I recently picked up an old argon ion milling unit. Got it for the vacuum system. 6" diff pump with LN2 trap, associated valving, Varian 840 ion gauge, and a couple 10kv power supplies. I got plumbed in this morning. Got it down to about 4.5x10^-6 torr. Not terrible.

But now I have a leak and I am having a heck of a time finding it. Anyone here done this?

I have removed the chamber and eliminated that by using a piece of 1-1/4" thick acrylic over the port to the chamber. I have isolated it to somewhere between the diffpump butterfly valve, roughing pump valve, and the chamber flange. I can pump down the system, close the fore line and roughing valves and the pressure starts rising immediately while the diff pump remains the same pressure.

One problem is all the fittings leading into the chamber flange/butterfly valve housing are threaded ports. 1"NPT for the ion gauge adapter, 1/8" for the thermocouple gauge, and a 1/4" port for the vent. And this was from the factory?!? I can see the remains of red glyptal on the threads. Now everything has teflon. Guess someone else was trying to find the leak too.

I pulled off the valve assembly, the o-ring was pretty flat on the flange between the roughing valve and the chamber flange. Will o-rings under vacuum suddenly go bad? This unit is older than I am so I can believe it.

So what are some simple ways to trace a gross leak like this. I dont have access to a helium leak detector so thats out. I have heard using helium will make the ion gauge jump when it gets into the system from the leak. I have also heard of something about testing for leaks with argon.

Any ideas???

Last question, the diff pump does not say how much oil it needs and the manufacturer is no longer around (Tecnics Inc, T-700). Any ideas how to figure how much oil it should take? The old oil is pretty dirty.

I'm no expert, but I've worked on some vacuum systems and I'm glad to pass on what little I know.

NPT threads are obviously not your friend here. I've had moderately good luck at getting a vacuum seal on them by going with a thicker/heavier grade of teflon tape. The yellow-colored stuff that is for use in residential gas systems seems to work pretty well - Home Despot carries it. If you want to replace the Glyptal with new, the stuff you probably want for vacuum seals is Glyptal 1201.

Deformed/flattened o-rings are a good source of leaks. I'd replace it out of hand, without bothering to try and convince myself that it was the true source of the problem. I don't know if this is compatible with your end-use, but I put a very thin coat of Dow-Corning 111 Silicone grease on replacement o-rings.

Some leaks can be found by pressurizing the system, but pressurizing is the reverse of drawing a vacuum, and if the leak has any compressible material around it (o-ring?) that has gone slightly wrong, pressurizing the system may re-seal it temporarily and mask the leak.

Usually there is a sight glass on the pump case that shows the oil level. On the pumps I've worked with, the oil was supposed to be half-way up the sight glass. The pumps get dirty on the outside because the vacuum oil tends to get on everything - it may have buried the sight glass in a haze of oil and dust.

Dirty pump oil is kind of tough on the pump. You may want to run replace the oil with new, run it for 20 minutes, and then replace it again. Since you mention the pump is no longer being manufactured, I would not use one of the various "vacuum pump flush" solutions you can buy: it might do in your seals and then you'd be in a bind.

I've not heard of an argon-based leak detection method, sorry.

It sounds like a nice pump that needs some care; I hope you can find the problems and get it running well!

Dave

Does it have Viton O rings?, I understand they outgas less than standard O rings. I haven't tried this either but have heard that Loctite blue neutral cure silicone is vacuum compatible (is this RTV?).

I believe that a fresh set of o-rings designed for high-vacuum service are in order.

In grad school, some guys set up a real hard vacuum system (routinely in the 1-5 x 10^-9 Torr). They used all Swagelok VCR fittings, welded to the system. Built in heater for outgassing. Now, I had a moonshine still I was running to get my PhD (studying batch distillation, not supplementing my income with illicit revenue). So I felt like the poor relative next to all that shiny electropolished stainless. Also, I remember that the guys doing the work bought one of the Craftsman tool sets with like 9000 tools and an 8 foot wide by 6 foot tall roll-around.

Anyway, there were no threaded connections, and all seals were metal-to-metal. So flanges had raised rings machined onto the surface (as if a metal o-ring had been placed in a groove), and two flanges were sealed together with an aluminum (or copper, or stainless, or nickel, perhaps?) crush gasket between them. This is the same principle as BMW power-steering fittings, and VCR fittings.

For retrofit, I'd definitely try Teflon(R) PTFE tape and new o-rings. If you still get leaks, you might try putting blanks in flange locations and watching how the blanks change pressure transients to run down the offending leak location.

Or, pressurize the system with 30psi (be very careful here!) helium and use a mass spec leak detector (you got one of those, right?)

Best,

Jim

The Diff pump oil quantity can be determined with the pump dismounted and the tree pulled out so you can see down to the bottom of the can. The bottom needs to be covered under all conditions, including the quantity of Diff oil that is in the "Vapor, condensation, run down the tree and walls cycle". You can guess, or actyually wet all surfaces when recharging the unit with diff oil. Some pumps have flat bottoms. they require less than 10 mm to cover the bottoms (assuming the system is not sitting on some off level floor ;-). Other pumps have ribs to increase heat transfer area. These must be covered at all times during operatiuon. If the oil level should drop too low, the heaters can burn out, or burn the remaining oil. That said, The diff pump oil volume is one of the limits of absolute pressure attained. Volitile fractions and high vapor pressure contamination are proprtional to the amount of diff pump oil present, so you want to use the leasy amount that is "safe". Oh! and diff pump oil is $$$$.

Gross leaks can sometimes be located with acetone. Swabing or squirting the suspected leak area with acetone will often show a "blip" in the convection gauge display of a pumping system (not static and valved off). First the vacuum level will blip up as the acetone vapor enters the system, then the vacuum will increase as the acetone freezes in the leak, msaking a temporary seal. Finally, the pressure will return to the beginning display as the acetone is pumped away. (No acetone near ferro fluid feed throughs! they will be ruined)

Oh, I have NEVER been happy with just giving taper thread fittings a bit more tightening to make a leak go away. The fitting will split if it is not so already.
Take all the fittings apart, clean with brushes and acetone, and then reseal with tape or your choice of glyptal, torr seal etc. I like teflon tape well enough. Easy to clean off when needed

If you find a leak by any method, and just want to seal it up, brush on some nail polish or varnish. If you have any kind of pump at all, low -7s will be possible.

I also have a vacuum system, though turbo molecular pumped, and fitted with an RGA, which works well for leak testing with any gas ;-)

CalG

If you don't already know this, 10-6 torr is also the range of virtual leaks. Once you eliminate all of the the leaks from the outside, you need to worry about virtual leaks. Even a fingerprint can outgas for days. Any built up crud will outgas forever. If you have not already done so, I would take the system completely apart, clean it mechanically (scotchbrite) and then chemically (acetone). After it is clean, handle the parts with gloved hands. Replace all of the o-rings with viton, lightly greased with high vacuum grease (very low outgassing). This should include the small o-rings in the butterfly shaft as well as the large one. I would also replace the oil in both pumps, after cleaning the stack and the inside of the diff pump. If the cooling jacket of the diff pump is clear (open), now is the time to clean the inside of the tubing, as it will eventually clog. I used to do this with small container of dilute HCl and a recirculating pump. This will dissolve the lime that will otherwise clog up and kill the cooling coils. If you don't do this regularly, the coils will stop off when the pump is operating (assuming you use it frequently).

4 X 6 is not bad. What is the oil in the diff pump?

If it is hydrocarbon oil you will need an LN2 trap to
go below that. If it is DC704 or DC706 you can get away
without it. Do you know the oil in the diff pump?
How fast does the blank-off pressure rise? How far does
it go - that is, where does it assmptote to?
What diameter is the diff pump?

How to leak check without a leak checker.

1) get a leak checker.

2) watch the highvac gage (what kind is this, cold
cathode or an ion gage?) and carefully squirt acetone or
alcohol on the suspect fitting. Watch for a *change*
in the pressure. If it's a small leak the solvent will
get pulled into the leak and freeze in there, lowering
the pressure. If large it will be sucked in and vaporize
in the system, raising the pressure.

3) a more sensitive test can be rigged up with a short
section of glass tubing. Put the tubing in between the
foreline and the rough pump, or better yet in series with
the roughing line only, so you can do the test with the
diff pump cold and out of the system.

You need a sparker coil (tesla coil) to excite the
gas inside the glass tubing. It will glow blue-ish
depending on the pressure, but when acetone gets into
the system via your leak, then it will turn pink-ish.

1) you can't find your leak by pressurizing anything.
Great for inner tubes but DOA here.

2) there's an even-odds chance you are seeing outgassing,
depending on your rate of rise.

3) O-rings do degrade over time. Enough to leak.
There's a lot of ozone around systems like this.

4) you can get mass spec helium leak tight joints with
NPT fittings. Just have to be careful.

5) glyptal is a *bad* sign.

6) do NOT use any epoxy as a thread sealant. NFG for that.

If you can beg borrow or steal a real mass spec helium
leak checker, you can proof the entire system out
100% within about an hour.

OK, going back, a six inch pump, varian ion gage.
BTW do NOT use acrylic for the blank off on the port!!!

You can find unusual large virtual leaks associated with
materials like that. Almost like there is gas trapped
inside the material, diffusing out. Change to an aluminum
or steel plate.

Jim

I'm with CalG on the acetone. I've used that to good effect when nothing else was available. IMO, nothing beats a helium leak detector or general purpose RGA. As said above, Glyptol is a bad sign. It takes good fittings and an artist to get reliable results with Teflon tape. If you want things to be permanent, Loctite 271 works well and doesn't seem to contaminate anything. I've built a lot of commercial systems with that. You'd need a lot of heat to break it free though, so not for use with stuff you have to get apart. Everything needs to be clean and o-rings should have only the slightest shine of a good vac grease like Apeizon M or N (expensive!). Dry o-rings don't seal.

No clue what the oil in the diff pump is. Probably old hydrocarbon. Has a light yellow color probably from age though. To get the tree out you have to take off the bottom of the pump. The spiral baffles are permanently mounted on the topside.

When I close the butterfly it goes from ^-5 to ^-3 in seconds.

I think the glyptal is from the original assembly of the system.

If your hair does not curl up when you whiff the
pump inlet, it's probably OK. Don't pull the tree out
unless it's badly coked up at the top.

(never, ever, ever, stick you nose down the inlet of
vacuum equipment and take a big sniff, you probably know
that)

That's a pretty fast rate of rise. If you can locate a
tesla coil sparker the glass tube trick actually works.
An old ford coil would do the job too.

Jim

I actually have one of those handheld units. I'll have to borrow it from my dad. I lent it to him for checking for leaks in plastic welds.

What about connecting the hv to the grid of the ion gauge? It's in the zone of the leak.

Looks like I wil take an inventory if o rings today and get them ordered up. I am thinking viton instead of nitrile?

Also thinkmk will go ahead and pour the diff pump oil into a clean container and filter it out at least and rinse out the diff pump. There are a lot of particulates floating around.

Thanks for the help guys.

I know discharge gages do pump, so if you have one of those do turn it off.

If you are looking for base pressure rise using the ion gage, of course you leave
that on. Those do not pump the way ion gages do.

There's a good book about vacuum technology, by a guy named O'Hanlon. Lots
of practical stuff.

Jim

Where is a good place to get o-rings?

Even a tight system will show a pressure rise when you turn the pump off unless it is very thoroughly outgassed. A discharge gauge is not going to pump enough to show up in a system that size. They are used on large radio tubes like monster klystrons because they are both maintaining the vacuum and testing the quality of it at the same time, but these are tubes that have been pumped down and baked for a long time to get rid of almost all of the adsorbed gases. The discharge gauge just acts like the getter in a smaller tube. If you really want to check the system after you get it back together, a mass spectrometer leak detector is the way to go. I had one of my liquid helium transfer lines die, probably because I didn't get a good joint and thermal shock opened up a tiny leak. I pumped it down with my old Veeco MS-9 leak detector and got down to -6 territory, even with the leak being pumped by the rinky dink pump in the MS-9. Then I blew a little helium on one end of the line and the mass spec meter slammed on the peg. If you can't find a leak with a mass spec detector, it aint leaking.

Bill

Absolutely true that the mass spec helium leak checker is the way to go. He does
not have access to one, unfortunately.



And, I apologize for any confustion, his system does NOT have a discharge gage.
I only mentioned that on the off chance he might. I was told by the Veeco
experts to always disconnect the discharge gage on those machines, when
leak checking the leak checker itself, for a variety of reasons.

I have indeed had varian discharge gages that were tough to identify as
leakers, when they were engergized. Pull the lead off, and it would show up
clear.

Jim

There is a chance I will be able to get access to a helium leak checker. He does service work on Lasik machines and they have a small portable unit.

I went ahead and pulled everything apart short of the diff pump today. Orings are cheap so I made a list and going to order them. Is viton worth the expense?

As for diff pump oil i think I will get some of the DC704 knock-off. ~$60 for 500cc. Most 6" pumps seem to take somewhere between 400 to 500cc it seems.

One option I have available is since I have the top valve assebly appart I can go ahead and tig on new VCR style fittings. Worth the trouble? I can easily put on KF16 flanges on the vent and gauge connections and for the ion gauge I would have to get a stainless version of the 1" ultra-torr fitting on there now.

Dow corning high vacuum grease [silicon], change your o rings [on our kit theres a collection of vac clamp fittings as oposed to threads]if any, a scratch is all you need to drop vac
mark

As to Viton - yes it is worth the money for vacuum applications.

Probably not. Likely a vacuum oil. Dow Silicon-based, or Santovac, Fomblin, or Octoil(sp?). In nasty chemical service (metal fluoride gases), we used Fomblin.

For the seals, you can use Dow-Corning, Dupont Krytox, or Apiezon grease.

Jim

I take it you are aware of these folks:

http://vacuumshopper.stores.yahoo.net/

and also:

http://www.duniway.com/

What kind of cooling does this diff pump have? Air or
water? It would not hurt to change the diff pump oil
but be *sure* to measure the amount you take out. They
often need less than you think should be in there.

My theory would be to find the leaks before you go into
the diff pump, it seems to be working tolerably well
right now. If you can find a manual for it they will
tell you the oil capacity. But do save and measure the
oil you dump out.

Jim