Second DSLR Cooler Box

Earlier this year, I created a Peltier cooler box for DSLR camera and lens.  The box worked very well but there were a  additional features I wanted:

• Preserve the lens "clamping" design of the first version
• Preserve the thermo-electronic hardware of the first version.
• Redesign the mounting for the liquid cooler radiator - first version had this as a bolt-on.
• Include the speed controller within the main box rather than as a separate module.
• Redesign electrical input to use a simple AC transformer
• Attach the cover with screws rather than rubber bands
• Allow use of a quick-connect for the camera
• Allow camera to mount in either portrait or landscape mode.
• Allow a future option of using the box with a large lens not supportable by the camera fitting.

The last two points required a complete rethink of the system.  As in the previous version, the bulk of the assembly is constructed of 1/2" Sande plywood.  This hardwood from Home Depot splinters much less than the comparably priced Luan plywood.

Camera Dimensions

I started by getting a reasonable estimate of the camera dimensions to. I came up with the measurements below. Though shape details are approximate, I tried to carefully measure the bounding-box for the camera and correct;y place the tripod threads and the lens fitting.

After much tinkering on the computer, I came up with a design in which the camera was attached to a L-shaped bracket rather than to the base itself. In turn, the bracket would be fastened into the box in one of two configurations, thereby positioning the camera in landscape or portrait mode.

Camera Bracket

For rigidity, both arms are formed by gluing two layers of plywood. Notches are cut out of the wood to accommodate a Manfrotto 323 RC2 quick release head in one arm and the USB cable and access door in the other arm.

This is how the brackets look once constructed.




The front view below shows how the bracket is intended to work.  Either arm of the bracket screws through the same holes in the base plate.  In either configuration, the camera lens aligns correctly with the front hole.

Clearly, the challenge is to mount the bracket on the base and to mount the camera on the bracket so that the above alignment actually works on the real thing.  Since there was no way I was going to get my measurements and cuts perfect, I left two points of adjustment in the design that I would finalize experimentally once the base and bracket were constructed:

• placement of the holes in the bracket for the fastening bolts
• thickness of the shim between the bracket and the quick-release

From the side, the bracket sits nearly flush with the front panel (I left a small expansion gap). I reused the 7" Losmandy plate.  After final assembly with Camera and zoom lens, the balance point was achieved with the plate mounted flush to the front of the box as shown here.

Here is the constructed bracket and base parts

Once the base and bracket were painted, I attached the quick-release block to the bracket. The additional foam behind the bracket provides additional insulation and braces my desiccant box in place under the heat-sink. The wire shown is a USB cable for the TemperHum temperature and humidity sensor. The wire goes down through a grove in the base. The cables for the camera USB and camera power cable also go down through the same groove.

Electronics Cover

As in the previous version, all of the electronics and cooling hardware are mounted into the top cover. Only the camera itself is attached to the base/bracket portion of the box.

The front elevation for the cover, shown earlier, is relatively straight forward. The only notable design element is the split front panel so that the base and cover form a clamp that seals around the lens. The side elevation, shown below, is more complicated. 

Unlike the first version, both the camera chamber and the liquid cooler are housed in the same overall box. A vertical center panel separates the two chambers and is well insulated with form.  This panel also supports the CPU heat-sink, the TEC chip, and the cold-plate/pump assembly.  

The motor speed controller board is also mounted onto the center panel, on the radiator side.

The radiator is now mounted vertically on the back panel with the inlet and outlet hose attachments at the bottom.  As in the first version, this orientation also ensures that, while the camera is pointing at the sky, the radiator is never in a position where the hose attachment points are uppermost.

Attached to the center panel, on the camera side are two blocks of wood composed of three layers of plywood.  A tee-nut is sandwiched within the assembly to receive the end of a 1/4"-20 thumbscrew.  This thumbscrew inserts through the base from below and holds the base and the cover together.

Here is the initial construction of the cover. Though not detailed in the drawings, this view shows some countersinking in the center panel.  This allows the cold-plate to recess into the panel. This ensures a tight sandwiching of the Peltier between the cold-plate and heat-sink base. 

Once I finished the painting, I glued sections of hobby foam on either side of the center panel. The hole in the foam is just large enough for the Peltier chip to sit in. This should minimize any air flow between the two compartments.

The electronics mount snuggly in the cover.  There are holes cut in each of the four sides of the back compartment to allow air to flow into the radiator.

The speed control board is now mounted internally so only two external lines are are needed, ground and +12V. The plug in the back panel is a 2.5mm jack for these power lines.

Larger Lenses

In this design, the camera body is attached to the cooler box and the camera lens is supported by the camera.  The foam clamping around the lens, provided by the upper and lower halves of the box, do provide some additional stability.  

This approach works well for smaller lenses, up to my 70-200mm/f4L zoom lens.  My wife has a larger 400mm lens that is better supported by its own tripod foot.  My approach to using a larger lens is shown below.

The bracket inside the cooler box is removed altogether.  The Losmandy plate is replaced with a longer 11" model which projects forward of the box.  A second Manfrotto quick-release is mounted to the front end of the plate on a riser block.  The lens is attached here and now supports the camera instead of the other way around. I consider mounting both the camera and the lens to be a bad idea as it would likely put stress on the camera's lens attachment ring.

Update 2013-11-03 - Testing at Canyon of Eagles - This weekend, we went to the Canyon of Eagles dark site and I tested out the box performance. Here is the box at the beginning of Friday evening. It promised to be a dew-heavy night, so I put heater bands set to 1/3 power on the lens and guide scope. Cable management is still a problem.  I need to route all of these more centrally as I was having some declination guiding issues which I attributed to the weight of the cabling.

On the second night, the temperature outside was 50F around midnight and dropped to about 35F by dawn. I did not capture a trace of the internal box temperature but did note a consistent 25-30F difference between external and internal temperatures during the early evening.

The images around midnight had an EXIF CCD temperature of 48F and that near dawn of 35F. As this is usually much higher than the ambient temperature around the camera, I am pleased with this result.

When we recovered the equipment on the second morning, everything was, in fact soaked with dew. The heater bands kept the lens clear, though. When I opened the box, things looked great in terms of moisture. There was none on the camera body.  A few drops worth of water had frozen onto the internal heat-sink and the desiccant box had completely changed color. Otherwise, nothing. I am very pleased with this result as well.