It seems like creating a vacuum should be a pretty easy job, but it turns out that sucking all the air out of something is harder than it seems. A cheap vacuum pump will get you part of the way there, but to really pull a hard vacuum, you need an oil diffusion pump that costs multiple tens of thousands of dollars.
Or, you need a bunch of thrift store junk, a TIG welder, and a can of WD-40. At least that’s what [Lucas] put into his homebrew oil diffusion pump. The idea of such a contraption is to vaporize oil in a chamber such that the oil droplets entrain any remaining gas molecules toward an exhaust port. His low-budget realization of this principle involved a lot of thrift store stainless steel cookware, welded together with varying degrees of success, with liberal applications of epoxy to seal up any leaks. And an electric smores cooker for the heating element, which was a nice touch. The low-budget approach extended even to the oil for the pump; rather than shelling out for expensive specialty oil, [Lucas] distilled some from a WD-40 silicone spray lubricant.
The video below details all the travails [Lucas] encountered along the way, plus the testing process. The results were at least encouraging — the diffusion pump was pulling vacuum far in excess of what the roughing pump was capable of. He clearly still has some work to do, but getting as far as he did with the scrap heap of parts he cobbled together is pretty impressive.
[Lucas]’ goal with all this? A fusion reactor. No, not that kind. This kind.
Thanks to [fruitbaticus] and [supereggbert] for the near-simultaneous tips on this one.
Thats.. umm.. yeah. Ill be surprised if he can get into the -4 range with that. And he wont be able to tell with that vacuum gage anyway.
“you need an oil diffusion pump that costs multiple tens of thousands of dollars.”
Unless one is buying a very very big one, then they tend to be cheaper than this. As in only a few thousand dollars, sometimes towards the lower end of that scale. And the oil isn’t wildly expensive either so distilling one’s own is rarely worth the effort if one values one’s own time.
To be fair, for 10 grand, one can by a turbomolecular pump, and with such one don’t have to deal with the oil vapors filling one’s chamber.
Since yes, oil diffusion pumps do spit out quite a bit of oil. If one’s application can handle a bit of an oil coating, then no problem. If it can’t, then a diffusion pump is a fool’s errand to use in the first place.
And no, water cooled baffles don’t stop the vapors, only slows them down. Though, some applications will constantly purge the chamber with argon or other gases, typically seen in plasma etching and various depositioning techniques. This purging gas will also help to keep the vapors even more at bay.
If one only cares about partial pressure of the various gases contained in one’s chamber, then yes a diffusion pump can go to practically perfect vacuum. If one cares about the actual pressure, one won’t be going bellow the vapor pressure of the oil.
Meanwhile, turbomolecular pumps don’t spit out even near as much oil, but these fast spinning devices have their own hassles. Like literally exploding if one gets a sudden large enough leak, or if one starts venting the chamber before the pump has slowed down sufficiently.
Have had pretty good luck with £500 turbo pumps from eBay (which work purchased as backups). Bit noisy compared to new ones but worked fine.
Used oil diffusion pumps aren’t too expensive either and much more likely to be OK than a turbo.
A good 2 stage oil pump can trundle down into the -4 range. But I don’t like them at all. Much better to stick with scroll pumps and turbos if at all possible. Or for the ultra spendy – cryo pumps.
For home projects, ion pumps seem to be feasible.
https://www.youtube.com/watch?v=_snzYepQTjI
Oil diffusion pumps are more like $250. But if you want to pump down quickly I would shop for a used turbo-molecular pump.
I’m pretty sure his thrift store expenses plus argon and rod exceeds the cost of an eBay pump. Now he needs a Cold Trap and a McLeod and ion gauges. Varian has a great high vacuum epoxy.
TBH this feels more like a Journey Not Destination project: exploring the operating principles by Making One rather than chasing cheapness
Am I the only one who feels like this video is one step away from Jim Gaffigan trying to make an oil diffusion pump?
Personally I quite like the Mercury drip style vacuum pump. In theory it could be made with a pump instead of gravity and achieve a fairly hard vacuum. It operates using a slightly different setup but the principal is more or less the same.
The Babo pump is a continuously cycling version. https://avs.scitation.org/doi/10.1116/1.1316999
Distilling your own diff pump oil is mad. The performance of your pump is really dependent on it’s quality and you can buy it for not crazy money unless you go for the really fancy stuff. There are absolutely loads of cheap diff pumps knocking around, sometimes for free, still dating from the point that turbos got cheap and everyone who could binned them. So I guess a fun project but not one to go for if you want a good quality diff pump rather than the experience of building a diff pump.
Got a good source for oil?
I don’t particularly like the idea of spending more on oil then I did for the pump
LIked the idea, but that was just too painful to watch, I could have done without all the overly loud audio effects.
Heck with fusion, make 1L6’s and Nixies!
Hello Dan
This is a good craftsmanship and I am thrilled that you have made from the remnants of your household a diffusion pump working in the basic principles! All comments about vacuum measurement and testing your results are not important, the pump works! Of course you can achieve even better results with special oil and baffle but I don’t think that’s what you’re wanted! The information that turbomulecular pumps are used today is only partially correct, there are still many processes where oil diffusion pumps are used! And this will remain so for a long time, especially in pumping speeds > 5000 l/s!
In high vacuum applications with a strong magnetic field (fusion reactors, Super Conducter) high vacuum pumps with moving elements (parts like rotors) do not work anymore! So even in high end processes the diffusion pump will still exist!
Your video is top and the comments remind me of some things in my workshop.
For me the message is clear “Can’t be done, doesn’t exist, you find a solution for every problem!”
Great job! Many greetings from Germany. Hans Werner (Professional diffusion pump engineer)