Fluid cooling is generally acknowledged as the most efficient way to cool a CCD camera. The default air cooling via the normally supplied fan is vibration prone and highly inefficient. The best I was ever able to get with my single-stage cooled camera was a differential of around 35°C and that was at almost at 100% power level. Living in Tucson, it is a challenge to keep a CCD camera cool during the hot summer months, where the evening temperatures might not fall below 25°C. Last year, I went with a fountain pump, 4 gallons of water, and some ice to get things going. Even then, the best I could muster was -10°C for the camera operating temperature. Clearly, some improvement was necessary.
I experimented with thermoelectric drink coolers, but the capacity was sorely limited. It was time to do it right. Some work by Doc Greiner on the cooler for his Canon EOS 10D caught my eye. After a lot of correspondence and help from Dick, the RAG Cooler was born. In recognition of his freely shared help and information, this cooler is named for him.
I use this cooler with my ST-10XME, which has a single stage cooler. With an ambient of 26°C, I can get the camera to -20°C with the camera cooler at 80% and the auxiliary cooler at around 16 volts out of 24 volts. I’m controlling the coolant temperature at around 13°C. Of course, the camera fan is off. My current cooling differential is 46°C with considerable cooling reserve. I suspect more cooling is possible up to a 50°C differential but I decided not to push it. I wanted to stay comfortably above the average dew point.
Figure 1: Overall view of the cooling system.
Figure 2: Key components identified.
The system is mounted on a 12" x 17" piece of scrap plywood. Fluid control consists of a TE Cooler to chill the coolant and a recirculating pump with an attached reservoir. The output line carries the cooled water up to the camera via a nominal 15 foot run. The return line brings the water back to the pump. The homebrew power supply consists of two 12 volt 6 amp switching power supplies that I had laying around. These are connected in series. In the upper left of the chassis can be seen the TE controller. The controller senses the cold plate temperature via a supplied thermistor and adjusts the power to the cooler via a switching regulator. The operating temperature is settable via 10-turn potentiometers on the controller board. Once adjusted, the entire system is automatic. The only maintenance is to check the fluid reservoir level from time to time.
Figure 3: Thermoelectrric cooler
This is a 61 watt thermoelectric cooler, TE Technology model LC-61, that operates as a heat pump to remove heat from the water circulating in the cold plate. The Thermoelectric junctions "pump" the heat to the heat sink where it is dissipated in the air by the cooling fan. The terminal strip is for the fan power and the TE power. The control thermistor is attached to the cold plate by aluminum tape. The cooler sits on some foam rubber to insulate the cold plate from the ambient to minimize cooling loss.
Figure 4: Pump
The pump is an Eheim 1048 aquarium pump. It has the advantage of being very quiet and adds minimal heat to the pumped water. It is specified to lift a column of water up to 4' 11". More powerful pumps are available and should probably be used if you have to lift the water higher. The reservoir, called a "Tank-o-matic plug-on" is mated directly to the pump. Both parts are available here. The white connectors are quick disconnects and are a wonder for this kind of application. See Figure 6 below.
Figure 5: TE Controller
The TE Technology TC-24-10 controller is capable of controlling up to a 10 amp TE cooler load. The electrical connections are for the supplied thermistor at the left and the power input and TE Cooler output. The controller should be mounted to a heat sink for highest reliability. Here, it is bolted to the base plate of the power supply enclosure. The 10-turn potentiometer in the lower left corner of the board controls the cooler set point.
Figure 6: Quick Disconnects
These magic parts make a liquid cooling system manageable. Consisting of two mating parts, the connection can be broken with a minimal loss of coolant fluid, typically a few drops, by pushing the metallic button on the lower part as shown above. With these disconnects, fluid systems can be broken apart as easily as normal electrical connectors. They come in a variety of sizes, shapes and materials. The pair shown above consists of part numbers A-06360-80 (upper part) and A-06360-65 (lower part) and are available from Cole-Parmer. Although a bit expensive, they certainly make removing the camera from the OTA a lot less messy. Here, the upstream connection is to the camera via short pieces of 1/4" tubing.
The tubing used is 1/4" ID, 3/8" OD C-flex tubing, part number A-06422-10, also from Cole-Parmer. It is very flexible and care must be taken in the cable routing to insure it doesn't get pinched.
I am currently using distilled water as the coolant but will ultimately use a 50-50 mixture of ethanol and distilled water.
First off, this is not an inexpensive system. It does indeed cost something to do cooling correctly. The overall cost, not counting the 24 volt power supply and enclosure, is around $800 US. The advantage is a maintenance-free system, temperature-regulated coolant and lower camera operating temperature. The ultimate cooling differential appears to be at least 50°C. In the final analysis, I believe it is worth it. For my camera operation, I can operate in the warm weather sufficiently cooler to reduce the dark current by a factor of 4, and the attendant noise by a factor of 2.
And best of all, the ice is available for the drink of your pleasure, and not wasted on camera cooling.
The RAG cooler would not have been possible without the considerable time, expense and knowledge of Dick Greiner. Once again, as he has many times in the past, "Doc" Greiner continues to add to the knowledge base of this fascinating, frustrating and demanding hobby. Thanks, Dick! Any errors are of course mine.