Engineering Question for Long, Digital ExposuresI'm thinking about a way to keep the

[TXForester]

I'm thinking about a way to keep the sensor in cameras cool during long exposures. Sensor heating is one of the sources of noise (that grainy look) that can be controlled to some extent. The easiest way to deal with it is do night time land/starscapes when the ambient tempts drop the temperature of the camera body, but that doesn't help during summer.

A cheap solution to summer temps is encase the camera in a modified cooler and put ice packs in with the camera. A high tech method is modify a cooler with Peltier cooling using a thermo-electric cooler (TEC). You can buy a cooler with a TEC system designed for keeping sodas and snacks cool on long drives, and modify it to fit the camera. You can buy a TEC and buy/make a cooler and put them together.

The common problem with the high and low tech methods is condensation. Condensation and electronics do not go well together. I think a venting system could help, but it would keep the system from reaching it's lowest temp for the camera body. Probably not bad if I can drop the body temp 15 to 20 degrees F. I could be wrong. To vent air from the cooler means warmer, water laden, air from outside would come in to replace it.

Any other ideas on eliminating or minimizing condensation?

 

[daveleo]

If you wrap the camera in plastic wrap, that will prevent "new air" from
touching the cold camera and condensing on it (it will condense on the
plastic wrap, not the camera body).
However, the air that is already inside the camera may condense if you
drop the temperature too far below the initial temperature.
How far below depends on the initial moisture content of the trapped air.

My first attempt would be simple: Put the camera in a modified styrofoam
box that has vent holes and a small fan moving the outside air in and out of the box.
Wrap the box in aluminum foil as a radiation shield. Put the camera under a beach umbrella.

Stick a cooking thermometer through the wall to monitor interior temperature.

EDIT : what was I thinking - - - skip the box. Use just a fan an the umbrella.

 

[semilog]

I love the Rangefinder forum, but this is probably not the best place to ask your question. You really want to be on an amateur astronomy site. Those folks routinely deal with all the questions you're asking.

That said, I have a friend, Tyler Nordgren, who regularly takes long exposures. He's an artist and also a professional research astronomer. Tyler uses a Canon 50D (IIRC) and although he sometimes takes and subtracts dark frames (to compensate for hot and cold pixels and pattern noise), he says that cooling the sensor is not generally necessary for the sort of work that's on his site. So you may be able to get away with not cooling the sensor below ambient, depending on the length of your exposures, your subject matter, and your technical requirements for the final images.

It can also help to average multiple shorter exposures. At low light levels, you're not collecting a lot of photons in the dark zones of the image, and in addition to accumulation of dark charge, read noise can be a problem. Frame averaging can help a great deal by smearing out the noise while preserving the signal.

I'd experiment with these approaches extensively before dealing with the hassle of cooling my camera.

 

[YYV_146]

I don't think most astro-photography requires cooled sensors. I have done exposures up to 30 minutes without much issue with my NEX-7, and if you intend to shoot for periods longer than that, there are greater issues than the sensor getting hot. Often a large number of short exposure are merged in post, which will eliminate most of the color noise.

 

[hkleinespel]

Why not use some small silica gel bags in the colder enclosure ? Or some other kind of humidity absorbing material ? (I think even rice works quite well...worth a try)

 

[iamzip]

I don't know anything about sensor heat, or if cooling is even required - but I am an engineer, and I can tell you that the most efficient method of heat transfer is convection. In other words, you need to get air moving around the camera. That should keep it at least close to ambient temperature. If you want to get it colder than that, you need to direct the intake air over some sort of heat sink, either ice or the peltier element.

 

[daveleo]

I don't know anything about sensor heat, or if cooling is even required - but I am an engineer, and I can tell you that the most efficient method of heat transfer is convection. In other words, you need to get air moving around the camera. That should keep it at least close to ambient temperature. If you want to get it colder than that, you need to direct the intake air over some sort of heat sink, either ice or the peltier element.

+1
which is why I think a fan (for convection) and an umbrella (as a radiation shield) is a good simple first shot at it.
Any "solution" that borders on condensation adds a whole subset of other issues to deal with.

 

[TXForester]

Thanks for the ideas.

 

[semilog]

Also: a Google search for "DSLR cooler" or "DSLR cold box" yields many links to relevant commercial products, DIY plans, and discussions of use…

Good luck -- please tell us how your project unfolds!