on 5/31/2000 3:54 AM, Neil Ogle at [log in to unmask] wrote:

> To users of Micromass Prism III SIRMS and all interested parties:
>
> Over the last few months I have noticed the following changes in the analyser
> pressure of my Prism III.  From an acceptable pressure of 1E-3mbar the
> pressure
> has gradually increased, first to alternate visually between 1 and 4E-3 and
> more recently to permanently sit at 4E-3.  While not disastrous (I can still
> run samples, albeit with reduced success) I would be greatful if the list
> could
> point me in the right direction to resolve my problem.  What I have already
> done is check my rotary backing pumps and they appear to be functioning
> adequately.
> I have also baked the system.  I suspect a small leak somewhere in the vaccuum
> tubing in the machine but I would welcome all other suggestions before the
> machine
> "goes under the knife".
>
> Thanking you in advance
>
> Dr. Neil Ogle
> Environmental Engineering Research Centre
> The Queen's University of Belfast
> Department of Civil Engineering
> [log in to unmask]
>

We have a PRISM II dual inlet machine with two Pirani gauges.  One measures
the backing pump pressure, the other the inlet rough pump pressure.  Both
gauge calibrations have drifted over the years, probably due to slight
contamination from rotary pump oil.  Both gauges are isolated from direct
exposure to the pumps by alumina filled traps that can be regenerated by
baking.  The traps themselves can contribute to the rise in pressure seen on
these gauges when these traps become saturated.  If you attempt to
recalibrate as suggested by Stephen Howe, it pays to determine which pot is
which.  One pot adjusts the response at atmospheric pressure, the other pot
adjusts the response at "minimum" pressure (generally 1E-4 or below for a
good calibration).  Unfortunately, both pots will affect the calibration at
both ends, and a couple of iterations are necessary.  Generally this
calibration can be done conveniently only on another vacuum line where
pressures over the required range can be changed quickly.

The only times that we've had true leaks in our backing pump lines were due
to damaged o-ring seals on the Edwards diff pump dip stick assembly.  These
seals melted when we temporarily lost cooling water to the pumps even though
the heater power was cut immediately.  Residual heat from the oil reservoir
was conducted to the nearby dip stick assembly.  These leaks were readily
apparent in both the Pirani gauge pressure and the analyzer pressure, as
measured with the ion gauge.


--
Howard Berg
Development Engineer
UCSB Department of Geological Sciences
Santa Barbara, CA  93106-9630
Voice:  805.893.3487     Fax:  805.893.2314
mailto:[log in to unmask]
http://www.geol.ucsb.edu