Hi
This post might be a bit OT. But considering how much effort we put into that signal to noise ratio...
!!! WARNING !!!
The following image could prove disturbing to sensible individuals.
http://pages.videotron.com/astrolo/Cool/Flake.jpg
Let start by the results. The following is a comparison of two identically
treated and displayed darks taken at 0C and -50C. Notice the shape of the
histograms. The standard deviation of the pixel stats. And finally the overall
esthetics of the frames.
http://pages.videotron.com/astrolo/Cool/darks.jpg
How to...
The principle is simple. But the lack of space in the camera makes part
selection critical.
http://pages.videotron.com/astrolo/Cool/peltier.jpg
At around -30C the optical window fogs up and ice crystals form on the CCD chip.
Heat wire and neoprene insulation takes care of the optical window.
http://pages.videotron.com/astrolo/Cool/Heat.jpg
A dry gas purge takes care of the CCD. Space is so restricted that I had to
install the intake needle valve on top of the desiccant plug.
http://pages.videotron.com/astrolo/Cool/purge.jpg
I did not want to be stuck renting high pressure tanks so I use Argon/Nitrogen
wine preserving gas.
http://pages.videotron.com/astrolo/Cool/wine.jpg
How cold can it go? No idea. I chickened out at -60C and the temperature was
still dropping like a stone. Anyways I can only reach around -50C before the
onset of fog and ice on the optical surfaces.
http://pages.videotron.com/astrolo/Cool/Close.jpg
I can't wait for the next clear night to see how the new results will compare with my previous TT ari time series.
Thanks
Juan Herrero
http://pages.videotron.com/astrolo/
That is some souped up data collection machine! I can't wait to see the results. Those ST9 cameras are really workhorses.
Please keep us informed.
...Tim
Nice work! Looks like you'll get near LN2 results. This would freak-out any "puriest" who doesn't like any kind of condensation or ice on their equipment.
After using an LN2 cooled camera at a professional observatiory, we compared bias frames (0 sec.'s) to a dark frames shot for 7min (420 sec.'s) exposures. There was virtually no differenace in noise count btw the two using LN2 as the coolant!
James Foster
Los Angeles, CA
Folks,
Just to be clear - professional CCD dewars are cooled by liquid nitrogen (LN2) but aren't operated at LN2 temperature (77K = -196 C). In fact, the chip is usually kept at -140 to -120 C. One of the design considerations in a dewar is how well to couple the LN2-cooled inner surfaces to the surface with the CCD chip which is exposed to the outside thermal environment by radiative coupling. You want enough thermal conductivity to balance the heating and cooling influences and keep the chip itself in the range -140 to -120 C where dark current is negligible and the signal-to-noise ratio is optimum.
Cheers,
Doug
To Doug,
Yea, I didn't mean to imply that Juan should fill his coolant lines with LN2 directly touching his CCD. As I understand it the professional CCDs are insulated from the environment with either a vacuum or argon mixture. The only thermal contact the chip makes with the coolant is through a copper "cold finger." At least this is how we ran our LN2 camera on JPL's 0.6 meter instrument.
James
Commercial sensors and signal chain electronics are not designed for use at "extreme" temperatures. Yes, the dark current decreases, especially for the hot pixels. At -100C, a typical operating temperature for science CCDs, you have essentially zero dark current and I don't remember seeing a hot pixel.
However, for commercial CCDs, several other effects start happening as you go cold. Probably the one most prevalent is a loss of Charge Transfer Efficiency (CTE). Look at the first UCAC paper (Zacharias et al., 2000AJ....120.2131Z) to see how they had to run their Kodak chip.
Basically, you are running things in a regime for which they weren't designed. You need to thoroughly test your system to see how well it performs at colder temperatures.
Arne
I've learned some interesting points regarding the suitability of different gases for dry purging.
As i mentioned before I did not want to start renting high pressure tanks, so I looked for alternative consumer solutions.
First I investigated common Dust Remover Aerosol. But the Flouromethane gas used has a boiling point of only -46C. That's well within my camera's temperature range. Since I did not want to turn the camera into a micro-climate experimentation chamber with storms et al. I did not even try it.
But I did try Party Balloon Helium. With a boiling point of -269C it should work great! Right?... WRONG! First, The CCD could only reach -50C at 100% power, when normally it can hold -50C at 47%. Second, the optical window fogged up right away even with the dew heater on. What was the problem? Well... it turns out that Helium has the highest thermal conductivity of all gases. 6 times greater than air. The gas was sinking the "cold" from the CCD chip and transferring it to the optical window!
So I ended up using the Argon/Nitrogen mix which has the same thermal conductivity as air. It works great to fend off ice from the CCD. There seems to be a time factor in the purging process. As the weeks pass I can reach lower temperatures before the onset of ice. The weak point remains the optical window. It fogs up by -50C depending on the RelHum.
I want to experiment doubling up the window, making a thermal optical window? A Xenon gas purge should work great, it conducts 6 times less heat than air. Finally for $260 I could get a two stage pump capable of reaching 5 Torr. A vacuum comparable to light bulbs and Thermos glass flasks. Now... How many Torr can that optical window withstand?
Juan