Contamination in the dewar
Since the last report of February 5th, considerable work has gone into
solving a critical problem - oily contamination in the dewar. On baking
and pumping the dewar, some oily condensate has been observed,
particularly on the cap covering the PTR port and the valve to the pump.
These are two of the coldest regions of the dewar during baking and
are reasonable sites for condensation of contaminants which are
evaporated from the hotter parts. A cold trap placed on the pump line
also accumulated considerable condensation but apparently mostly water
based although not completely clear and of a yellowish tint.
Contamination is a potential serious issue in a dewar. This is
because it can change the optical properties of lenses and filters
and, in some cases, permanently compromise the sensitivity of the
detector. The HAWAII detectors, however, have an inert, sapphire
cover over the detector itself and are unlikely to suffer damage
from hydrocarbon contamination. (CCDs, on the other hand, can be
In an effort to trace the source of this contamination, considerable
effort was put in by Anton and Benjamin over the summer, removing and
cleaning the various components, especially bearings. Each were baked and
pumped individually and in a test dewar, but no obvious source was
found; even some probably non-specification greasy lubricant on the
wheel bearings does not seem to have had a definitive effect.
A residual gas analyser was purchased in July 1999 and sent to CUO to
be used to analyse the composition of the best vacuum that could be acheived.
The only anomalous result was the discovery of a considerable
proportion of mechanical pump oil in the system. This is present even
without NOTCam connected but with just the pump and the RGA. This
is an odd result because the pump is unlubricated as appropriate for
high quality work.
It is entirely possible that the root cause of this problem will
never be identified. The oily contaminant is undoubtedly spread throughout
the dewar and particularly coats the huge surface area of anodized
The only solution, for now, appears to be to bake and pump NOTCam
until no contaminants can be condensed out, slowly driving off the
oil from whatever surfaces are affected. At the time of writing,
this arduous process is more or less complete.
The watery contaminant is likely precisely that, water. It is a
perennial problem in large vacuum systems.
Work on the motor controller is largely complete and all driver
software checks out.
The removal of all lubricants within the dewar as part of the
decontamination effort increased the load on the stepper motors
sufficiently that additional shims had to be introduced to reduce
the pressure on the gears. Also, the focus mechanism has had to
be geared down by 2:1.
As part of his degree report, French student Benjamin Noel carried
out some experiments to determine the gettering effectiveness of
anodised aluminium. Unfortunately, the results imply that it will
have little effect. A new, large, activated charcoal getter has been
designed into NOTCam.
Work has begun on adapting CUO controllers to run HAWAII arrays.
The test dewar should be up and running within a few weeks.
The order for narrow band filters has been made to the Gemini Filter
Consortium. A few have been received. Please see the
filters page for complete information.
The heavy delays now experienced by the project and the departure from
NOT of Michael Andersen, who is the primary driver of this functionality,
mean that the spectroscopic capabilites of the instrument have been
temporarily shelved and the first priority is to make the instrument
useful as a simple imager.
An addendum to the original NOTSA/CUO contract has been drafted in
order to formalise and constrain the large amount of unforeseen but
necessary work that CUO has been forced to accomodate.