Camera Peltier Cooler


[Gary Honis] has been modifying his Canon Digital Rebel XSi in order to do astrophotography. He previously removed the IR filter and replaced it with a Baader UV-IR cut filter that lets most infrared light through. However, in order to reduce noise in the pictures, he had to cool the camera down. He based the project on a peltier cooler that he salvaged from a powered beverage cooler. He made a small aluminum box and insulated it with styrofoam to hold the camera body. The peltier cooler was then attached on the side. It takes just over an hour to cool the camera down to 40 degrees, but the shots come out a lot clearer.

[thanks adam]

36 thoughts on “Camera Peltier Cooler

  1. I am looking at the images on his site that he took. Can you see them that well with your own eyes through the telescope? I have always wanted to get into this but it just feels so unreal when I see the images people can get at home.

  2. Instead of waiting an hour, why not put the camera in the freezer first, then use the peltier to keep it cold during the shoot? Might shave half an hour off cool-down time.

  3. [quote] (watercooling + peltier can easily go bellow -20C) [/quote]


    The freezing point of water is 0° Celsius or 32° Fahrenheit.

    So unless you are using the camera at some bitchin atmospheres water cooling cant get to -20

    1. Peltier devices have a hot side and a cold side. The cold side has no moving parts, and is in contact with the sensor (or the air stream that carries the cooled air to the sensor). The water removes the waste heat from the hot side of the Peltier device, so the water is always well above freezing!

  4. Luke, no they do not look that way in the scope. To get images like these you have to use a long time exposure. The scope will track the object in order to account for the earth’s rotation.

  5. samuel, the water cools the hot side of the peltier cooler. The water never gets near freezing, but the more effective removal of heat from the peltier allows the cold side to get more cold.

  6. “The freezing point of water is 0° Celsius or 32° Fahrenheit.”

    Very true. Unless you add this stuff called antifreeze. Then it’s more like -27F (-33C). Generally speaking, the term “water cooling” doesn’t mean straight water.

  7. There’s no way a single man with a dslr camera, a telescope, and miles and miles of atmosphere can get pictures like that.

    Where’s the pictures from observatories that surpass this quality?

  8. Gary is a god amongst amateur astronomers, the point of the peltier is that it is small, Astronomy mounts have a weight limit, getting water cooling involved is more effort than is really needed.

    Getting down to -27 is fine and all that but its not about getting it as cold as possible, its about getting it to a point where you can control the temperature, consistently!!

    If you start taking long exposure astro images at different temperatures then you need all sorts temperature matched frames to deal with the noise. Astro imaging isn’t just point and click, You need the actual images themselves (known as lights or subs) then dark frames, flat frames, bias frames, etc.

    @ _matt, There is every way that that man with a dslr can take pictures of that quality :) They do it regularly too. check out

    You might have to go back a couple of months as its not really the best time of year in the UK for astronomy.

    I have a similar setup to Gary’s (- peltier) and have taken my own long exposure astro shots with some degree of success. There is a technique used called ‘Stacking’ where you take a set of raw equal length long exposures (sub frames or subs) of your target, you then run the subs through software (Deep sky stacker or maxim dl) which ‘stacks’ the images on top of each other, doing this increases (in theory) the SNR and also has the bonus of removing artifacts like satelite/plane trails from your final image, the resulting image after post processing (levels and curves in PS) also has far more detail than any single sub on its own.

    I’ve been into astronomy for just over a year, its a fantastic hobby all round involving physics, robotics, electronics, computers and imaging and hacking. On a personal level I have built an eletronic focuser (555 timer based) and a Dew heater controller (555 based again) and have a few arduino based mods that I’d like to build too (current one is an automatic dew heater controller that keeps my scopes at a constant temperature using some LM35Z sensors).

    sorry for the length of this post, I’m passionate about the subject, I’ve been visiting hackaday for yeats but it was astronomy that finally made me pick up a soldering iron and start with electronics :)

  9. @samuel: As others have mentioned, most water cooling systems use some kind of anti-freeze coolant mixed with the water. It has the added bonus of being able to hold more heat per volume than water too (or so I’ve been told).

    @_matt: I’m not, personally, into this stuff but I’ve orbited around their community due to the really cool hacks they have been doing with software and digital cameras. After seeing what I’ve seen, I can believe that those pictures were home-made. There’s a reason why the impact detected on Jupiter yesterday was found by a hobbyist before NASA/ESA/etc. Some of these guys are INCREDIBLY dedicated and down-right brilliant. They’re willing to (like this guy) modify expensive digital cameras, buy/build the highest quality private telescopes, set up extremely long shutter rates, and implement some of the most complex atmospheric distortion compensating algorithms in order to produce the pictures you see. I rank amateur astronomers up there with a select number of hobby communities like Amateur Radio operators and Amateur rocketry enthusiasts where the top tier of the craft have shown themselves to be capable of accomplishing things that even most geeks wouldn’t think possible outside a corporation or government agency. If anything, I think the amateur astronomers even have a leg up on the other two hobbies as, at the very least, they don’t have to worry about, nearly, as much government regulation of their chosen hobby. One last thought, if you check out the caption besides a number of the pictures, it suggests that he is, also, going to special locations (with, I’m assuming, much lower light pollution) in order to take his pictures).

    As for this project, I’m impressed by his results. Of course, as an armchair quarterback, I have to chime in with suggestions for how he might be able to improve the efficiency:

    * A larger heat sink on either (or both) side may help.

    * More importantly, he may want to replace the Styrofoam with some of the pink Dow-Corning rigid foam insulation you can pick up, extremely cheaply, from any Lowes/Home Depot store. The stuff he’s using right now looks like it might be the cheap stuff I got from Home Depot when I needed to block off the space to the right and left of my window air conditioner. I only went with it because I was feeling cheap that day and I wasn’t too worried about paying for a few extra kWh of electricity. For as sensitive a job as what he’s doing, I would think that he’d want every edge he could get. It comes in multiple different thicknesses and even the same thickness as the present styrofoam should provide much higher insulation values due to the fact that it’s specially designed for that purpose.

    * Also, it looks like some of the edges of the aluminum box are showing through the styrofoam. He, probably, should make sure that the foam overlaps it’s self at the edges and that any other visible aluminum gets covered as any visible parts of the aluminum can be a bad source of heat gain. He may even want to go that extra step and look into using a small piece of foam pipe insulation (which can, also, be picked up extremely cheaply from Lowes/Home Depot) to wrap the first few inches of the telescope eyepiece where it connects to the camera box.

    * He could, probably, further cut the time it takes to cool by removing the extra components he doesn’t need from the camera (the casing, the flash, mount the camera battery remotely, etc.). In the case of re-locating/rewiring the battery, this could be especially important as a battery being drawn from represents a significant heat source even closer to the image sensor than the Peltier system. An added bonus might be longer battery life as most batteries don’t work well in cold environments. Of course, if the removes the casing, then he’d have to be more careful about the potential for condensation on the electronic components. This might be avoided though the use of cheap silica desiccant like the stuff you find in the bottom of new shoes. Even if he didn’t want to crack open the camera’s case, he should still be able to wire up a battery mock up that allows a wire out to an externally mounted battery.

    * While it may not improve cooling, he may want to consider trying to put a piece of fine wire mesh (like bug screening, but being careful to make sure it’s made of metal) between the inside fan and the camera. That fan on the cold side of the peltier set-up is very close to the camera and those kinds of simple motors are notorious for producing lots of electromagnetic interference. He may be introducing a new source of noise for the image sensor while trying to eliminate the noise caused by excess heat (though, obviously, the work he’s done has, definitely, produced a net positive).

    * An obvious solution would be to buy a second cooler, construct another peltier unit, and mount if on the opposite side of the camera but, I’m sure, that’d add a lot more weight.

    Like I said, I’m just armchair quarterbacking here but if any of these ideas help him, or anyone else, refine his technique even further (assuming he’s even reading these comments) then I more than happy to throw them out there.

  10. Hehe, looks like Reggie made many of my points just before I posted. BTW, I think it’s also worth pointing out that the instructions for removing the IR filter on the Rebel camera should also be very useful for anyone crazy enough to want to spend that much money on a high resolution IR-based night vision camera similar to the hacks people have been doing to web cams. I’m sure any regulars probably already realized that but I figured I’d mention it anyway. Of course, at the price of those cameras, it might be cheaper to just buy real night-vision equipment…

  11. The webcam hacks are a very simple and cheap way to get into astrophotography, IR filter removal is the simplest of all to do (and necessary), just remove the lens, the IR filter is usually attached to the back of the lens, the more interesting mods are long exposure mod and amp off. anyone that is interested should look for toucam long exposure mods on google, the one I modded for long exposure was a philips spc900nc, I believe there is some work being done on i2c control of these devices….

    The IR filter is replaced in Gary’s camera with a Baader filter which does still filter IR but at narrower bands so that the juicy stuff (Ha, Hydrogen Alpha) still gets through and the grotty stuff gets blocked. So removal would be useful in IR photography, replacement useful for astrophotogrpahy + it also makes the cameras autofocus work correctly for ‘normal’ photography.

    Whilst Gary does visit special places to take some of his images it is entirely possible to capture incredibly good images in light polluted skies through the use of LP filters or narrow band filters, the decent ones don’t come cheap, a hutech (borg) LPS IDAS filter will run you around £160, with narrowband filters being a similar price and most narrow band imagers will want 3 (Ha, OII/III, S2) so they can do the ‘hubble pallette’ images :)

    @ colecomanm, the significance of the filter replacement is probably more than the peltier cooling, it increases the cameras sensitivity to Ha by 5x, this means 1/5th of the exposure time is needed to capture the same amount of data, this has many benefits, less noise in the image, less heat build up, less sky glow noise etc.

    As for the guys worrying about the cost of a DSLR camera should take a look at the price of ‘amateur’ astro imaging CCD cameras, for a sensor the size of the canon it is ridiculous amounts of money.

    I would say that 75% of amateur astro imagers haven’t got the funds (or justification with their partner) for a full on astro cam, so they get a DSLR as it has the added bonus of being good at normal photography, + when funds allow at a later date the filter replacement can be done either DIY for the brave or £200 from guys like

  12. Quote:

    “@ colecomanm, the significance of the filter replacement is probably more than the peltier cooling, it increases the cameras sensitivity to Ha by 5x, this means 1/5th of the exposure time is needed to capture the same amount of data, this has many benefits, less noise in the image, less heat build up, less sky glow noise etc.”

    @Reggie: I would, certainly, have thought so after seeing his graph showing the difference between the pass spectrum of the old filter vs. the new filter. My suggestions were intended to be in addition to the, already existing, filter swap.

  13. Hi Colecoman, your suggestions were all cool (pun intended :D ) no issue with any of them at all, I was merely making the distinction between removal and replacement in my post for clarity :) Both are valid methods for astro work, whereas the replacement wouldn’t be so good for IR work.

  14. This is awesome, I remember a few years back modifying a parallel port QuickCam for thermal imaging. Nothing fancy but it does work for higher temperatures. A company I worked for at the time was having problems with warping of metal after welding. I used the camera with the IR filter and colled down to show them heat transfer and develop a better “jig” and welding patern to keep warping from staying “warped” because the guys would weld the seam AFTER it has already shifted. Then when it would cool down. Bam… one bent and warped object. Any way. great work, I didn’t realize it could be used for this purpose too…..

  15. not sure what the bayer filter does for things apart from add colour, not sure where abouts that is in relation to the sensor, but I’m guessing as there is no mention of removing it for DSLR and astro imaging then there is nothing to be gained from trying, as I’m sure someone would have by now :D

  16. I don’t get why there’s a fan on the cold side, it supposedly is to blow the hot air on the peltier and keep the air moving, but since the thing is connected to a aluminium box surrounding the air inside on all sides that seems a bit pointless and the fan creates heat itself operating (you can feel its heat when you touch the back of the center when you a fan is operating) so it’s more a disadvantage I’d think, but maybe it spins very slowly? I guess you could use one that spins at like 500RPM or something to slowly circulate air without getting too hot itself like a mini-ceiling fan.
    Perhaps if the box was much larger than the camera it might have made more sense though.
    Also I’d look into adding some copper into the mix, it’s much more efficient than aluminium and might help to avoid having one side much cooler than the other, a few strips of copper to aid the aluminium.

  17. @Reggie it’s explained with sample pictures, when you remove the internal IR filter the colors go all off since the camera expects the normal visible spectrum, so to compensate when you take normal pictures with the camera you have to either use manual whitebalance and point it at a reddish card or use the simpler method of using an external bayer filter that does the same thing as the filter you removed namely stripping the IR out, that way you can still use the camera for normal photography but can simply unscrew that bayer filter to do IR photography, and put it in the cooling box for astro IR.
    The bayer doesn’t add color, it removes IR.

  18. @wwhat: At the bottom of the step-by-step construction page, there is a link called “Correcting Color Balance”. On that page you’ll find a graph comparing the color spectrum of the different filters. Check out the difference between the old Cannon filter vs. the Baader. Both filters block IR but the the Baayer allows much more of the visible spectrum through. Basically, they both do the same thing but the Baayer is just a higher quality (and probably more expensive) filter. This, probably, isn’t all that noticeable if you’re just taking normal pictures (which the camera was designed for) but, I’d imagine that, in astronomy you want as much of the visible light to get through as possible. So, in this case, the Baayer does end up adding color.

  19. No the filter isn’t used when doing the astro stuff surely, the whole project is about getting the IR in a clean no-noise way, you don’t then add a filter again, the filter is for when you aren’t doing astro as I get it, so the camera is still usable.

  20. Seems I was partly wrong, they use the baader UV+IR filter to cut out the IR+UV in telescopes so the sensor gets more sensitive to the visual spectrum and thus gets more color in too.
    Here’s the guy on his experiments with the baader filter and the reasoning:
    The site baader uses and linked in this hackaday bit seems down though, but thy have this one (in german)

  21. the hackaday summary incorrectly says it cools “to 40 degrees”. The article actually is describing a “temperature drop” of 40 deg F.
    So for example, if the ambient temperature is 50F, the camera can be cooled to 10F.

    Re the “Bayer filter”, it is part of the CCD silicon chip, and cannot be removed by mere mortals. This is entirely different than the “Baader filter”.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.