Jump to content

SkySurveyBanner.jpg.21855908fce40597655603b6c9af720d.jpg

Imaging Source CCD Cooling DIY - Possible Success


MarsG76

Recommended Posts

Hello Everyone,

I have been astro imaging for a few years now and in that time I collected a bit of equipment. My deep sky imaging is done using a Canon 40D full spec self modded DSLR and one thing in particular that I have in the arsenal that I literally never use is a Meade F3.3 focal reducer.

I don't use it on the DSLR because I basically get a view through a circle, like a periscope view on the frame so for that purpose its useless, but I wanted to use it because at f3.3 my SCT exposure time would decrease by a factor of, allegedly, 10??!! to get the same brightness of objects. That would be great but how do I utilize the f3.3 FR? 

So I thought... I own Imaging source cameras the DMK 21 and 41 CCDs and these CCDs would work with the f3.3 FR. They are great to use for what they are intended for, Solar System imaging, but not deep sky objects.

There is very little info on deep sky imaging using these cameras online and mostly the answers are that they are prone to noise in anything longer then 30 seconds of exposure and useless.

Of course I didn't want to believe that so I decided to do a experiment and attempt some deep sky imaging using the DMK41. The forums didn't lie, there was more noise then image!!!! even when stacking. Dark frame subtraction only left black "anti noise" effect and like I said, there was more noise than image so now there's more black gaps than image... pointless, the DSLR is much better for DSO. :icon_rolleyes:

I thought, what if I cool the CCD? And so starts the experiment.

I ordered a few Peltier TEC 40mm x 40mm TECs, and used a couple of small fans on top of a 60x60x10mm fin heatsyncs, the types used on older graphic video cards. This setup was not even close to good enough to dissipate the heat generated by the TECs. It got so hot that in a couple of minutes it ended up heating up the "cold" side and in turn the CCD chassis. Prototype 1 failed miserably... and I'm thinking that I'll need to get some beefy heat dissipation happening here.

post-43662-0-34237800-1436541104.jpg

For the next test I used a couple of heatsyncs and fans from some old Pentium 4 and AMD PCs I had in the garage and assembled the unit by placing some thermal transfer fluid between the DMK41, TEC and heatsync. As on prototype 1 I put a TEC/heatsync combo on both sides of the DMK41 body and used some cable ties to hold it in place.

The whole thing was powered using a 50A 12volt PSU.

post-43662-0-06951700-1436540637.jpg

Once powered up the I left it siting on the bench and first monitored the temperature for about 5 minutes, the heatsyncs seemed to get to a certain temperature and not increase any more.... Here I knew that the cooling was sufficient enough so it's time for a dry run.

I turned off the unit, allowed everything to reach surrounding temperature and took a 10 minute sub with the cooling turned off. Even though I took a dark frame with the cap on, I still got a frame that looked like I imaged a star field!!! ALL NOISE. 

For the cooled test I turned on the cooling and left it to cool on the bench for 30 minutes before doing a cooled 10 minute sub. The result and difference was staggering. :icon_compress:

post-43662-0-34260100-1436540645_thumb.j 

I thought that there is something in this and might have a cheap deepsky mono CCD. 

I could hardly wait for the next clear night to setup and field test the cooled DMK41 in the real world. :icon_bounce:

One thing I noticed was that there was a lot of dew developing on the chassis. The only concern I had there was that there might be dew developing on the actual chip when imaging. I wasn't concerned about dew developing on the circuit board inside since dew is distilled and would not conduct to short anything out.

On inspection after the dark frame test I didn't see any dew or fogginess on the CCD to I thought that as the chip is powered the slight amount of heat generated by it will be enough to keep the dew off it and seeing that the CCD will be orientated in such way that any excess dew anywhere else should drip off to the ground, but only a proper test will tell. Also another thing to think about was whether the fans spinning on each side would introduce any vibration. I'm left at a stage where there is only one way to find all that out... field testing.

post-43662-0-14807400-1436540657.jpg

I setup the equipment outside and did a few 10 minute subs on M104 with no focal reducer, just at the native prime F10... I didn't focus accurately enough but as this was a test as long as I was seeing something I was happy.

post-43662-0-51637300-1436540665_thumb.j

There was a lot of dew developing on the chassis, as expected and the cable I used is obviously not designed for so much current that is being drained by the TECs and fans (more TECs then fans.. much more :grin:) because the cable was getting quite warm, not hot but warm.. again it stabilized at a certain level and didn't get any warmer so I knew it was not going to melt the insulation and short out... not in the short run, not tonight so I lived with it.. still living with it. If the project becomes such a success that I'll be using the DMK cooled regularly I'll replace the cable with a heavy duty main 20 amp cable.

post-43662-0-38849100-1436540674_thumb.j 

End result was soft but once stacked it was smooth, any random pixel/noise that remained on the final stack was taken care of by a 1 pixel despeckle in PS and there was not many of them... but certainly there is something into this.. definitely a poor mans cooled CCD.

Before I packed up for the night I though I'd do a couple of quick rough and dirty 2 minute exposures through R, G & B filters at maximum gain and use it for color data.

post-43662-0-36248100-1436542310.jpg

To conclude the end result was actually better then I thought it was going to be. Whether it is a useful mod to use for luminance frames, or just RGB data etc will show up in future and proper more controlled and serious testing.

If it's usable and useful it'll be great because using the F3.3 FR with the DMK41 will give a similar field of view as the DSLR APS-C size CCD with a f6.3FR so could be used for quicker luminance frames or quick RGB subs...

Need to mention that there was no vibration caused by the fans so that "worry" was settled and a limitation to consider is the amount the final stack can be stretched during processing since the DMK ADC is only 8bit. Theoretically stacking 16 x 8 bit frames should be somewhere around the 14 bit mark that the Digic III and later DSLRs capture RAW files.

I wonder how much difference it would make to cool the DMK 21au618 and capture planet video to stack, might get more detail by lowering noise levels by cooling, able to have less noise at higher gain hence enabling higher frame rates and might make the chip more sensitive to light because it is cooler.... but that will have to be experimented on and I'll post part 2 of this experiment once I have some results.

So many questions, so many ideas but so little time.

Thanks for reading.

Mariusz

Link to comment
Share on other sites

  • 1 month later...

Do away with the cooling fins and fans, they will cause vibrations. Fit cpu water coolers and use silicon piping to transfer the water to and from the cooling blocks. A small 12 volt pond pump is fine for supplying the water pressure. Just sit it in a bucket of water.

I do all my water cooled computers like that bar the bucket and pond pump, ( bit more sophistication used).

This way better cooling and no vibrations transferred.

Derek

Link to comment
Share on other sites

I realize that my previous post about water cooling may seem over the top to some, but QSI do just that. Their CCD cameras come with the ability to use their own add on water cooling system. I purchased their system several years ago to use when I get the observatory built. Usually it is not necessary here in the UK during the dark months, as it is so cold anyway. But it ensures that there are no vibrations due to fans and also allows better cooling to another 5C degrees lower than on air cooling alone. This helps reduce current consumption at the camera and reduces the load on the Peltier element. In an observatory situation it is not hard to add such a device, as long as you add antifreeze to the water in the bucket or tank there should be no problems. The lower load on the camera electronics should also be of benefit as to lowering electronic noise in the CCD, ( current consumption changes as temperature is regulated and the fan speed changes).

 In computers I use a small header tank and air cooled finned radiators, but in an observatory there would be no need for the fans to cool the radiator if that is what is wanted.

Derek

Link to comment
Share on other sites

I realize that my previous post about water cooling may seem over the top to some, but QSI do just that. Their CCD cameras come with the ability to use their own add on water cooling system. I purchased their system several years ago to use when I get the observatory built. Usually it is not necessary here in the UK during the dark months, as it is so cold anyway. But it ensures that there are no vibrations due to fans and also allows better cooling to another 5C degrees lower than on air cooling alone. This helps reduce current consumption at the camera and reduces the load on the Peltier element. In an observatory situation it is not hard to add such a device, as long as you add antifreeze to the water in the bucket or tank there should be no problems. The lower load on the camera electronics should also be of benefit as to lowering electronic noise in the CCD, ( current consumption changes as temperature is regulated and the fan speed changes).

 In computers I use a small header tank and air cooled finned radiators, but in an observatory there would be no need for the fans to cool the radiator if that is what is wanted.

Derek

Thanks for the comments... It did get me thinking about a design to integrate water cooling but in a closed system design.

Interesting idea but I'm not going to integrate water cooled cooling for this experiment since I didn't see any vibration effects and as it is it's a nicely contained package... I think that having a water source out in the field might not be practical or comfortable. If water cooling was integrated within the CCD and around the OTA totally closed up without any open water or pipes going into a bucket, than it would be great.

For this experiment I do feel like it would be overkill and the extra cost not justified by any improvement in setup or picture quality.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.