This brief note simply serves to show a few results of a test of the cooling system on the Rochester Mark IV. It isn't definitive, but a first step to more comprehensive tests.
On the (cold) night of Nov 6, 2001, the sky in Rochester was mostly clear. I attached the liquid coolant lines between a garbage can filled with about 10 gallons of water/antifreeze and the Mark IV, and turned on the pump. I set DACs 14 and 15 to lower and lower values -- which ought to increase the current running through the TEC units which cool the CCD chips. I recorded the values for water temperature and chip temperatures reported by the STAMP, and also took a set of 5-second exposures of the sky.
Decreasing the DAC 14,15 values does cause the coolers to work harder, as the graph below shows:
The curious thing is that when I ran similar tests indoors, back in May, 2001, I found much larger values for the difference (CCD temp - water temp) at a given DAC setting. For example, a DAC setting of "2" then yielded a temperature difference of about -34 degrees Celsius; in the recent tests, the temperature difference was only about -12 degrees. Why? Perhaps the pump has lost some of its power, and is driving less water through the hoses. Perhaps there's a blockage in the hoses somewhere. I'll have to check the volume of water moved by the pump per minute.
How did the lower CCD temperatures affect the images? I should have taken dark images, with the shutters closed; unfortunately, I didn't. I tried to find a column which was effectively dark -- covered by some layer of material. However, the edge pixels seemed to be
So, with this data set, the best I can do is look at the sky value as the temperature dropped. If the sky was uniformly dark during the tests (clouds appeared near the end, so probably not), then we expect to see the sky value drop to a limiting value as the temperature-induced noise becomes negligible compared to the photon noise from the sky.
As the graph shows, CCD 0 did approach a limiting value, so it appears that a temperature of -10 degrees Celsius is a reasonable operating temperature for it (if the Rochester sky conditions are always the same as this particular night). However, the sky value in images taken with CCD 1 appear still to be dropping at the end of the test, so lower temperatures would improve its images.