7th February 1995

From Michael Ashley.....

Just when we thought everything was working smoothly, the IRPS mirror rotator begins to act strangely. When driving it from zenith to nadir it vibrates alarmingly and skips steps. Thinking the problem could be a missing phase, John spends some hours checking for broken cables and electrical problems (it is not easy removing the backshell from a connector on the roof). This time is considerably extended due to John's need to persuade the LeCroy 300 MHz Digital Sampling Oscilloscope that all he requires is a simple trace of voltage versus time rather than a fourier transform or cross-correlation with something it has previously measured. The LeCroy has a brain the size of a small planet, and is reluctant to do something as basic as displaying a simple waveform, without first providing the operator with every opportunity to sample its prodigious capabilities. Only after John's threat to unplug it does the LeCroy agree to masquerade as a dual-beam oscilloscope.

All connections to the motor check out OK, and the drive electronics is also fine. We then begin to suspect a resonance between the rate at which we are driving the motor, and its natural torsional period. By trying various inter-pulse delays, we find that at practically any rate other than the one we were using (30 milliseconds between pulses) the motor behaves perfectly. It appears that a slight change in the calibration of the software timing loop, and a small change in the motor driving voltage (due to using a UPS), conspired to bring the rotator into resonance. Resolving this problem consumed almost all of the day, and left us somewhat concerned about what new surprises IRPS had in store for us with only 24 hours left before we had to leave the Pole.

Dinner was superb: lobster tails and steak and no ``toxic purple'' in evidence. Coincidentally, three senior NSF representatives were visiting for the day.

Jean is analysing the data from his first two balloon launches, and is getting ``seeing'' figures of around 1 arcsecond. Jean is also excited to discover that he has won a lottery to travel back to Christchurch aboard the vessel Greenwave. The trip will take 6 days.

John is pursuing his goal of videotaping the remaining women on the base, and manages an extended session with Emily, the cook. (John says it is important that our video properly acknowledges the vital role played by women in the running of the South Pole base. Next time, he can write the software and I'll video the women.)

Erik completes his goal of being the fastest person to unicycle around the globe, completing all 360 degrees of longitude in about 10 seconds.

While using the computer to fill the inner can of IRPS, John notices that when the can is full a great spray of LN2 comes gushing out of the pressure relief valve. I decide to use this as an indication of fullness, and install a silicon diode in a suitable place to be hit by the LN2. An intermittent solder joint in the control box thwarts by first attempts, but in the early hours of Wednesday the system is working smoothly. We can now sense when the inner can needs filling (by monitoring the detector temperature) and when the can is overfull (using the newly installed diode). The IRPS computer now keeps track of the time at which each fill occurred, and for how long the LN2 solenoids were open. With experience we should be able to use this information to decide how full the Taylor-Wharton dewar is, and to alert Jamie to the need to refill it.