Accurate Temperature Controlled Peltier Fridge

[avandonk]

As some of you are aware I built a Peltier cooled fridge some time ago to cool my Canon 5DH. Below is a plot of ambient temperature and temperatures inside the fridge from a warm start. It is obvious that as the ambient temperature changes the fridge temperature will follow.

I have ordered a Peltier temperature controller from here

http://www.tetech.com/Temperature-Controllers/TC-48-20.html

and it is arriving today. I will post up results later. The main aim is to get a constant temperature inside the fridge so all darks and lights are perfectly matched in the temperature they were taken no matter what the ambient temperature is doing. This is the main reason why ICNR works better than any other method. If the temperature of the camera does not vary by about 0.5C dark subtraction should now give better results than ICNR if the ambient temperature changes quickly over an imaging run.

Also all you folks with cooled Canon cameras and QHY8 cameras should be able to control their temperatures as well. They do have cheaper versions without the display and computer control.

My guess is that with care the temperature would not fluctuate by more than 0.5 degrees C. The controller is capable of holding the temperature at about + or - 0.1C variation. This of course would depend on the implementation.

Bert

[peter_4059]

Thats a couple of impressive peltier coolers Bert. I imaging the temperature controller will be a big improvement. How much does the whole thing weigh with the camera inside?

Also how does the cockroach in the 3rd image fair at -20 degC?

Peter

[avandonk]

Peter it weighs just on 2kg but it is supported independently of the camera so any minor vibrations (by cooling fans on the heat pipe coolers) are damped by the lack of direct connection to the optic or camera. In other words the whole mount needs to vibrate before the optic or camera is affected. Cockroaches will survive nuclear fallout and this one snuck in during the day just for the pic. He was gone before the next ice age.

Bert

[iceman]

That looks fantastic, Bert. Have you got any dark frames taken with the cooling on/off to compare?

[avandonk]

Here you go Mike.

http://www.iceinspace.com.au/forum/showthread.php?t=32683

I have played with the new thermoelectric controller and will carefully wire it up tomorrow. It is a microprocessor controlled integral and differential controller. What that means by adjusting the parameters it will quickly go down to the set temperature and hold it there without any oscillation. Why reinvent the wheel when it has been done by experts.

Bert

[avandonk]

Set up the fridge with the Peltier controller inside the house with the doors open so ambient was at about 12C. I then shut the doors at 20 minutes on the graph and turned on the heater. The controller sensor is mounted on the Peltier coldplate inside the fridge. Then plotted the results.

I have not optimised the settings for integral and differential yet I just went with the default.

I think it works very well.

Bert

[MrB]

Has the PID got an option for self-learning? makes life easier, though the temp looks pretty stable already!