by Doug Coulter » Mon Feb 24, 2014 5:45 pm
Nice looking stuff! I'd definately replace that oil diffusion pump with a turbo, a turbo-drag if possible, and get rid of one of the forepumps too - I kind of doubt you need both with a modern turbo. For reference, my own big tank only "needs" 512 l/s to be fine, even that is probably excessive once you get way down there. The issue is that in molecular flow, vs viscous flow, the O rings leak more than happens to find its way into the pump intake by random collisions with the walls, without getting up to viscous flow pressures, and at that point, only things like sublimation pumps and so on have much effect - you just need more square feet of absorbtion ability.
Viton leaks water, but not air. Nitrile the other way around. Why someone hasn't yet invented a dual-seal approach using both, I dunno - I might have to, or you.
If something's been up to STP/shop air conditions for a long time, it takes awhile to get back down there - it looks like that's your situation - tight, but wet (viton, perchance, too?)
In fact, earlier this month, during the very worst solar conditions ever, I turned off my turbo, which I rarely do, for a few days, and even with the forepump valve closed, and mostly CF (copper) but some viton (door and a couple tubing couplers) it went up to e-1 mbar in 3 days, and a week+ later (and a very moderate hour bake) - still isn't back to its normal base pressure, about a factor of two high.
Didn't check - my Pfeiffer mass spec is offline, but you just know it's all water, since the only possible intake is through viton. I can let it up to shop air for an hour and it not take anywhere near this long to get back to base pressure. I can get back on the same day if I take it up to STP using argon or nitrogen instead of shop air, and don't open the door much while I replace a grid or something.
That's one nice transformer too. I re-wound a harbor frieght spot welding transformer with some taps on the secondary for use here, with a variac and a solid state relay - it's a working thing I can count on. I noticed in my own evap work that most of what you might evap - I messed with Al, Ti, Pd and Cu - is a decent "getter" itself. It's why some pro lashups have a "shutter" over the source. You preheat the stuff you're going to evap, maybe even evap a little - and it eats all the contaminants itself - only then do you do the deposition, after the evap source itself has cleaned things up. Kind of like a sublimation pump for free. Al and Ti are particularly effective at this, and especially with water. I accidentally coated much of my tank innards with Al on my first try at all that - and I believe it is better than stainless at not having water stick back onto it later on. It ruined all my insulators, of course - I had to do a teardown and clean them all again - but it went back to base pressure a lot faster....hmmm. I may have to do a more complete job of that coating.
I'm surprised maglev needs a backup power source. My Pfeiffer thingie makes enough back emf to run the controller for 15 min or more in a power failure, if there's no leak. It's almost freaky how low the ratio of friction to stored energy in that rotor is. It only draws about 35-40W to run the entire mess at base pressure.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.