View Full Version : High Vacuum Pumps

01-05-2013, 02:41 AM
Has anybody ever heard of building your own high vacuum pump (I am thinking below one torr, probably rotary vane?

01-05-2013, 03:14 AM
Titanium sublimation pump.

Oh, I'm sorry, when you said "high vacuum", I thought you meant 10^-8 Torr.

01-05-2013, 07:32 PM
No, that not that high. That sounds like I would never be able to get the cutting oil to stop outgassing!

01-05-2013, 08:04 PM
You can get down to ~100 milliTorr with a single stage rotary vane oil filled roughing pump. You can get down to about ~10 milliTorr with a two stage rotary vane oil filled roughing pump.

What are your goals for this pump? How far below one Torr do you want to go?

I've rebuilt enough roughing pumps to know I wouldn't want to try machining something like that. You can buy a brand new single stage RV pump from Horror Freight for less than $100. You can get some very nice used two stage RV pumps from your local university surplus auctions (that's where I get mine). I could not possibly machine something like that for as little as I have paid for mine.

What do you plan to do with it? If you are experimenting with vacuum high voltage glow discharges I can tell you from experience that a single stage RV pump will get you down into the range for at least basic experimenting (you can get glow discharges at those pressures).

01-05-2013, 08:57 PM
How about a couple Refrigeration compressors connected in series?

01-05-2013, 09:41 PM
Under 1 torr is not considered high vacuum. My systems will get down the the .00000001 to .000000001 torr range (10^-8 to 10^-9.

What exactly do you want to do? You can get cheap refrigeration vacuum pumps at harbor freight. Craigslist is also a good place to pick up used pumps.

No, you do not want to build your own pump. Rotary pumps are actually pretty high precision pumps and the tolerances are quite tight. You would need to have a grinding setup to build one.

01-05-2013, 11:26 PM
This is interesting. I'm wondering how good a vacuum can be obtained using oil as one of the functioning elements, whether as a lube and seal or whether the height of oil in a tube acts as a piston-

Does it take a special oil to not vaporize under high vacuum and prevent a complete vacuum from being generated?

Something else that would be interesting to know is how high a vacuum is required for various things. Some see a vacuum as being a space where most of the air has been evacuated (pretty vague, and not really a vacuum)- others see it as .00000001 torr (whatever that is- it's just a very small number at this point). Is there a chart that would illuminate the range and show the usefulness of it at various points?

We could start with sea level air pressure, then go down into the earth for some distance, then look at it from a height above earth viewpoint, then get into the very low pressures where other things become possible, like vacuum tube operation, metal vapor deposition, etc. I think that would be interesting, and probably show it in such a way that it becomes very meaningful to a lay person.

01-06-2013, 12:26 AM
This is interesting. I'm wondering how good a vacuum can be obtained using oil as one of the functioning elements, whether as a lube and seal or whether the height of oil in a tube acts as a piston-

see "diffusion pump"

Does it take a special oil to not vaporize under high vacuum and prevent a complete vacuum from being generated?


Something else that would be interesting to know is how high a vacuum is required for various things.

That depends a lot on what you want the vacuum to achieve.
Usually, you talk about how far each molecule of gas travels before it hits another molecule or some other object. There's a chart on the "mean free path" article on wiki. So these "low" vacuums are used for things like glow discharge, controlled thermal conduction, etc where you expect the gas molecule to bump into something else from point a to b. You can also calculate how many molecules hit a surface area per second, for stuff like oxidation, etc. Higher vacuums are usually used to prevent surface contamination, like in scientific experiments of surfaces, particle accelerator experiments where particles whiz around long distances and had better not hit anything along the way, or whatever other process where you need an absolutely clean environment.

01-06-2013, 12:35 AM
I am definitely thinking about the 'nice rotary vane pump' level of vacuum. Glad to know that this isn't really doable. (although I also wonder about rebuilding of possibly cheaper clapped out pumps...)

Where are the ground surfaces? I imagine that the smoothness of the cylindrical wall and the flatness/right-sized-ness of the end plates and ends of the rotor would be.

01-06-2013, 12:57 AM
Hi T_M,

It is possible for an amateur to rebuild a RV pump though the first time may be quite difficult. First you need to obtain a pump then you have to obtain a rebuild kit for that particular model of pump. If I were you I would avoid any pump that is "totally clapped out". At the vary least you want one that has a working motor and isn't really corroded on the inside. Pump motors go bad when people stop taking care of their pumps and the internals seize. Avoid any pump which has been used to pump on anything containing water. Water and oil filled roughing pumps don't get along.

Safety first (or second): Researchers use oil filled roughing pumps to pump on all sorts of nasty chemicals. Wear rubber or nitrile gloves if you do decide to actually rebuild, or even change the oil in a used pump.

If you get to the point where you actually need to buy a rebuild kit google "Capitol Vacuum" and "Precision Plus". A rebuild kit costs ~$75 to ~$250 depending on the pump and weather the kit includes replacement vanes.

Now, what are you planning on using the pump for?

01-06-2013, 05:59 AM
A variety of things actually, including some glowy/ dischargey stuff, perhaps even sputtering...

01-06-2013, 07:04 AM
Oil in rotary pumps is used as a sealing and lubricating element. It is typically a high grade mineral oil which does have a very low vapor pressure, somewhere around 10-7 to 10^-8. You can even use rotary pump oil in diffusion pumps for medium vacuum applications. Though the problem with that is the oil will carbonize when exposed to atmosphere when hot and make a big mess. I use silicone ester based oils in mine.

High vacuums are needed for things like coating. Typically for evaporative and e-beam coating you want pressures below 10^-6 otherwise you can trap gas molecules under the film being deposited. Sputtering is done at a much higher pressure but you still need to initially get down to a high level of vacuum to make sure the chamber only contains what you want in it. Basic sputter coaters for SEMs dont bother, they are just a mechanical pump with a simple DC magnetron sputter head which is fine for depositing non-reactive gold on samples.

I have rebuilt my own pumps. Direct drives are easy, you can buy a kit from Duniway Stockroom. Belt driven pumps like the old welch pumps are a lot more difficult. The assembly is very critical and there are usually fixtures involved to get it right.

Pretty much everything in a good quality pump is ground.

I use a big 6" diffusion pump on the sputtering system I am putting together. On my little mass spec system I have a small turbo pump.

Vacuum stuff is stupid expensive. Almost everything is stainless, tig welded, and helium leak checked. You might look over on the fusor.net forum for some FAQs about Vacuum systems. There are several high schoolers over there working on putting together fusors for science projects. A fusor is an electostatic fusion device invented by Farnsworth (The TV guy). They use high voltage to accelerate deuterium and cause it to fuse releasing neutrons. They are good for doing neutron activation of materials for nuclear research, but that is about it.

Lew Hartswick
01-06-2013, 10:05 AM
If anyone wants to build a "vacuum" pump better than a Hoover carpet cleaner
and asks here how to do it, then they CAN'T. :-)
(as someone who has overhauled Welch pumps in the Chem. dept.)

Jon Heron
01-06-2013, 10:42 AM
I have a couple large rotary vane pumps I use to dewater used vegetable oil and to distill the methanol out of the biodiesel/glycerin in my biodiesel plant. My new pump is a Busch 1.5hp with an 1.5" input connector :p It gets my 500 gallon processor and 100 gallon recovery tank down to 29Hg in no time!
Both my pumps came from commercial sausage meat machines where they use vacuum to move the meat around.
I found the secret to making them last is to use an air filter on the input and keep the oil dry and change it frequently.
The actual rotary vane portion is pretty strait forward using aluminum vanes that rotate through an oil bath in the pump body. The complicated bit appears to be the oil separator assembly and placement of the many oil lines...
Here is a picture of my first pump (it needs to be rebuilt now), you can see the cylinder that the aluminum vanes rotate in (between the pulley on the right and the oil separator box on the left) and the oil bath is on the bottom, not much to it, not sure how the magic happens in the separator box as I have not had to take that apart yet... The shaft runs on bushings with packings instead of ball bearings.

01-06-2013, 02:20 PM
Don't let anyone tell you that you cant, of course you can.
That is the sort of thing that Home Shop Machinists can excel at.
I've also rebuilt pumps along with a couple of turbomolecular pumps.
However, if I were you, I'd rather rebuild a used pump and end up with a really nice unit for far less effort.
Vacuum can be rather expensive, but, many of the components can actually increase in value.
Last year I sold a Bell Jar for many times what I paid for it and the buyer was tickled pink. Plumbing parts can also return good value.

Tom M.

George Schmermund
03-31-2015, 08:11 PM
If anyone is interested in building homemade vacuum tubes it is possible to construct a working triode tube with only a good 2 stage rotary mechanical pump. The tube's performance will be on the order of what de Forest was able to do in the early 1900's. The glass parts can be made demountable for easy access to the filament for replacement.

If you want a sealed tube you can use something like Hysol 1C epoxy or JB Weld. A tungsten filament won't last more than a few hours at the pressures achievable with a mechanical pump, but neither did the early Audion tubes. Any crystal radio set can be converted to a vacuum tube set by just replacing the crystal detector with a DIY triode as the detector.

If you get caught up in making vacuum tubes you'll want to get something like a diffusion pump for achieving a much higher vacuum than can be produced by just a mechanical pump. I have a collection of homemade triodes that were built in 1994-95 that were sealed with epoxy and they still work!