I'm refurbishing a Edwards Diffstak diffusion pump that I got on ebay. Turns out it was really gunked up on the inside and is going to take some cleaning. When I took the first look, I didn't have any specific knowledge about this design, which looks a bit different inside that most diffusion pumps.

manual pdf

I got these parts removed: cold cap, retainer strap, and spring that holds the jet assembly down against the boiler plate, revealing this:

Unfortunately, this part, which I later verified is the top of the jet assembly, did not move or come out. I knew from other disassembly instructions and videos that this part usually is loose and will even fall out. The nice people on http://www.fusor.net (a good DIY vacuum resource) were able to confirm that this was supposed to just pull out. I tried soaking acetone and mineral spirits in there; this did not loosen things up much, but flushed out a bunch of soot and carbon flakes.

I was able to get a manual pdf just by asking Edwards customer service nicely, though my saying I was from CMU on the contact form may have helped. This is for the “Mark 2” version; mine is slightly different, and is presumably older. Diffusion pumps were the first practical kind of high-vacuum pump, and many of the ones you see on ebay are 50 years old or more, made by companies that are long gone. The Diffstak is a late design, with refinements like a built-in cooled baffle to reduce backstreaming of oil vapor into the chamber, and also all stainless construction except for the heater mount.

Cutaway drawing from the manual:

It turns out that the tapped hole in the top of the jet assembly is for a puller, although the tool is not described in the manual. The threads are M5. I came up with this:

(I should have put a block under the lever to keep from marring the flange).

I decided to try heat to loosen things up. The base of the diffusion pump is designed to get hot, >200 C. Because I had already put solvents in, and who knows what was in there before, I did this outside.

I was prepared to escalate to a blowtorch if necessary, but did not have to. I connected up power (with no cooling water) and waited for it to heat up. Since this is a 220V heater, I used a variac to step up the 120 to 160. When it started sizzling and smoking and smelling like mineral spirits, I tried the puller again, and it came out with only moderate force.

You don't want to pull super hard because you might just rip the bottom nut out of its attachment. That cross brace you see in the bottom view is just sheet metal. 10's of pounds of force should be ok, but not 100. The heat made a big difference.

This is the jet assembly and some loose carbon that fell off the bottom of the jet and out of the main tube when I turned it over. I was afraid this would be a hard carbon deposit, but it mostly seems more sooty. There is also some brown deposit that isn't completely carbonized and will require solvent removal.

This is looking down inside the empty tube:

I think the sticking was mostly due to the fairly close clearances between the bottom of the jets and the outer tube. This gap only needs to be wide enough for the trickle of condensing pump fluid to flow back down to the boiler, and should not be too wide or vapor will come up there rather than going out the jets. The manual says you are not supposed to use abrasives or mechanical cleaning because it is likely to damage the pump. I'll try scrub brush first.

I just got the pump reassembled and working.

I decided pretty early on I would use abrasives where necessary. At least in my case most of the carbon seemed to be loose soot. There were some patches of hard glassy carbon on the jet assembly which I decided to just leave alone, since they were harder than the aluminum and really stuck on. Before getting serious I tried various solvents (paint thinner, acetone, xylol) and water-based cleaners (pine sol, citrus degreaser), and really nothing did anything. I don't think anything will easily remove plain carbon. Hot concentrated hydrogen peroxide was suggested in an old vacuum book; this is part of Piranha solution.

Fortunately the carbon in the main body just washed out with a solid blast of water. I was especially worried about the filler tube, which is maybe 1/4“ ID, and neither end is very accessible. So the water may be worth doing, but I would skip soaking in caustic cleaners, as I found that you can rust the bottom end of the main enclosure (which is mild steel, not stainless, likely chosen for higher thermal conductivity and weld compatibility). The visible bottom of the inside is aluminum, but this seems to be just a thin disc sitting on top of the steel. So once you use water, use heat to dry it out again.

I removed the varnish and most of the carbon on the jet assembly by using a Dremel tool with a brass wire brush and oil. I used tapmatic cutting oil, but I suspect any oil would do. The oil speeds the removal and reduces the tendency for the soft brass to smear off on the cleaned surface. I used 1000 grit silicon carbide paper and oil in a few places. I also did some buffing with emery, but I'm not sure it was that good in comparison to the brass brush.

Do not try to remove the heater unless it is dead. I decided to remove it because I wanted to soak the whole thing in water, but the cement heater element was stuck to the rust on the bottom of the steel base, and the heater element pulled apart. I was not too upset because I wasn't crazy about powering the 210 volt rated heater. I was able to replace it with a fairly similar US-made 120V Chromalox heater.