|
CCD Notes
July 24, 2004 Connelly's Springs, NC. Heat in the kitchen is rising...  This is last week's 45 minute exposure of M101 revisitted. I've used 12 passes of Lucy-Richardson deconvolution based on a captured point spread function to remedy some star bloat. And I'd like to think I've applied some recent lessons about noise to get at some fainter details. A couple of background galaxies have come through in the full-res image, and outer spiral structure is visible in the data but not cleanly enough to show off in a pretty picture.
The water-cooling was so successful, I repeated all the measurements at
2x2 binning and experimented to see how much heat gets dumped into how much
water. I'd like an informed view about how big a pot needs to be carried
into the field. The 2x2 binned stuff is more tabulated data to look over
(and to type), so I played with the pretty picture above first. The data will
be along by and by. Instead of using a 2 gallon basin for cooling, this time I used just a 1 liter jar. I turned on the pump and half the water disappeared into the tubing. I topped it off with ice and a little more water. The ice vanished very quickly, but the chip promptly dropped to -36.5C. It held -35C as long as I kept adding ice, a handful every 10 minutes or so. At -30C and at -25C the jar's appetite was noticably less. It would hold -25C for quite a while with just cool water (15-16C) and it would hold -20C essentially forever as the water heated up to 20C.
I decided to quantify "essentially forever" so I've been running a liter of
water through the camera for three hours to see where it all ends:
Starting water temperature: 23.5C. In 15 minutes, the CCD stablized at -20C. One hour later, the water was 25.5C and the CCD -19.5C. After three hours, the water was 26.8C and the CCD -19.1C. A control jar sitting beside the active one did not change. I shut down the pump, drained the water, and let the CCD stabilize without water assistance. The single stage Peltier cooling holds the chip at -15. So this relatively warm water is getting me only an extra 4.1C. That's barely enough cooling to be a realistic trial. I need to know how fast water that is cool enough to get me an extra 10C at 80% cooling capacity warms to the point that I'm getting those same 10 degrees at 100%. That would define a useful configuration. More experiments with cooler water and vessels of different sizes, please.
I'm thinking of using a 10-20 liter carboy and a buck's worth of supermarket ice.
Hook that up, get the CCD to -30C and 80%, then watch what happens as time goes by.
|