Slashdot Mirror
Cooling a Digital Camera?
thusson asks: "I work at a lab doing intrinsic signal imaging of cat and mouse visual cortex (brains). We are using a Dalsa 1M60P camera, and we want to cool it about 30-50 degrees C to improve picture quality by reducing noise. Does anyone have any suggestions about how to do this? So far, heat sinks have helped but only by a few degrees. I figure the overclocking community is a good place for novel ideas."
What should you be cooling? How are you measuring your temp? Answers to these questions will drastically change the answer.
=================
Unix is very user friendly, it's just picky about who its friends are.
30-50 degrees Celsius?
#1 - Is your claim valid
#2 - Will the camera work at that temperature
#3 - Try some liquid nitrogen. Or, some spray dusters turned upside down.
A lot of people people use these to cool there optical sensors for telescopes.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
Or so I'm told.
I've hit Karma 50 and gotten a Score:5, Troll... I win!
we want to cool it about 30-50 degrees C
So far, heat sinks have helped but only by a few degrees
What makes you think that heat sinks will lower the temperature by 30-50 C? At best, heat sinks will lower the temperature to that of the surrounding room.
If you want something cooler, you'll need to use Dry Ice or refridgeration or some external input.
Peltier elements are the way to go here. They're small, accurate and powerful. Anything you want in a scientific environment.
I do research in a neurobiology lab and for one of our setups we water cool our camera. I just looked at it now (it is not my setup), it seems to run water from a nearby pipe into a specially made heatsink mounted to the back of the camera and then outagain towards another pipe.
30-50 degrees Celsius is above room temperature (~22 deg C). I highly doubt that your lab has an ambient temperature above 50 deg C. Fahrenheit temperatures don't any sense either, because you would begin to condense water vapor on the camera in the 30-50 deg F ranges as well.
Would you like to re-write the question so it actually makes sense?
As a number of people have correctly pointed out briefly, Peltier effect devices, also known as thermoelectric coolers (TE coolers) are the most likely way to accomplish this. There are companies (Products for Research, e.g.) that specialize in pre-packaged TE cooling units for various types of equipment. The basic coolers almost all come from Melcor.
With very careful design and construction, you can use a multi-stage TE cooler to get more than a 50C offset from room temperature. It requires careful attention to insulation, and to heatsinking of the first (hottest) stage.
Before you go cooling your camera, though, you should check whether it is designed to run cold. Any device, cooled to -30C (or even to 10 degrees below room temperature, in Tennessee humidity) will start to condense water. To run a camera cold, it must be designed to be water-tight, and must almost certainly have a dry-gas-filled double window (or vacuum double window) on the front, with the outer window heated to prevent fogging.
Overall, unless the camera was specifically designed to be cooled, you may be better off buying one that is appropriately designed, unless you have a lot of time, money, and expertise to put into the engineering. Certainly talk to the manufacturers of the camera you have before you start cooling it, and see if they sell a kit cor conversion.
Peltiers are probably your best best. You'll need to mount heat sinks on the hot sides to radiate off the waste heat, but you can get a good 60-70 C temp difference between the hot and cold sides.
There's a good example here
Even people that believe in pre-destiny look both ways before crossing the street.
The astronomy folks are way ahead of you. Google and ye shall find.
"You might as well get your son a ticket to hell as give him a five string banjo." -unknown minister
Numerous companys make scientific-grade cooled CCDs for astronomy and microscopy. One of those is probably what you should be using.
He didn't think sinks would make it go 30 - 50 deg cooler. He said they helped make it cooler than without heat sinks. Reread the sentence, you misunderstood it.
Photos.
But I will try:
There are two solutions to this problem. The first, is to cool your sensor like you asked. A peltier cooler in between the sensor and the heatsink is the only way to do this. A heatsink will reduce you to ambient temperature at best. If you want to go below that, you need a peltier cooler at least, or something much more exotic like a compressor-based refrigeration unit, evaporative cooler, or liquid nitrogen cooling.
Alternately, you could use a CCD with a cooler already built in, such as those from Santa Barbara Instrument Group.
Finally, you could simply use a sensor that isn't so noisy. All the digital SLR cameras nowadays use CMOS sensor technology, because it's bigger than CCD primarily, but it also has a lot less noise since it isn't crammed into such a small space.
Another approach often used by astronomers is to take the noisy CCD, do an exposure of a given length, then cover it so no light can reach it, and take another exposure of the same length to create a "dark frame" that contains nothing but noise. The noise in both images will be approximately equal, and can be subtracted out using photoshop or similar software, resulting in a very clean image.
Random and weird software I've written.
Stick it in a biiiiiig block of ice. It's a perfect solution!
Unlike most applications in which they are misapplied, Peltier Coolers are ideal for the purposes of cooling CCDs and thermal imaging devices. To achieve a 30-50 degree C cooling you may have to use a 2 or 3 stage configuration. In multistage configurations you employ more coolers for each stage to compensate for the waste heat generated by the previous stage. Thus in a two stage design one junction is employed in the first stage and thermally coupled (use plenty of thermal goo) to four junctions in the second stage and 9 to 16 in the third stage backed by a good size heatsink. The entire assembly should cool in the 30-50 degree range. Of course, this will result in condensation or even frost so be sure to encase the CCD in a dry nitrogen overpressure chamber.
Why not just throw it in the freezer for a few hours before the experiment. As long as the experiment isn't too long, you should be able to take it out, use it, and return it to the freezer for the next round.
It's good to use your head, but not as a battering ram.
It is a Peltier device. It's real. It works. One side gets hot, the other cold. It's just a solid state heat pump. Do your own googles, try "thermoelectric" if "peltier" doesn'te net enough, for further info.
Good judgement comes from experience, and experience comes from bad judgement.
- W. Wriston, former Citibank CEO
More specifically, I'd consider installing a ductless split air conditioning system. Mitsubishi makes many different types of them. As the name would imply, they don't need ducts and can simply hang on the wall or from the ceiling.
If the ceiling is too high, you could install a duct in front of it to force the air downwards. Only downfall is that this is an actual air conditioning system, which means you will need to have the condenser fan installed outside, as well as the reefer lines to said condenser.
They aren't cheap, but are also not that expensive. You could probably get one installed for under $2000. Depending how far away the condenser would be, mind you.
Is that a real poncho? I mean, is that a Mexican poncho or is that a Sears poncho?
Not sure if it will help with a digital camera but I would suggest getting this book:
CCD COOKBOOK by RICHARD BERRY
it's available here
http://hardinoptical.com/ccdcookbook.html
and universityoptics.com has most of the components
good luck
Ok, so what you're doing, let's just take a wild guess here, is clamping a cat's head so that it cannot move. Then you're removing the back of the cat's skull, so that you can get at its occipital lobes. Then you show it pretty pictures. The classic fun would be to measure neuron firings invasively with insanely thin electrodes.
/.) and some propylene glycol. Or would that be too cold?)
(To quote Dr. Hugh Wilson, "Well, what else can you do with a cat?")
While I sometimes agree with Hugh, I don't quite get what you're looking for with the camera. Why cooling?
Maybe you'd give us a few lines about what you're doing? What's the method, and what are you looking for? Inquiring minds want to know.
(As for cooling, just use whatever system you're using to make ratsicles, along with a Jefi (see July 17's
Many people have mentioned that you can use peltiers to cool the sensor below room temperture, This is probably the most cost effective way to do it, but since this is /., there's another really cool way to do spot cooling. If your lab has a clean source of compressed air, check out exair.com. These things are called vortex tubes, and can create incredible temperature deltas, that are ideal for spot cooling, provided that you can supply the amount of air it requires. I've been told that it's actually relatively quiet given how much air it consumes, but i have no personal experience with them. just make sure you keep the exhaust end away from anything you don't want to melt. I woulda used it to make coffee while cooling my processor, but not everyone can afford a pump and a tank that can maintain that kind of airflow at 100 PSIG. Oh, one other thing... have you considered the potential issues w/ cooling the chip to those temperatures? I'd imagine condensation would kill your image quality, and depending on how well the boards are made, the solder joints may develop stress fractures. I remember using cold spray to cool PCB's to make sure there aren't any cold solder joints on PCB's we test. Oh yea, you could also use cold spray instead of the votex tube, but something about that stream of hot exhaust seems strangely apealing.
Chop the cooling system out of a cheap old dorm fridge, install the cold part in an biggish old tupperware tub (that has room for the camera too), cut a hole in the tub's side & install a little 2-layer window of optical quality glass for the camera to see out through, and either dry & seal the space 'tween the panes or rig a couple tubes to blow heated air through there (preventing condensation either way). Then cut holes in the tupperware top, run your camera controls through the holes, and insulate the whole thing with spray-on foam.
If you don't have anybody who can be trusted to fill in the details and do this safely, then spend money on something pretty.
It's easy to make up & spread cool- and credible-sounding stuff. Finding & checking hard facts is hard work.
There has been research done in image based rendering and computer vision that deals with removing noise due to heat. Take about a thirty second exposure with the lens cap on. This will give you an image with only thermal noise. Now you just have to take your images and subtract some constant times the noise image to get the denoised image. There are a few algorithms out there for finding a good value for the constant.
My father is an amature astronomer and bought a CCD camera kit this Richard Berry guy put together. The CCD chip is mounted on an aluminum block that fits inside an aluminum "eyepiece". There is a peltier device mounted (cold side) to the bottom of that. Then the hot side of the peltier has another aluminum block mounted to it where water flows through.
Just go lookup "Richard Berry CCD camera kit" in google and you could probably find something.
You are free to do as we tell you.
We want your soul.
www.wewantyoursoul.com
I'll ask a second time, and I suppose I'll get modded down again, as apparently, Slashdot just doesn't want to know.
WHY DO YOU NEED TO COOL THE CAMERA?
I have done this sort of thing to reduce the noise level of CCD's. Peltier type devices are fine if you have access to the CCD unit. If you are using commercial digital cameras, the CCD is not accessible for direct cooling.
./ a link to some of the before and after picture.
In my lab, I keep the camera in a zip lock bag in the freezer. When I want to take pictures, I take it out of the freeazer and the bag and attach it to my microscope. This method will let take the pictures for several minutes, before the camera warms up. If you employ this method, watch out for condensation on the lens!
If you manage to get this or some other method working, do post to