In this short Note, I describe the performance of the Rochester Mark IV's ThermoElectric Cooling (TEC) units on the evening of Aug 8, 2002 EDT.
My plan was to send a series of command voltages to the TEC units, asking them to drop the camera temperatures further and further. I wrote a script which included a double loop:
The idea is to watch the camera temperature as it drops to the next equilibrium value. Tom Droege suggested giving it an hour at each setting, but I wasn't able to start early enough in the afternoon to do so. I began the experiment at around 6:30 PM on Thursday, Aug 8. It had been a warm day, with temperatures in the high seventies or low eighties. The sun shone all day on the garbage can which holds the coolant, so its temperature was pretty high (about 38 degrees C, according to the Mark IV thermistor). The Mark IV sits in a little plastic shed, which holds the cameras, telescope, control computer and UPS. It gets very warm inside, apparently: the air temperature (which is indicated by "VCO temperature", I believe) was over 40 C at the start of the run.
Here's a graph showing various quantities as a function of time. I'll explain below.
The horizontal axis marks time. The data start at 6:21 PM EDT on Aug 8, and runs to about 11:00 PM EDT. It was easiest for me to plot in terms of hours since 8:00 PM EDT on Aug 7, which is the same as 0:00 UT on Aug 8.
The orange dots near the top mark the temperature of the water/antifreeze coolant pumped through the camera head. The green dots mark the temperature of the air around the Mark IV. You can see that both cool off as the sun sets and night begins: the air inside the shed cools by nearly 20 C, and the water cools by about 10 C. Something I don't understand are the vaguely periodic variations in water temperature. Note that the sudden drops and recoveries occur at the same time as small increases in the reported CCD temperatures.
The light-blue (I-band) and dark-blue (V-band) dots show the reported temperatures of the CCD chips. The I-band camera seems to stick at around 0 C, and the V-band camera bottoms out around -10 C.
Above the dots showing temperature are small symbols connected by lines, with the same color-coding: light-blue for I, dark-blue for V. These show the TEC current values, multiplied by a factor of 10 for clarity. For example, the V-band camera's TEC current rises immediately to a value around 1.25, and stays there for the duration of the test. The I-band camera, on the other hand, has a TEC current of
TEC command voltage I-band camera TEC current
+3 1.0
+2 1.2
+1 1.4
0 1.4
-1 1.4
-2 1.4
-3 1.4
So, the I-band TEC unit doesn't run as hard as possible until
the command voltage drops to +1.
The magenta lines near the bottom of the plot show the command voltage sent to the TEC units. They have been displaced downwards by 20 units for clarity. The first stretch represents a value of +3, the next +2, and so on. I had to cut short the times for values of -2 and -3 in order to start some other tests.
Note that after both TEC currents rise to their maximum value, when the command volage is +1, the CCD temperatures simply follow the gradual decrease of the water temperature. It looks like, when running as hard as possible,
V-band TEC unit cools to 30 C below water
I-band TEC unit cools to 40 C below water
I think it's clear that I need to find a way to prevent the water from heating up to 38 C during the daytime.
However, I am puzzled by the small, mildly periodic variations in water and CCD temperatures. Any ideas?
Last modified Aug 9, 2002 by MWR