Dear James,
Back in the late 70s/early 80's, we used a stainless bellows pump made by
METAL BELLOWS CORP, SHARON, MASS.  (This is all the address which is given
on the label). It was Model MB-41 and as I recall not terribly expensive at
the time. I don't believe we had any fail. However, there is an obvious
problem to cope with. If you are circulating gas at a low pressure, you
have pressure on the outer bellows surface of about one atmosphere. This
means that the drive rod pulling the bellows  face, the joint which fastens
the rod to the face, the associated eccentric on the motor shaft and of
course the motor are "fighting" this pressure differential during
expansion. Of course during the next half revolution, the bellows contracts
with a thump under the combined atmospheric/drive rod force. In other
words, the pump works in a saw-tooth pattern rather than continuously and
the wear on bearings, heating of the motor,etc are quite damaging.  However
there is a solution which can be as simple (cheap) or complex (expensive)
as you wish. I hope my explanation with a crude drawing is not confusing.


^^^^^^^^^^^^^I
BELLOWSI------DRIVE  ROD-------O >>>>>>
^^^^^^^^^^^^^I                               /I    Spring
                                               /
      Collar on drive rod within which the
     eccentric on the motor axle rotates

On the side of the drive rod collar opposite to the bellows, drill a small
hole and tap it so as to attach a matching threaded rod which will bend
readily into a hook. To this hook attach an adjustable spring. In our case,
we used a loop (band) of small diameter rubber tubing, one end of which
could be tied by a cord to a nearby aluminum post which was part of the
preparation line support frame. The motor was turned on and the rubber band
stretched until the pump ran its smoothest at which point, the cord was
knotted a two person job)  A sleeve of small tygon tubing on the hook
prevented it from cutting the rubber band. One could use a metal spring
with a long threaded bolt/nut arrangement but such fine-tuning is pointless
if the gas pressure changes. Although it never happened, the rubber band
could break and take off in an unfavourable direction but one could shield
the band appropriately.

Roy Krouse