Thursday
27th November 1997 - Disaster!
From John
Storey.....
Today
probably would have been a good a day if we hadn't tried to
take some data with the Mism. It quickly became clear that
there was a large signal no matter where in the sky we looked,
and that although the signal varied with postion it was not
in the way it was supposed to. A little investigation, with
Ant waving first his hand and then a soldering iron around
in front of the window, showed that only one beam was present.
That gave two possibilities, both ugly: either something was
blocking one of the beams, or one of the mirrors had fallen
off. Given the glass-rattling-about noises we heard the other
night when adjusting the rotator, the latter and uglier possibility
seemed the more likely.
We
removed the optics plate - a nasty operation to perform in
the cold because the plate has to be lowered an inch or two
and then held while the D-connector is unplugged and the window
heater wires unscrewed from the terminal block. We put the
optics plate in a plastic bag and allowed it to warm up inside
the AASTO for a few hours.
It
turns out that one of the little rectangular beam-steering
mirrors has fallen off. Of the three dabs of epoxy holding
it in place, one has simply dropped off both the glass and
the aluminium (it appears not to haave properly adhered in
the first placce), one has remained attached to the aluminium
plate and the glass but has pulled a big (5mm) chip of glass
out of the back of the mirror, and the other has come off
the aluminium *and* also pulled a chip of glass off the mirror.
The
possibilities now seem to be:
1.
Bring the optical boxes back to Sydney and make proper optical
mounts, returning the boxes in January. This is probably not
practical as we would need complete new sets of mirrors. It's
something we certainly should do before the end of next year,
though.
2.
Re-glue the mirror, using cryogenically rated epoxy. If we
take the second option, we need to realign the optical system
with a laser. This will be time-consuming, but possible. What
we would do is position a piece of paper above the optics
plate at the same height as the outer surface of the window.
Then, set up a laser so that its beam passed through the "intact"
channel and comes out through the centre of the window. Finally,
align the repaired mirror so that its beam crosses the first
one at the paper. Sound ok?
Some
questions for Max:
1.
What is the f number and diameter of the beam as it passes
through the window? 2. What epoxy was used the first time?
3. How critical is the alignment (we've found all the little
shims - we think)
We'd
like to think about it for a day before we proceed - advice
welcome!
Some
other software issues came up with the Mism:
1.
"cvfma" throws the cvf into continuous rotation.
Is cvfma in steps or degrees? Only "steps" is meaningful,
of course. 2. The figures given to convert rotator steps into
degrees appear to be correct. A quick "eyeball"
calibration gives:
ccw beam on horizon ~5,000 steps cw beam at zenith ~15,000
steps box horizontal ~23,000 ccw beam at zenith 30,000 cw
beam at BB 37,000 (assuming of course that there is a ccw
beam, which there isn't)
Can
I stop being depressed now?
It's
possible to cheer yourself up a bit by looking at:
http://141.224.128.11/ago/daily_reports/ago_today.lis
where
you'll see the AASTO data from the ARGOS transmissions (listed
as AGO-A2). All the numbers are complete nonesense until we
send Augsburg the correct calibration file. (Actually, the
wind speed and air pressure are correct, and the TEG volages
are nearly right).
With
mcba's help we got the 386 super set up so that people not
at the South pole can talk to it. It turns out that the gateway
and domain addresses on the hard disc had not been configured
for this location. The 386 super still crashes (or hangs)
if a telnet command has an unsatisfactory result - eg if it
tries to talk to an instrument that's not there.
Before
firing up the Nism again we tracked down the faulty Dallas
problem and checked the limit switches again. I was worried
that whatever had killed the Dallas (assuming it wasn't Larry
Hagman) might be the same thing that caused the limit switch
to go non-functional a few days back. Anyway, the limit switches
are fine (heaps of megohms, says the Fluke) and the Nism fired
up without any problems.
Here's
a rough calibration of where things are in absolute degrees
(ie, using "rotma")
ccw beam peaks at BB ~175 deg (abs) box horizontal ~390 cw
beam hits snow ~670
There's
a good strong signal that does all the right things. However,
looking at the adc readings it's very noisy. Even when the
signal has a s/n ratio of 100:1, the output from "det"
and "detx" is all over the place. I hope the data
acquiistion software is making multiple samples.
"adc
amps" also gives a very noisy result, although this cleans
up when the Stirling cooler is off. It's likely that the current
*is* in fact noisy.
"adc
volts, bb, cooler all give very stable output with fluctuations
of between 2 and 5 ADU.
The
faulty Dallas turned out to be just that. It was the one on
the Nism mounting bracket. Even after I cleaned it in bicarb
soda, washed it, dried it and brushed it, there's still no
response. It's been replaced with another one, which the software
correctly identifies as "no name found". (The Dallas
whose name can not be spoken.)
Occasionally
we get "sbit" as a response when we do a DS INIT,
but we're not losing any sleep over it.
When
I was wrestling with the Dallas I kicked the bucket of sodium
bicarbonate over, and now it's all over the AASTO. (But it's
a hell of an improvement on either brown slime or rockwool,
believe me.) There's an outside chance that when we come back
to the AASTO this time next year it will have grown stalagtites
and stalagmites, and be really quite attractive.
The
evening picked up considerably when we helped Fred unbolt
the 300lb friction disc from SPIREX. Yes, Fred had lifted
the crane into position single-handedly last night, while
we were asleep. Once we got the disk down, Fred lifted it
onto the trolley and wheeled it across the roof of MAPO. (It's
not clear now why we went to the trouble of installing a crane.
It would have been much simpler just to bolt Fred to the platform.)
Preparations
for Thanksgiving are now in full swing (It's celebrated on
Saturday at the South Pole - I'm not sure what time zone that
corresponds to!) The Dome was full of smoke this evening and
no-one was worried - the cooks had just built a turkey smoker
and fired it up (actually, given the beer situation, I suspect
that home-brew bourbon might also be part of the plan).
Abu
is still on the pump; the pressure is slowly going down, as
well it should. Ant leak-checked the T-tube and found it leak-tight,
which is a worry because everyone knows that fluorosilicone
O-rings are slightly permeable to helium. Ant is going to
try a different kind of leak checker, and possibly a different
kind of helium. Actually, Ant just came in very pleased and
said yes, he *had* got the O-rings to diffuse helium. Good
work, Ant. He's also replaced the "warm" O-rings
on the T-tube with Viton ones. They be fine as long as the
heaters don't fail - in which case Abu will leak anyway because
it contains most of the entire Parker O-ring inventory.
I
visited the ham shack in pursuit of someone who could make
up a 25-foot length of 10-base T cable for me, and found the
operator in the midst of a series of 30-second conversations
with a bunch of anonymous people. He seemed to be enjoying
himself, though I'm not sure why. If I get a chance I'll fire
the rig up and see what's out there.
Nothing
else good happened today.
John
 
|
Further Information
|
Contact:
| 
