the
cap radius for the HV feed but on the same side. If not I'll need another port on the side, I guess.

I intend to place some lead bricks under and built up along the sides of the volume to prevent as much radiation leakage as possible.
That means I'll have to look over the top of the wall to see into the plasma. This is not such a good position
for one's head so I thought I could use a motorcycle mirror at 45 degrees there to see with.

Then I remembered a nuclear lab where we investigated gamma scattering and I was quite surprised to find that almost any angle you could get some scattering.
If I try this method of viewing do you think I will Will I get scattering from the mirror?

Thanks, Bill

[Doug Coulter]

The extension tube is more or less a part of the mass spec head - it holds the guts of the thing, and is a non standard size to do so -- it comes with the thing. Gammas and such don't bother it, but magnetic fields (and that ion gage is killer-bad that way) really hose it bad -- a few gauss and it's not a working thing. That's why you see that gage way out on an elbow in that system, it was messing up the mass spec and the ionizing process itself when right on the tank.

You can rig a mirror/periscope to view, but yes, if you let gammas out into the room from anywhere, they'll scatter all over the room and into you-- off the air itself. A leak on the side opposite where you're standing is still a leak, and some will be coming back. My geiger and activation counter is sitting behind a lead glass shield, behind a 2" thick lead brick, wrapped in 1/4" lead on 5 of 6 sides (top is open) and it still counts 3-4k cpm when the fusor is running (the only direct path to the fusor is through that 2" lead) -- because there's a hole in the shielding on the far side, in the back -- and that much gets around the tank and comes in straight down on the other side, obviously after scattering more than one time. If you're going to have a window, get some thick lead glass to cover it. You will be wanting to cover the entire thing with lead, not just build a leaky box around it. Feedthroughs, you name it -- all leak badly. The thick flanges aren't so bad, but the tank walls themselves get real leaky above 40kv main power, or in the presence of any good amount of fusion, so you cut and solder lead sheet all over if that sort of thing worries you. Then you run around with a counter and find the leaks you wouldn't believe were possible, and do it again, this time with more overlap and better joints.

No need to weld in the screen; it will stick fine if cut right ( a little oversized ) and stuffed into the pipe. What would you have done if you melted it and you had welded it? Remember, it's the tank that gets most of the heat, and a little screen can easily go "poof" if hit with enough charged particles -- and you're talking about putting it in the worst possible place for that. Some will get through and ruin your window too if you don't have a sacrificial piece of glass in there (I think I mention things like this elsewhere as well).

You can get blank or pre-drilled flanges and weld whatever you want into them. If you're going to do any amount of this -- it's cheaper to buy a (good) TIG welder and learn how than to buy it all from Lesker or even surplus, and a lot more empowering. We tend to find blank flanges and then bore and weld them as needed, those are cheap. But they're not fun to bore -- that SS is tough stuff and it takes a little practice to work it without making it work-harden so badly you can't cut it anymore.