[Prev][Next][Index][Thread]

Dif pumps, etc.

> Our glass diffusion pump has inbord heaters and we have not had a problem
> with flashback. Inbord heaters (heating coils sealed into the pump) is the
> way to go in my opinion. It makes service a little more complicated , coils
> are said to burn out after some time. I've had mine for two years and they
> still look the same as when I put them in. Maby by using silicone oil instead
> of hydrocarbon they last longer. The use of silicone also enables the pump to
> be dumped to atmosphere with out doing to much damage to the oil. In my
> opinion this inovation eliminates the need for a two mechanical pump system
> in our industry. A single two stage mechanical pump and a diffusion pump seem
> to perform very well. Improvements to a diffusion pump? Inbord heaters, water
> jackets around the condencation section and an optical baffel (no direct line
> of sight) between high vacuum side of the diffusion pump and the main
> stopcock  also with water jackets to reduce the amount of backstreaming.

I too think these inbord heaters are the way to go.  I have never seen 
submerged coils burn out as a result of normal use.  Flashback, is 
another thing.  If you have no grounding electrode (or grounded vacuum 
gauge sensor, or other metal) in your system, then your delicate 24 gauge 
nichrome wire in the dif pump will be forced to absorb this load.

I believe a big mistake many scientific glassblowers make when fabricating
dif pumps is to run the _completed_ pump (with inboard) heaters thru their
annealer.  The nichrome wire, and worse, the brass/silver solder/whatever
(used to secure the element) are heavily oxidized in the oven - often
close to the failure point.  I've seen excellent results from
oven-annealing the pump, then inserting the completed glass/metal-seal &
nichrome-wire assembly and carefully flame-annealing these.  In use,
inbord heaters enjoy a long, long life.  They operate in a protected
atmosphere, and well below their service limit (ie: "red hot"). 

I'm a big fan of dif pumps, and I couldn't imagine not haveing one. I
think a single stage, air cooled, glass unit with DC705 silicone fluid
will suit even the most demanding system.  I have a slow (25 l/m)
mechanical pump - yet after the bomb I'm to "0" (the lower limit on my
thermo gauge) within 30 seconds or so. I've changed my oil a couple of
times - not for any degradation in performance, but just because the fluid
had become dark with age.  People who buy one thinking they can just plug
it in are usually dissappointed.  They require attention, and a basic
knowledge of how they work.  In my experience, each pump seems to have a
different "personality" as well. 

Water cooling is common on the 2 and 3 stage dif pumps, but not on the
single stage models.  I'm not sure why.  I made a crude copper-wire
radiator for mine.  It's simply pieces of 14g bare Cu wire wrapped around
the neck just above the internal jet (where the oil should condense). A
simple touch-test says the Cu is dissapating a lot of heat. Seems to
facilitate condensation, but then again it worked pretty good before this. 
I got the idea from an old 50's vintage physics book (Proceedures in
Experimental Physics - a great bathroom book BTW!)

     -John Anderson