Mechanical pump oil really matters. Especially with “new” pumps.

Every fusor and fusion system seems to need a vacuum. This area is for detailed discussion of vacuum systems, materials, gauging, etc. related to fusor or fusion research.
Post Reply
JoeBallantyne
Posts: 313
Joined: Tue Jul 20, 2010 4:08 pm
Real name: Joe Ballantyne
Location: Redmond, WA

Mechanical pump oil really matters. Especially with “new” pumps.

Post by JoeBallantyne »

I have had my eye on a Welch 1402 DuoSeal mechanical pump that was listed repeatedly on the Seattle area craigslist for the last year and a half or so. The listing claimed it was for a used once pump that was for a commercial chiller project funded by the local Portland gas company. When I first saw it, the listing was at $800 for the pump, but slowly dropped as there were no takers. Last week I checked, and a new listing popped for what appeared to be the same pump for $495. I jumped on it, and offered the guy an even $500 which he accepted.

I tracked down my old 1.5 inch diameter red rubber vacuum hose with a KF25 connector on one end, and took a couple of MKS901p vacuum gauges with me, and drove the 3.5 hours down to Portland. Tested the pump which was setup with refrigeration line connections instead of standard KF fittings or a proper hose barb, and it pulled down to under 3 microns. This was with my hose clamped down onto the threaded top connection of the refrigeration connector on the pump – definitely not the optimal setup for a vacuum tight connection. I figured it was good enough for a first 30 minute run after sitting for a decade or so, paid the man, and hauled the pump back home.

This was a new style Welch 1402 (has the rectangular oil sight glass instead of the keyhole style sight glass), so I bought a “new style” Welch 1393G KF25 adapter for it off Amazon from Brownian Motion Technologies. Ordered the adapter about 1AM in the morning, and it showed up later in the afternoon the same day! One of the perks of living near Seattle where Amazon was headquartered for many years – some stuff shows up same day – which is really nice.

Connected an MKS901p directly to the newly installed KF25 adapter, fired up the pump and let it run. After several minutes it pulled down to 1.05 microns, but after letting it run for several hours and getting up to temperature, the pressure rose to about 1.7 microns. Still a great pressure, but I decided I wanted to see what would happen if I replaced the original stock Welch pump oil that had shipped with the pump and was still perfectly clean in the sight glass, with Inland 19 Ultra which has a vapor pressure of 1e-6 torr at 25 degrees C. This is about 10x better than the specs on standard vacuum pump oil – 1e-5 torr at 25 degrees C.

Pump after running for several hours with original oil,

IMG_20230501_204602691.jpg

I set the pump up on the seat of a chair, ran some 5/16 ID vinyl tubing from the drain stopcock to an empty clean 1 gallon HDPE container (originally for windshield washer fluid) on the floor, and drained the original oil. Then I turned on the pump, uncorked the input, tipped the pump up 10 degrees or so to maximize drainage out the oil drain spigot, gently put my hand over the bare exhaust output (no dust cap or exhaust filter installed) – partially blocking the exit of air from the pump exhaust and thus slightly pressurizing the exit of oil from the drain – and with the drain still pressurized then poured 2-3 cups of the new Inland 19 Ultra oil directly into the pump intake with the pump running. This flushes out most of the original oil from the pumping mechanisms and from the body of the pump. The whole flush took 20-30 seconds before the new oil was pretty much all flushed out through the oil drain. Partially blocking the exhaust vent with the pump running and slightly tilted really helps get most of the oil out of the pump. You do NOT want to completely block the exhaust vent as too much pressure inside the body of the pump can damage it. When I was done, the 1 gallon container holding the flushed out oil was about three quarters full. The Welch 1402 has an oil capacity of 2.25 quarts.

I shut off the pump, closed the drain stopcock, put it back on the floor, filled it up through the exhaust port with Inland 19 Ultra oil, connected an MKS901p back to the input, fired it up, and then was absolutely shocked!

I figured that maybe best case the pump would get down to its spec’ed 1e-4torr (0.1 microns). Well, it did, but then it kept going down! It actually pulled down to the minimum the 901p will measure – which is 1e-5 torr. It took a while to get there – 20-25 minutes of pumping, but it just kept pulling lower and lower until it bottomed out the gauge! I was completely flabbergasted. Furthermore, even when the pump got up to temperature – 120 degrees Fahrenheit – the pressure didn’t rise at all. Stayed rock solid at 1e-5 torr. It wasn’t the gauge either, because the pump bottoms out multiple different 901p gauges.

Pump after filling with Inland 19 Ultra

IMG_20230504_221625233.jpg

So basically, the new oil was more than 150x better than what was in the pump. 0.01 microns vs 1.7 microns. Pretty amazing. Some of the modern mechanical pump oils are really very excellent. I think the Inland 19 Ultra is quite a good deal, as you can get it from SIS https://www.sisweb.com/vacuum/sis/inland19-ultra.htm#1 for $140 plus $75 shipping for a 5 gallon container. Which works out to $40 a gallon and its vapor pressure specs are much better than the more expensive stock oils that ship with most pumps.

I think one reason the oil made such a huge difference in this case, was because the pump itself is basically brand new. It probably was only used once, and then sat. I have another “old style” 1402 that was backing a diffusion pump, which I bought with a fusor system from Tom Bales, that pulled down to 2.5 microns after 10 hours of pumping with its original Alcatel 200 oil. After replacing the oil with Inland 19 Ultra, the same pump pulled down to 0.72 microns after 2 hours of pumping. So a factor of 4x better instead of 150x better. A ~4x improvement from using Inland Ultra 19 is probably going to be more typical with most older pumps.

Still, the fact that a new Welch 1402 mechanical pump will pull down to 0.01 microns with good pump oil boggles my mind. I thought for sure I would have to use either a diffusion or turbo pump to get that low. I’m really happy that is not the case.

For folks that are using 2 pump solutions (mechanical + diffusion, or mechanical + turbo) for their fusors, a backing pump that pulls 15 microns is just fine, so pretty much any vacuum pump oil will work OK. But if you, like me, want to use only a mechanical pump to pull your fusor vacuum, then what kind of oil you use in that pump can really make a difference.

Joe.
Aleksa Djokic
Posts: 58
Joined: Thu Dec 22, 2022 2:22 pm
Real name: Aleksa Djokic

Re: Mechanical pump oil really matters. Especially with “new” pumps.

Post by Aleksa Djokic »

Very interesting indeed. The idea of not using a secondary pump may not be such a good idea. If the system is going to be used for a fusor, or some other plasma project, there needs to be a constant pumping speed and still maintain high vacuum, the roughing pump alone may not do the job. Keep in mind that pirani gauges are one of the most inaccurate out there, the only time a understood how, for example, leybold or edwards claimed the vacuum of 1x10-4mBar for their two stage pumps is by using a ion gauge that is cooled by liquid nitrogen, a bit of a cheat in my opinion, not stating that the liquid nitrogen trap is a pump of its own that allows to reach such low levels of vacuum. When measured with a standard gauge like TC, Pirani, Baratron, Penning, the levels are in the low 1x10-3mBar range. I have measured both new and used pumps, never saw a better vacuum then 9x10-4mBar measured with the above mentioned gauges directly coupled to the inlet of the pump.
JoeBallantyne
Posts: 313
Joined: Tue Jul 20, 2010 4:08 pm
Real name: Joe Ballantyne
Location: Redmond, WA

Re: Mechanical pump oil really matters. Especially with “new” pumps.

Post by JoeBallantyne »

I already use only a mechanical pump on my fusor. A different Welch 1402 DuoSeal from the two mentioned above. It pulls down to ~1 micron, and works just fine and dandy with my fusor. I typically run the fusor at about 5 microns with D2, and have the vacuum mostly throttled off using a manual butterfly valve. The butterfly valve is almost fully closed.

So, yes, using just a very good mechanical pump on a fusor is certainly an option, and much simpler than having to deal with a dual pump setup.

As to the accuracy of the 901p, I doubt that MKS would ship a product that was grossly inaccurate over its usable range.

You only get single digit precision on the e-5 range of the 901p, but the point is that anywhere in that decade is well below where most folks (Joe Gayo excluded) run their fusors.

To the point of this thread, what kind of oil was in the mechanical pumps you were measuring pressure on? Was it spec'ed at 1e-6torr vapor pressure at 25 degrees C? My guess is that it was not. Typical vacuum pump oil will bottom out at a higher pressure than excellent vacuum pump oil. That is the whole point of my post.

Joe.
User avatar
Richard Hull
Moderator
Posts: 14976
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Mechanical pump oil really matters. Especially with “new” pumps.

Post by Richard Hull »

Great story on what was hailed as effectively a new pump and use of a super oil in it. those are very impressive numbers. If one's goal is for 10e-7 of 10e-8 with a secondary pump this would be the way to go, I would think. Bottom out the mechanical and let the turbo go even deeper.

My formally seized precision 5 CFM, that was had for free in the grass at a hamfest does 5 microns currently and with the Turbo drags to 10e-5 due to a forever leak in my fusor. All of this is just fine for fusion, of course. Even a 20 micron bottom-out on a mechanical would do fine, of course.

Still, it is nice to know with the best oil in a nice pump will do wonders rarely seen in the world of mechanical pumps. Good report for folks with a truly good pump and the money for the oil.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
User avatar
Dennis P Brown
Posts: 3147
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Mechanical pump oil really matters. Especially with “new” pumps.

Post by Dennis P Brown »

I have an old pump at work but it easily reaches 0.2 micron's on a very large system (Bell Jar.) The saying is true: Size really does matter. That pump you have is a monster.

My fusor's large mechanical pump quickly pulls my dirty system (exposed to air) below 2 microns. When clean the gauge drops below 1 micron; however, that TC gauge isn't exactly accurate below 1 micron and my cold cathode just starts there. Bottom-line, getting below 0.1 micron for that pump is believable.
User avatar
Richard Hull
Moderator
Posts: 14976
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Mechanical pump oil really matters. Especially with “new” pumps.

Post by Richard Hull »

I wouldn't call a 1402 a monster pump. It is a 5CFM pump. That is about normal for a good vacuum system with a moderate size chamber. I have two good 1402's plugged up with new oil in them in the lab. I also have two welch 1397 monsters in my lab, 17.7 CFM. I have been trying to sell these one Horse power motored big boys every year at HEAS for $800, no takers. They are sold new for over $6000. If you like the 1397 used, Duniway has a rebuilt for $3100. Is it the 240 volt operation that kills the sale of one of my 1397s? Perhaps it is the near 200 pound gravitational lock with terraferma that makes folks demur?

My free Precision is a 5 cfm and working fine. Why don't I use one of the marvelous, much heralded Welch pumps?? Let the sleeping dog lie...Always...

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Post Reply

Return to “Vacuum Technology (& FAQs)”