old vacuum valves

[randy baron]

In the long (long) road to assemble a vacuum system I'm at the point of looking for at least one valve, maybe more.
This first one will be on the foreline, between turbo pump and rotary vane pump.
I am trying to find economical options, so, ebay etc.
I found what must be a very old vacuum valve, supposedly Leybold but trying google image search I found no similar Leybold valves, but did find some from Balzers.
Anyway, I have no experience buying old valves so to someone who does: do old vacuum valves tend to be good, or a risky proposition?
I know it's a kind of "it depends" question but maybe someone who's bought a lot of old valves might have some knowledge to impart.
I can't test it since it's far away and I doubt the seller can test it either.

It looks like this. I guess it might be a bellows valve based on the 2nd image.

[Shawn Cochran]

I've purchased used MDC valves on 'the auction site' and cleaned them successfully. Mostly manual valves, and mostly needle-style. I got a used gate valve for a project that I replaced the o-ring in and it's just fine too. Personally, I like VAT and MDC - but your mileage may vary.

[randy baron]

Shawn, thanks for your input.
I decided to take the risk and buy it. It was on about $30 and from what I saw other used valves going for that seemed like a good price. Hopefully it will work, actually, hopefully the guy sends it and he isn't a scammer.

I'm still interested in others' thoughts on buying used valves, what to look out for etc. If manual is worth it or should I focus on getting actuated ones (pneumatic or whatever). Typical prices to expect. Brands most likely to have old but working valves. What can be repaired and what not (e.g. flange scratches?, o-rings, etc. but probably not a crack in a bellows valve).
Valve types and their tradeoffs/ad-/disadvantages, etc.

[Ryan Ginter]

I've purchased a total of four or five manual right angle bellows valves from eBay, never had an issue with any of them. Only cleaned the flange faces by hand with acetone, for the insides I just swished some anhydrous alcohol through. All of the valves held under vacuum.

[randy baron]

that sounds promising...

[Dennis P Brown]

For any such valve the bellows is really the key. If they fatigue and have a pin hole leak then they are only junk. O-rings can be replaced through sizing it might be an issue. Air driven valves require electronic control and a source of air pressure. Hence, manual valves are the main ones most here buy. Also, when buying (higher cost units) only buy from someone that permits returns.

The unit in your picture looks clean and if the o-ring is in good shape don't remove it unless it has issues. Use absolute alcohol for cleaning; avoid aggressive solvents that could damage the o-ring.

[randy baron]

The seller said the valve "worked perfectly until it was removed about a year ago" so I'm hopeful.
Unfortunately I can't test it until l get some more KF25 hardware, and even then I can only test it with my diaphragm pump that goes to 80mbar. Rotary pump should be getting up and running "any day now" though.

[Dennis P Brown]

For $30 I think it is well worth trying.

[randy baron]

I just got it today. The valve switch/action feels very nice. I haven't been able to test it otherwise in any way. I wiped it all down with ethanol and paper towel, outside and as far as I could reach in through the flanges, some grime came off but not much.
I am wondering if it turns out I need to change the o-ring if I can do that by pulling it out through the flange opening rather than taking apart the valve (or maybe taking it apart is super easy and there aren't fiddly bits that don't want to go back in when I try to reassemble, like so many devices have).
The only potential problem I see at the moment is that the flanges are a bit scuffed up (there are only two flange/holes but below are multiple pics of each from different angles to help the viewer estimate the scratch sizes):

Not sure how much the scuffs will matter. Those pits (first image) might be a problem. I could probably sand them out or maybe use my diamond sharpening stone (300 grit one side, 1000 grit the other) and manually grind by spinning it as best I can around the flange axis to grind the scratches out (and only make new radial scratches in the process). Or use silicon carbide grits I have for telescope mirror grinding and grind the flanges against a piece of plate glass, progressing from 220, 600, 800, 1200 grit (1200 = ~15 micron). That's probably smooth enough, otherwise I have some Cerox and aluminum oxide polishing powder for literally a mirror finish but I seriously doubt that's required for a foreline pump o-ring.

[JoeBallantyne]

Just test it first before you do anything to the flange faces.

Put a vacuum gauge directly on the input of your mechanical pump, pump it down, and see how low your pump will pull.

Then put the valve between the pump input and the vacuum gauge, open the valve, and see if it will still pull down just as low. It will take longer to get there, as your valve will be outgassing quite a lot especially at first.

If after pumping for a while, you cannot get the pressure down to what it was without the valve, then you might have a leak. The worse the leak, the bigger the difference in ultimate pressure between the two setups.

You can also close the valve, and measure the pressure rise with the pump valved off. Then you can flip the valve around and do the same tests.

If one side is worse than the other, the pressure rise will be faster on that side.

Put whichever face is worse facing the pump input. So that it minimizes the impact on pressure rise in your chamber when you valve off the pump.

My guess is that even though the flange faces look pretty ugly, they may still hold a good vacuum. In which case, don't spend time messing with them.

Some folks here (I think it was Rich Feldman) fixed a badly gouged KF flange face that was leaking by filling the gouge with epoxy and sanding it smooth.

Joe.

[Dennis P Brown]

I agree with Joe. Do not replace that o-ring. Unless it leaks that is risking a lot of problems.

If the flange leaks, dressing its surface is easy; use very fine Emory paper (wet) and dress it on a glass plate backing. Then polishing compound on a cloth lap (glass plate backing again) works well.

[randy baron]

Ok, I'll leave the o-ring be, innocent until proven guilty.
However, I couldn't help myself on trying to smooth the flanges.
I designed a KF25 flange sanding bit for my drill: 3D printed fixture and laser cut pieces of sandpaper to fit it (320, 400, 600 and 800 grit for the moment), sandpaper with tacky adhesive sprayed on the back so it doesn't spin on the plastic fixture.
I was surprised how well it worked. Less than 10 sec. at 320, then about 10-20sec. each of the higher grits, not so much because it needed longer--probably 5sec would have sufficed--but because I struggled to avoid chatter when using the drill. I found I needed to reduce speed and add more hand pressure between the flange and 3d printed sanding piece. I also added a couple of drops of water which helped both the chatter and also the sanding itself. Another thing for "version 2" would be to use a shorter threaded rod (it was just a piece I had lying around).
I put a cork in the valve tube to prevent any grit from migrating into its innards. I really messed up on the 800 grit, deciding to try it dry first and that created some gouges. I wetted it and spent a minute or so to sand those (mostly) out.
I put the flange on my flatbed scanner after each grit to show the progression of before to after.
I see the cross-ways scratch on the 800 grit picture. I'll probably go back and fix that, maybe even do 1000 grit, but I just wanted to show my first attempt at this technique.


Before any sanding:

After 320 grit:

After 400 grit:

After 600 grit:

After 800 grit:

[Richard Hull]

Nice bit of work! You have shared a great save for scratched KF flanges. 3 atta' boys go to you from me.

Richard Hull

[randy baron]

thanks. I can use all the encouragement I can get on this seemingly infinitely long Vacuum Road.

[Dennis P Brown]

Ingenious use of 3-D printing for repair of a flange. If not too costly, I'd think you could even sell such a kit to people on e-bay.

[randy baron]

thanks a lot. I might be able to sell such a kit but I think ultimately it would be at a loss.
I could maybe sell digital plans although the sandpaper would be a little tedious to cut without a laser.

[Oscar Anderson]

As an FYI, the chinese KF25 angle bellow valves on eBay (the ones priced at around $50 - $55) are actually fantastic. Ironically they're some of the better manual valves in our lab.

[Daniel Harrer]

I flattened my flanges with an orbital sander. It is nowadays easy to even get ones with grits like 10,000 and that turns the surface into a quite good mirror.

[randy baron]

That's interesting about the Chinese valves. I guess you mean these

[randy baron]

I can get an old Pfeiffer "EVB 025 PA" KF25 pneumatic corner valve for about $50.
I already have the one manual valve shown in this thread.
For my aluminum coating system I will have a chamber, turbo pump connected directly below it, then the foreline to a Leybold rotary vane pump. I have a foreline trap which I'll put in the line near the Leybold. I think I should put the one valve I have near the exhaust port on the turbo pump.
So would I have any use for this Pfeiffer valve, and is it a good price for it?
I do have compressed air in my shop.
I'm just not sure how many more valves I need, ideally.

[Dennis P Brown]

If you don't have compressed air it is useless. Hence the low price because few people want that type. In any case, using that between the chamber and turbo is a bad idea since it will badly affect pump speed by the turbo - assuming that was your plan. A valve between the turbo and foreline is fine.

[randy baron]

Thanks for the response.
When I wrote "turbo pump connected directly below it" I meant to imply directly with no valve but I see that was a little ambiguous.
So, kind of sounds like I anyway probably don't need it unless I want to have a computer controlled valve between the turbo and foreline rather than my manual one, and at this point I'm not sure why I'd want that.

[Richard Hull]

I have always had a manual bellows valve right at the chamber AFTER the turbo. I use this valve to choke off the turbo as the deuterium gas is let in.
Once down at the bottom pressure with the valve wide open, I close the valve down against the D2 flow so that I don't have to let in so much of my valuable D2 to maintain the fusion gas atmosphere against the ravenous exhausting turbo.

You must have a manual bellows valve between the fusor and the turbo inlet. You must develop a skill to pinch off the turbo against incoming D2 gas flow. It is an art. Not a science.

Richard Hull

[Dennis P Brown]

Yes, Richard is absolutely correct that for a fusor a valve to control deuterium loss is critical to save that gas. But for an aluminizing chamber like they have, maximum pumping speed/efficiency is of importance and having the full throat of the turbo pump against the chamber is best (through adding a screen is very useful in protecting the turbo blades.)