Electron beam device tests

[Doug Coulter]

To get some sort of gut feel for how charged particles behaved, we built an electron linac. Nothing too impressive here, just a test device. Some parts of of it worked better than others, but as usual, the failures were as instructive as any other part of the project. Now that there are a few here who know a thing or few about beams, perhaps they can add to this and illuminate us all.
Here is a picture of the thing as built up. As usual, click the pic for a bigger version.

Experimental E gun on big tank

After reading some literature, I decided to try a "Pierce diode" as the source. This should have helped get most of the electrons going in a nice straight beam out of it, and it seems to have worked. We had to spend some time building this, and finding out about how to get decent emission under the conditions. We actually did the emission tests with another jig I'd made to be a twin filament "triode" which could be used as either a real triode, or an ion gage. We found that with the filaments we were using that things simply got too hot if we heated the pure tungsten hot enough to get decent emission, in our case about 10ma was wanted. So we did the work to try emission enhancing substances, starting with the things that work well in electron tubes, which was a mix of barium and strontium carbonates, later activated in vacuum. The two filament jig let us run comparisons of two things with only one door opening on the tank to replace them. Well, the Ba/Sr mix did work and at a much lower temperature, but...post activation it was quickly ruined by air exposure, and in our rig we didn't control the activation temperature well enough, and found that if you melt the stuff, that kills it too. So based on a hint in some paper, we obtained (thanks Bill) and tried some Yttria, which worked about half as well, but didn't degrade after that -- it was rugged and allowed us to get the desired emission at much lower temps than pure tungsten did. Here's a pic of our test rig.

Dual Filament Triode for emission and other testing

We tested this as though it were a normal tube triode (mu about 2, and yes, you could make an oscillator with it) and as an ion gage (it worked fine) but in this context as a way to test emission from various filament configurations and coatings. We used small 6v quartz halogen bulbs we cut off with a diamond wheel in our toolpost grinder, or just crushed with pliers, with a full voltage filament current of around 2 amps. With the yttria coating, we were able to get our target emission with about 3.5 V on the filaments which had them in the low yellow to orange heat. Great! Now to make a diode to drive some lenses and so on for acceleration and focus. Here is the cathode assembly.

Pierce cathode

We found a drill bit that made the angle about right for a Pierce cathode, and did the various machining steps on the lathe in OFHC copper billet. The step is to accept the glass, and the backside is bored to match the little light bulb and position the filament correctly in the hole. We also made the anode of pure copper, with a small exit hole and a bit of shield after that, bored out large, also stepped to fit the glass tubing we used. To get the required close spacing to have decent emission at reasonable voltage, we spaced the two with a washer we punched out of india glass (mica) from McMaster. This gave us the basic diode. Now, copper is kind of hard to machine well, not to mention not being real cheap, so we found some very thick wall bronze pipe (yeah, but it was the only thing we could find the right size) and made the lens elements for an Einzel lens for the rest. This is all shown in this mockup here -- we did adjust the lengths of the glass segments (pyrex) to make it all fit better, using that toolpost grinder we made for the lathe to cut the glass.

E-Gun Mockup on bench

I had thought we would not need a short focal length system here, as it was a long way to the tank, and didn't yet know the math for couloumb (space charge) spreading. I tried this with fairly low voltages generated from CCFL inverters driven off adjustable regulators. About a kV or two for the Einzel lens and a couple hundred volts for the main diode source. Well, we learned some things. For one thing, small stray magnetic fields are a problem if you have low energy charged particles flying long distances -- the earth's field alone were troublesome. Also, space charge spreading was severe, so we wound up shortening the lens element spacings and using more voltage on the lens elements and extraction. The idea was going to be to provide most of the acceleration potential between this whole thing and the tank, using the tank coupling as the last lens element, which did work, but was kind of dodgy to be around when we ran 20kv or so there. All of this was sealed off with Apiezon wax W, which is why we wound up with the copper flashing "corona rings" which aren't that -- they are just heatsinks, and once we made the jig to make one, well....we just made more for all the elements, so it was easy to solder banana jacks to them for playing around easily. We could have sealed this up another way that would have made heat less important, but making it leak free and easy to change were paramount design parameters. What was nice about the large spacings was we could see the beam hitting the glass when we adjusted things wrong, so it was educational if nothing else. But the basic takeaway was, get them going fast, fast -- short spacings, and quickly get to voltage where things like small mag fields and stray charges on the glass wouldn't mess things up in a semi random manner.

The intention here was to make a useful E beam heater, or perhaps a nice focused beam we could scan using a TV deflection yoke. We never took it that far, it was time to build a fusor instead, so we set this aside for now. The main goal was to get that all important gut feel on how this stuff worked, and that goal was accomplished. And of course, we had some fun with it too. I happen to have some neon, so we let a little in the tank to see the beam for example.

Beam in neon

We also used a 1L pyrex flask in good vacuum as a fluorescing target. This worked fairly well, pyrex is moderately conductive so we didn't have too much trouble with charge buildup on it. However, we did see one strange effect with it. After it had been irradiated for awhile, we opened the tank and took it out. After sitting about a day, we saw a fine white powder on it where we'd been hitting it with the beam. After another day or so of sitting, it deliquesced and was wet. Wiping it off left the glass looking fine but -- I'm not using the flask for anything critical again. Edit: found a picture of the target beaker.

Funky beaker after some milliCoulombs of E hits

[johnf]

Doug

I had a look at what you did for the focusing on your e-beam

Einzels work with the gap between electrodes producing the E-field hence the beam bending. All of ours the elements are much closer together ie 0.25" max separation. The element length for each piece of tube can be got by simple drawing it out on paper ( ray tracing). This tells you the position for the three elements and their length, the potential can be got from simulators or by guessing. Most of our einzels will have a few kV less than terminal on the active element typically 0.75 to 2kV less than terminal.


When using ion beams the focus point is near the minimum current draw for the active Einzel element

[lutzhoffman]

Hello: Just looking at the physical size of your chamber brought this old reference for a super RF X-ray setup to mind, from the old Berkeley lab. You would have to shield like crazy, or run like hell. It may be handy in a practical sense for radio-chemistry, cross linking of plastics, etc. Or who knows even termite and pest control, the only downside would be returning home to purple windows, and brown Pyrex. On the positive side you would not even need to run the fridge, no spoilage, and the wife's, or SO's diamonds might look real nice : )

Enjoy.....

[Doug Coulter]

John,
I did this one just to learn, and thought I needed a longer and sloppier focal length than it turned out I did -- I was going for big aperture with the idea that I might scale it to large currents, and it served my purpose -- I learned my lesson (at least some of them, anyway) doing this. It was fun, and now that I have a better feel, yes, I agree with what you say in spades, and it's obvious looking at the E guns in CRTs what it should look like at least for these current levels. I'd love to have more like 100ma of course, for heating with low X rays. What dissapointed me most in this setup was the oddball spreading of focus, which I think I can attribute to one or all of the following -- charge pickup on exposed glass, imperfect machining, and probably contamination on the lens elements (leading to non conductive spots where it matters), which I only found out the importance of later on. I mean, I had like a quarter inch spot size where a modern CRT does quite a lot better at the same speeds and feeds. Good enough for a heater, but I wanted a lot better for other things. Doggone commercial CRTS are hard to replace the cathodes in with the Y2O3 I find works best in systems that get exposed to air, though, or I'd just adapt one -- CRT guns are dumpster stuff all over just now, and I've collected a number of them.

Lutz, yeah, that thing is cool for sure, I'd read that one myself when you sent it to me earlier.

Does anyone with experience on this PHPBB board software know why when you click on a pdf attachment it just wants to save it in firefox, rather than the usual "open or save" dialog I get on everything else on the web? Did I set something up wrong? That's a pain!