Cooled 550D

After having completed successfully some projects for cooling various models of Canon DSLR cameras, I decided to change my old 350D with a new 550D. The Canon 550D is a camera that incorporates significant improvements over the 350D, the most relevant for astrophotography are: increased quantum efficiency, lower readout noise, lower dark current production, live view and control of  all functions of the camera from the USB cable.

The main features of the cooling system that I produce are:

  • Temperature drop of up to 18º C below the ambient temperature. Considering that after several long exposures the CMOS sensor of the DSLRs heats up to 10º C above ambient temperature, this means that in reality the temperature drop in the sensor is up to 28º C below the operating temperature without the cooling system.
  • Active  anti-dew system on the sensor: when the temperature drops below the dew point, condensation may occur on cold parts that are in contact with the warmer air. So cooling the CMOS sensor can cause condensation on the low pass filter located in front of it. Heating it slightly, prevents this condensation if the ambient relative humidity is not extreme.
  • Passive anti-dew system in the inner parts of the camera: All internal parts which are cooled directly or indirectly are isolated with thermal foam, this improves the efficiency of the system and prevents condensation on them.
  • USB and HDMI connectors remain in the same place, only the 'bulb' exposures connector is moved to the plugs box located over copper heatsink / fan.
  • Remote control with digital display that allows regulating the working temperature of the camera and supplies power to it, eliminating the need to use the Canon batteries or external power supplies. The temperature regulation is especially useful if you want to do darks at the same temperature as the light exposures, or to moderate the temperature to save energy and increase the life of the Peltier element.
  • Compact and integrated with the aesthetics of the camera.

Below I briefly describe the process I followed to make the cooling system, including photos of some stages:

Once the camera is completely disassembled, the fun begins. A Canon DSLR camera has two filters in front of the sensor, the infrared filter and low pass filter. The first thing is to remove the infrared filter to let the camera be much more sensitive to wavelengths corresponding to the nebular emission lines H alpha, SII and NII. The low pass filter remains in the filter holder, this filter features the self-cleaning piezoelectric system, so this feature will still work after the modification. The next step is installing a heating system for the low-pass filter. A slight temperature rise in this filter prevents condensation of water vapor on it when the cooling systems is working. The photo below illustrates the filter cell with the heating system installed. The power to this system is supplied through the 2 thin wires protruding from the upper right side:

Once the CMOS sensor is mounted again in its cell, it must be assembled back into the camera. The sensor cell of the 550D is collimated by three Torx screws which fix the cell to the chassis at three attachment points. It is imperative to very accurately measure the height at each of the 3 attachment points before dismantling the cell so that when you mount the sensor back is perfectly perpendicular the optical axis.
Next is to put the copper cold finger attached to the sensor and for this we must make some modifications to the chassis of the camera.

Once in place the internal temperature probe on the cold finger just beside the sensor and the cable that supplies power to the camera is soldered, it is time to cover the cold finger and the adjacent parts with insulating foam:

At this stage the assembling of the camera can be started by placing the main circuit board. The "bulb" plug (2.5 mm jack) has been removed from this board to make room for the cold finger. Three wires are soldered to the board to relocate the connector in the external plugs box. All cables are grouped and protected with heat-shrink tubing:

Some more work and the camera will be fully assembled, showing the external part of the cold finger:

The next steps were to put in place the Peltier cell, the cooler fan and the plugs box. Once all the elements assembled, I proceeded to remove the front cover of the camera to remove the mirror. That mirror cause vignetting when using fast optics. Particularly f4 and faster optics cause a pattern that can't be eliminated with flats. The withdrawal of removing the mirror is that the measuring sensor gets invalidated, but this has no negative consequences for astrophotography. Here you can see the camera without the mirror:

Below there is a comparison of 2 darks taken with the camera, both 10 minutes long at ISO 1600, one done before modifying the camera (on the left), with an ambient temperature of 16 ºC (the exif information of the raw file shows that the CMOS sensor was at 23° C). The other one done after installing the cooling system (on the right), with an ambient temperature of 20° C and the camera working at 4° C (the exif information of the raw file shows the same temperature as the remote, 4 ° C). Both darks have been applied the same histogram stretch.
It can be seen the big difference in accumulation of dark current and the thermal noise associated to it. The improvement is very noticeable in the dark made​ with active cooling:

This page can give you ideas on how to initiate your cooling project looking for your own approach. Some have asked me about more detailed information, but unfortunately I don't have more detailed guides or tutorials about the process.