My sputtering system

[Joe Jarski]

The new PVD chamber is still a work in progress, but so far things are looking pretty good... I think. Before I continued with the chemical cleaning I wanted to feel somewhat warm and fuzzy that I didn't have any obvious leaks at this point (there's almost 18 feet of weld joints). I'm still acquiring some of the valves so in the mean time I made a temporary adapter for the turbopump just to see where I am with it. All of the flanges are sealed with O-rings - the cheap ones for now, but they'll be replaced with the good stuff soon. The base pressure is in the low -4 Torr range right now so I'm thinking that it's sealed up pretty well and I'm just dealing with outgassing and not real leaks. Proper cleaning and bakeout will hopefully get the base pressure well into the high vacuum range so that I can try evaporation while I work on the new sputtering guns.

[Joe Jarski]

I spent some time with my poor man's helium leak detector (helium & ion gauge) and managed to find a leak in my chamber. I had it pumped down to 7.0e-5 Torr according to my ion gauge and started going over all of the joints & weld seams with a hypodermic needle and helium tank. I found one spot that dropped the gauge pressure reading to 6.4e-5 Torr. It was difficult to pick up on the changes at higher pressure and I was looking for a more significant change in the gauge reading. Now that I have the technique - it's a great way to search out any leaks.

[Bill Fain]

Joe, Hi. Nice looking set up! Are you by chance on the Vacuum-X forum on Yahoo? -bill

[Doug Coulter]

Yes, really nice looking! Another trick is plain old water in a spray bottle. It''ll make a pretty big swing on a pressure gage --first lower, then higher as it gets in and evaporates in the tank. Good for the larger leaks. Of course, when the leaks are so large you are limited to about a millibar, you just go around with a rubber hose stethoscope and find those by ear -- they're easy. It's all the tiny ones that put the gray hairs on your head (or make them fall out).

[Joe Jarski]

Thanks for the compliments! After trying He without finding any leaks I decided to spray it down with liquid and that's how I first found it. I then went back to try and find out why the He didn't work. It happened to be at the lower pressure then (e-5 Torr range) and showed the leak. The next day I went back and tried again and it still wouldn't show anything in the e-4 Torr range. I narrowed the leak down to about a 1 inch long area, cut out the weld and re welded it. When I shut it down for the night it was at 4e-6 Torr and still dropping.

The one thing I didn't like about liquid is that it seemed to have a delayed reaction and took a long time, like a half hour, to return to the original reading and stabilize. Using He it would take about 5 seconds which made it a lot easier to pinpoint. On the other hand, there was no doubt that I had leak once it was sprayed down.

Bill, no I'm not on that forum - I'll have to get over there and check it out.

[Doug Coulter]

Right, the liquid is very slow. Of course, when I find the leak, I do one of two things -- open it up and re-weld, or change liquid to shellac or glyptal and fix the leak! Glyptal is best for that, as it likes heat and thermosets for a pretty decent permanent fix. Shellac does too (the good "blond" stuff anyway) but won't take the higher temperatures as well...Either will just get sucked into a small leak and get hard before it makes it to inside the tank. Other sealants that work well would include most epoxies -- but if and only if they are mixed to a level of "thorough" that most people can't imagine, else you have this high vapor pressure resin or hardener left and that takes forever to finally get gone. Hysol 1-C seems about the best epoxy for vac systems, the solids are silicon dioxide in that, but some like JB weld too. The longer cure time ones seem better than the 5 min variety, probably because it takes >5 minutes of stirring to get epoxy mixed well enough.
(and you use from the middle of the mix puddle to avoid the poorly mixed edges).

I get hysol at mcmaster, glyptal at Caswell plating (its the real thing!), and shellac in flake form from woodworking supply houses. Zinsner "Bullseye" shellac from the hardware store is the good stuff too. I've used the flake shellac to mount windows over a hole in metal. Heat the window and metal up first, put on the shellac, lay on the window, heat further, then when it cools off, it's good and has just enough give to appease the tempco monster. It's permanent -- you can't re melt it and remove the window, as it thermo-sets.

[William A Washburn]

Before I retired I worked for a company that had two semi-local plants. One is here in Fresno, CA
and the other is in Salem, OR. I managed the systems stuff at both locations so traveled to Salem
2-3 time a year.

Their mission was coating the plain glass we sent to them. They had a huge vacuum system with
which I had some familiarity. The vacuum containment consisted of a multiple rectangular x-sections
that would take a 72in x 120in sheet of glass, and move it horizontally from section to section sputtering
silver, gold, titanium and other stuff one layer at a time onto the glass sheet. The only thing I didn't
read in this current "Coulters blog" was that we use an extremely intense magnetic field which causes
the ions to form minute closely spaced spirals that bunch together and lock horizontally.

This was explained to me once by one of our research guys but I didn't get much out of it other than
without the magnetic field we couldn't build the tiny vertical towers of ion spirals we need to make
a product that works and is tough...Bill

[Jerry]

Bill, that sound like a pretty typical magnetron sputtering system.