Microwave Ion source

[Doug Coulter]

I'd looked at cavities at first, but was daunted by "how to get ions into the tank" from one, as well as the lower fields per watt usually obtained. It's illuminating to look at Terman pg 273 as he shows field evolution as a 1/4 wave stub is evolved into a cylinder cavity. That real high concentration near the hot end was what got me going on the stub design, where I could run the gas tube right through that spot and then into the tank via a simple tubing compression coupler. There were some significant tradeoffs here. I couldn't increase the diameter of the stub design without risking "monkey modes" in the large space, rather than transmission line behavior, and I couldn't make the holes for the ion pipe any bigger as they'd become significant microwave leaks. So that's where I wound up for that one -- as is it's pushing some limits, it's kind of too big to be a good transmission line, almost a quarter wave in diameter! In the new one, the issues are going to be more like sputtering, since there's no way to have an electrode-free area where the ions and fields will be and still put it where I want it, and I'll have to use some sort of low-sputtering, high temperature material for the screens (probably Ta wire for the first go, since I have some).

I had another scheme for fusion in mind, that would have required chopping and bunching ions while they were going "slow", and was looking at that klystron type of design you mention to do that. But the doggone heavy ions have such slow transit times at reasonable energies I couldn't make it fly even on paper with reasonable sized stuff. You'd need so much power working over such a short gap you sorta couldn't make what I had in mind at the time.

I'll be making and trying the new design fairly soon, pretty much as soon as I complete the turbo driver design -- it's close now, just fighting some noise issues in the switcher I use to replace the nasty high power series resistor in the original. I might have some fried parts in the proto, so...gotta do some more work there. The additional software needed on top of the opsys I have posted up here is pretty trivial, and will give some nice advantages in control and data acquisition.

[Joe Jarski]

I was thinking of using a thicker wall copper tubing - something like 1" OD X 1/8" wall to reduce the size from a transmission line standpoint and if I understand correctly, the thicker wall will reduce leakage around the holes also. I'm not sure if it's enough to make a difference though. There are a few other sizes, but a thicker (1/4") wall would be messy to couple from the coax hard line and a larger OD would probably push the magnets too far apart for the required field.

[Doug Coulter]

Well, the big downside of the thick wall stuff is getting your hands on it or making it, I suppose -- if you look at the pix, the magnets clear by plenty, they are nice an loud; being able to spread them out also means I got a little more uniform field in the middle, not that it seems to matter all that much in practice. In one version with a 3/4" pipe hole, we had to wrap some conductor around the glass tubing to keep the leaks down, but that one wasn't as good for other reasons -- needed more power anyway to get to voltage across the larger end gap. As the skin depth at these F's is tiny, no other real need to make things thick. In general we don't want super high Q, just enough to make the thing a good stepup transformer till the gas lights off -- then the Q drops anyway due to the load.

So, I'll not super-finish-polish or silver plate the new one unless I have significant heating losses there. Just make it basically right in the first place (so I hope!). I'm hoping to get tuning into the ballpark at STP with a full power "normal" microwave oven source, so I don't have to tune in vacuum, which is a pain. Else I'd need a swept frequency generator in that region, which I don't have and don't really want to buy right now. I'll probably just electropolish/bright-dip the new guy and see how it goes.

Ever tig welded copper? I'm thinking braze, but nearby there's going to be a SS stub that gets tigged to the tank, and I worry about remelting the braze if I do it that way. Seems like the old welding of thick to thin issues would be present in any case, ugh.

[Joe Jarski]

I've TIG welded a little copper years ago, not on anything real, just to see what it welds like. Thin stuff is a bit of a balancing act between too much heat and not enough - it's a lot like oxy-acetylene welding aluminum in that respect. You either have to move real fast once you get started or throttle the heat in between adding the filler otherwise the puddle will get huge. Surface tension seems to hold it in place pretty well without blowing a hole through. It's not too difficult with a little practice. I haven't done any thick stuff, but it would probably be a bit easier. The welds look much like they do in back gassed SS. If you go the brazing route you should be able to find stuff with a melting point very close to that of copper, so remelting the joint is a non-issue.

[Doug Coulter]

Well, I tried putting this together, first brazing the lower disk to the ISO nipple, and just plain failed to be able to do it. The thick to thin issue, combined in this case with the fact that the thick thing is also super thermally conductive means I either couldn't get the brass hot enough, carefully enough without slumping it, while I got the SS so hot it wouldn't wet due to it oxidizing. I have a decent skill at torch brazing but this one I just couldn't pull off. Rats!

I'm now considering either trying to TIG the dissimilar metals together, or furnace brazing, neither of which I have any experience at all with, or redesigning this thing so as not to have to make those particular joints. Could be I could redesign the cavity dimensions so it can all be inside the nipple (no fat place) with just some disks and pipes there. Don't know if we'll have enough heat to make soft soldering that a problem....which is a little scary.

Suggestions welcome.

[Joe Jarski]

Finally got around to finishing up the machining on this thing. It's probably a little on the fancy side for a test piece, but I wanted to get some good practice at machining OFHC copper before I start whittling on some larger chunks for other stuff.

All of the pieces parts

I still need to get something to cut the quartz tubing and finish up the anode & cathode along with the power supply.

[Doug Coulter]

Wow, that looks super nice! A thing of true beauty! A lot better than my kludge at any rate, it should work great once tuned (which can be a little tricky). Yup, OHFC copper is something that takes some practice to machine well, if at all. It likes to climb lathe tools for one thing. Kind of "gummy" stuff. I've had no special troubles with it on the mill, but on the lathe....not so easy sometimes.

I cut quartz on the lathe here, it's the only thing I've found that will work, and I'll cut you some if you like. I put one of those cheapo diamond cutoff wheels in a kind of toolpost grinder I made, and cut with the lathe going slow (about 60rpm) with water lube, mostly to keep the diamonds on the wheel, and the dust out of the air. Think dremel tool mounted on a toolpost and you'd be fairly close (and I actually started out with just that, but it's too hard to make a dremel go slow). But then I made one with a 24v servo motor, a belt, and a speed control to chuck 1/8" bits in a collet I made, works like a champ, also for ceramic tubing. Both of those you just cannot score and break, they simply shatter all over the place and ignore the score mark, even one halfway through the piece. Nothing like pyrex, which is literally a snap. Since quartz isn't cheap..it's better to get that right on the first try.

Edit:

Aluminum seems best for the cathode, as it doesn't sputter much and you can machine a super-thin tube of it for that part. I'm letting in the gas through the anode, but that's very optional and has "issues" the way I run it now, because you then have to insulate the gas line, for which pyrex works, but at some pressures it becomes a "neon sign" and lights up back towards the grounded tank plumbing, not good. It's probably better to just machine a plug and O ring seal it as I did, and get it right up close to the action for best results. You could make the end pointy, but I'm not sure if that helps much -- mine is (it's the capillary tubing I pointed via electropolishing) but in mine it's too far back for the best ion push action, to tell the truth.

Edit2:
In case I didn't mention this clearly enough somewhere else, to tune this at first, start with a gas pressure high enough that the DC field from the electrodes will light off the gas, when then becomes a very sensitive indicator of any RF present, and tune for max light, without arcing. We had troubles both near the end of the feedline from the magnetron, and near the end of the tuning stub. Both of those need to be adjustable, and I had to round both off to make it not arc at reasonable power levels. You'll have to find the right current for your magnetron, but mine wanted something from 20 to 50ma tops. Below, it doesn't oscillate at all, and above you get into other problems and fried tubes, they weren't meant to be CW.

[Joe Jarski]

Thanks - my back-up plan if it doesn't work is to have an interesting conversation piece. The OFHC copper wasn't too bad to machine. I ended up using some transmission fluid for lubricant that was left over from a rebuild, which seemed to work well. The outer piece of the coax that goes around the antenna has some "decorative" contouring though that was added to get rid of the marks from when it screwed itself out of the chuck when I was boring it out. Just when I thought I knew what I was doing, it had to throw me a curve. Keeps me on my toes.

I'll get one of those diamond wheels and take a shot at cutting the quartz. If I start scrapping too much, then I may take you up on the offer.

Once it's finished I'll probably just mount it to my sputtering chamber for educational purposes and fiddle with the gas feeds, extraction and stuff like that. I want to get some practical experience with beams and focusing and all that other stuff that I know nothing about. Both the inner and outer coax portion is adjustable, same as yours, but the tuning was one of my biggest concerns and something I wanted to talk you more about. Seeing yours in action should help out with that.

[Doug Coulter]

Extraction is far from perfect in mine at the moment -- there's just too much distance between the glow and the tank for it to be super easy to get right. The cathode does act like a lens, and so increasing the field can make the focal length so short focus occurs inside it, and after that the ions spread out again and hit the cathode walls. A better design could certainly be done with a bit more complexity there -- there's some interesting literature on it in our library. The real issue is that long skinny path from the glow to the tank boundary, anything that makes that shorter will help a lot. I have a 1.33" CF flange tubing coupler bolted onto the thick 8" CF flange, which is just a lot of thickness there, because that's what I could get done with what I had. In theory, one ought to be able to get around ten times the ion current I'm able to get from this due to that issue, which was why I was looking at other ways to get this done as 10x would be truly fat for the fusor.

Those diamond wheels are one of the few things worth it at Harbor Freight, I really stocked up on those. Takes a little finesse to make them centered on the crummy mandrel, but worth it. Or one could make a better mandrel. They do wear some, and that makes them centered if you are careful not to break workpieces while they are getting worn in.

In mine, the whole useful tuning range is inside 20 degrees on a 1/4-20 stub bolt -- and the really good spot is well under 5 degrees of turn. The coupling tuning (the length of the hardline and protrusion into the cavity from the side) seems less critical. We hit something that worked in a couple of tries (I also had the length of the center conductor separately changeable via a screw threaded into the end of it). I found that putting in any kind of tweaking tool to adjust that messed things up so bad as to be worthless, and with some materials, just burned up the tweaker! So I wound up turning off the power, tweaking, power back on, and so on. At the time, that was easy as I was running a separate filament power, so I could do it pretty quick and not lose track of where I was at there -- I only had to switch the DC power. To tweak the DC current, I just used (and still do) a variac on the power supply input. That's also moderately sensitive, as about the 89 to 92 volt indicated range is "it" and it doesn't work right outside that with the particular transformer/tube combo I'm using.

[Joe Jarski]

Doug, the diodes that you're using seem kind of overkill (in speed & current) to me. Am I missing something or was it that you just had those handy and put them to use?