My First Run of Diff Pump

[Jason C Wells]

First of all, thank you for making this resource available. I've been lurking and learning here for the last couple years. I am attempting the proton-boron reaction so my setup is a little different than that of a fusor. A VanDeGraaff will accelerate protons toward a grounded boron target. I'm trying to recreate Rutherford's 1934 results with a different apparatus.

I am running my diff pump for the first time. I have a CVC 4" water cooled pump with a Robinair two stage roughing pump. I have a rough mechanical gauge, a 1 Torr baratron and a .1 Torr baratron of unknown calibration. All equipment was bought used or was a hand me down from the university. I know it's a big pump but it was free.

I cleaned the pump with with achohol and bought new DC-704 oil from Duniway. I've read through the FAQ and I think I've covered all my bases.

My roughing pump is rated for 20 micron, though I question whether I achieve that. I should be able to run the roughing pump until my 1 Torr gauge comes off the high peg. The 1 Torr gauge did not come into range after 1 hour run of the roughing pump and diff pump. I was all excited to see some really low pressures. Then I went to scratching my head.

It seems that I might have a slight blind spot in my vacuum instrumentation. By the advertised spec of my equipment, I should be able to see that pressure drop into the range of the 1 Torr gauge. I did not. I show 30 in hg on the mechanical gauge so I have even better than perfect vacuum according to that. I was (hoping I was) on the borderline of your recommendations for starting my diff pump, which I did.

Do you think I need a higher pressure baratron to see pressure from the rough to high vacuum range? I'm thinking 1 Torr is sufficiently intermediate, but since I don't see the gauge indicating any decrease in pressure I am wondering.

How long would you expect a first run of a vacuum system to produce good vacuum? Is this something that takes a day to cook all the gasses out of my oil. Was I naive in my hope that vacuum was low enough on the roughing pump to start my diff pump? I hope I didn't coke my brand new oil.

At this stage, I need the benefit of a little experience. Any pointers?

[John Futter]

Jason
firstly you will have to see your roughing pump pump almost to its rated best vacuum before switching on the diffusion pump.
this may include an extended period of gas ballsating of the roughing pump to clean the oil of water and other contaminants.

only when you can get to this stage do you switch on the diff pump

if all is well your roughing pressure will rise as the diff pump now cleans its oil.
the stall pressure for most diffusion pumps is around 1.5 x 10 to the minus 1 millibar or 9 by ten to the minus 2 Torr.

if you cannot get this vacuum in the foreline you have either a leak (most probable for a first time evacuation) or a heavily contaminted chamber and associated vacuum lines or both.

from your readings of the baratrons you do not have a good enough vacuum

you need a thermocouple or pirani gauge that is useful over the atmosphere to low ten to the minus 2 millibar / torr range before you go on.
your mechanical pumps rating is ultimate vacuum 2 x 10 to the minus 2 millibar so you see you do not have much riggle room.

[Richard Hull]

John's right! No need to turn on the diff pump at all until you are well below 30 microns on the foreline. You either have a seriosu leak or a rotten forepump.

A good forepump should yank the enitre system down to under 50 microns in a matter of 3 minutes unless you are trying to evacuate a 55 gallon oil drum with a 1.1cfm pump.

Richard Hull

[Jason C Wells]

The keyword I got from your reply is ballasting. I see now that I missed that part. My rough pump oil was replaced when I bought that pump. It has been used several times over two years since then. I have to admit that I've use mechanical pumps in the past for refrigeration and didn't really read up on the mechanical pump information.

[Jason C Wells]

Since the last time I posted here, I installed a 0-100 Torr MKS manometer. This is really the perfect size manometer to allow you to see pressure dropping as the mechanical gauge pegs low. I also upgraded my roughing pump to a new Superevac unit. My Robinair had some pretty bad wear on the vanes when I inspected it. My complete system can achieve 6 Torr on the rough pump alone. With the pump blanked off I can pull .5 Torr which still isn't good enough.

30 Microns is hard.

Regards,
Jason

[Richard Hull]

With just a half-way decent used pump, 30 microns is easy! I have never changed the oil in my old used Precision pump in the last 7 years and it still pulls to 12 microns over the entire fusor vacuum system, every time in about 4 minutes after ballasting. That's when I start the diff pump heater and cooling fan.

As the others advise, you either have a terrible leak or a weak or poor pump.

Richard Hull

[JakeJHecla]

Jason, where are you in the greater Seattle area? I have a Welch 1402 in Renton that I can lend you. I'm in Boston until Christmas, but if you can wait that long, I can get you something capable of 10 microns.

[Jason C Wells]

My pump is nice. My system is half baked. No wait. If it was baked it'd probably do better. My system is leaky I'm sure.

[Jason C Wells]

After setting aside vacuum apparatus for a while to work on a pelletron, I am returning to the vacuum side of the house. I've decided to divide and conquer. I've redesigned some of the "tubulation" (MKS term, not mine) connections that I cobbled together for instrumentation. I've disconnected equipment to isolate where leaks might be occurring. Vacuum immediately dropped to .3 torr upon starting up my pump with the ballast port closed. That's an order of magnitude over what I was doing last fall. The pump is ballasting as I write this message. I hope to leak check the accelerator column and target chamber real soon.

Regards,
Jason

[Dennis P Brown]

If your mechanical, two stage pump, is taking your system only down to 0.3 torr (300 microns) you have a really bad leak or leaks OR your pump is at fault. So, when you said a while ago:

With the pump blanked off I can pull .5 Torr which still isn't good enough.

First off, did you recheck the mechanical pump against your gauge? If you are not down below a few microns in seconds - look no farther - your pump is at fault and you are utterly wasting your time. No two stage pump should ever "only" reach 300 microns (0.3 torr) just pumping on a gauge! That is, unless the gauge is in error - I will assume you have ruled that out and yes, that is trivial to confirm even without a second gauge to cross reference.

Next, (if it passes) add components (if possible) and determine when the system is starting to fail - at that point you've found the first leak problem (may be more, may not.) Continue until the system is sealed.
If you cannot do this then try the following.
Another method for a leak detection (for large leaks only like you appear to be suffering) is to pressurize the system (say 1.5 atmospheres of air- a tire pump works) and use a soap solution to locate the bubbles such a leak will produce. I am assuming you have carefully checked all seals visually and used sound (pumps off!) to try and detect any gross vacuum leak(s)?
A photo of your setup could aid people here to help you in correcting this problem.

Aside: If your oil is that bad (only 300 microns!) ballasting alone will most likely not cure the problem. More likely, your system is just leaking (if leaking that bad - do ballast the pump but find those massive leaks.)

[Peter Schmelcher]

The ballast makes a big vacuum difference and could easily make one think a leak is reducing. The dashed lines are with the ballast open on a family of two stage belt driven pumps from Japan. You should of course expect a different curve for your pump.
-Peter

[Jason C Wells]

The pump is fine. The plumbing is poor. I did positive pressure checks today. I plan to go buy some quality fittings on Monday morning.

( Edit: This reply should have come after Dennis. I didn't realize there was another page.)

Regards,
Jason