Science/Engineering/Technical forums

As some here know, I've been dealing with a tiny hard to find leak for a couple of months now.

No more, gotcha.

A pinhole had opened in a solder joint between some 1/8" Cu tubing and a piece of capillary tube soldered into it. What made this devilishly hard to find was that, since that point was run at 2kv as part of the ion source ion extraction, I'd slipped a piece of silicone tubing over it as an insulator. So, running around with a hose spraying He didn't find it easy -- it was too tiny a leak to draw in He along the silicone tubing right away. I''d run around back there, and see nothing, then go somewhere else, see something minutes later -- He finally finding its way down that channel and through the cap tubing....around and around we go.

So, DON'T DO THAT.

Further, the big saftey/emi shield screen I had back there made it real hard to get to. The way I'd done it (hose clamp around the flange with a lot of cutouts) made that so hard to remove and replace I didn't get too up close and personal when it was time to look for the leak, until in desperation I finally ripped it off (it takes two talented people to put back on).

So, DON'T DO THAT EITHER.

OK, so I've fixed the leak and now I have super nice base pressure. See pic! Since that made it purty for the mass spec, I made a couple special runs at very slow speed (1 sec/amu) while hard-baking and after. My hard baking is 2kw worth of quartz heaters at full power, which make enough UV to break up water, and make everything in there very hot very fast. I probably pused a little too hard here and saw some lines from things in the tank decomposing (even lines from the solder on my LV instrument feedthroughs). Maybe a little oil from when I had that motorized throttle valve in there too, who knows. Then I let it cool off and took another spectrum. Now, one last problem to solve. How to have EMI shield that I can put on by myself. So I made this thing, which was more work than it looks like. I had to heat-stretch a piece of 8" PVC from the junkyard to fit an 8" flange (the pipe was about 1/8" less ID) and final sand to a snug fit. I drilled and tapped the flange for a 10-32 bolt to hold it on, and re-jiggered the inside ballast resistor insulator and air channel for it all to fit together. The result was worth it, however, as now I have zero exposed HV wiring and no EMI to speak of at all anymore. The pure copper screen and good grounding scheme sees to that one. The system was back to 1e-7 mbar by the time the turbo spun back up from standby speed (normally I do runs with it at 29% of full speed to make gas control easier).

Good to hear that you finally found the leak! That's a pretty nice base pressure that you have now too.

Yeah, a big whew on that leak. I doubt I'll see that nice base pressure number very often, but it's nice to know it's possible. More normal is something e-8 unless the system has been pumping for days. I temporarily took the microwave magnetron tube off the thing, as I have that "other ionizer" in the 2nd grid that works well because I have that outsize tank.

Did a nice (but not record setting) run after all this, didn't report it as I didn't do my good data aq on it. Got indium really hot for once, though, in a higher than normal ratio to the silver, as the second part of the run wasn't as hot as the first part and silver is kind of a "right now" thing, where In's longer half life integrates better. The really sweet thing on that was that even though I was fooling with the boundaries of stability, which tends to create a lot of EMI as the current comes and goes, I had nada -- that shield really works. To even see the EMI I had to put a sensor in there (antenna).

Also got at least a temp fix on the HV FT issues, we'll have to run like this some more to see if the ad-hoc design change I made is the real McCoy or not. And, time to hook up the active AC drive on #2 grid again and work that path some more, it's where the records got set, and things are for once, all working at the same time. Heck, I never even finished the conditioning of that graphite/ceramic vane grid and never got comparative numbers out of it -- lots to do. In fact, there are (at least) two issues there alone. Since that grid is also far more accurate than any other I've made -- the effect of that, sans ceramic, needs tested. And, it's a different OD...lots of variables here, and much data needed to sort out what helps and what doesn't.

I found some tiny solenoid valves cheap at Burdin's surplus center (read the whole page, this is just one possible candidate), hope to have them in a day or so. If they check out, I will use one as the gas inlet instead of running it through the uWave source, and make it at least possible for automation to hold an operating point with closed-loop feedback. At least they are cheap! I will put the uWave source on another fusor that can't take advantage of huge path lengths in an oversize tank for ionization -- that thing is good, but needs more work on the extractor. Having the gas go through it "seemed like a good idea at the time" but has since proven more hassle than it's worth. Without that constraint I should be able to get far better ion extraction from it -- I can put the pusher electrode right up almost into the glow, and ditto the puller, which I think needs to be a two element thing. Still working on how I'd do that and get it hooked up without wires inside the tank where they get in trouble from charged particle bombardment and such. Maybe MarkB can make me some special quartz with side-arm connections for that.