DSLR Active cooling MOD process - Part 2

[MarsG76]

20 hours ago, benzomobile said:

What camera do you want to chill?

A Canon 40D, the one pictured. 

[Peeetr]

Hi, I´m preparing my 40D for cooling and I measured there is no space between sensor and metal cover on the mainboard. Please how did you solve it, are you spaced board? How thick you used the copper plate? How thick is your peltier? If you space it, how you connect bottom flat connectors? There is no additional length. Hmm and if board is spaced, how to cover it all? Thanks Petr

[MarsG76]

On 17/03/2021 at 10:29, Peeetr said:

Hi, I´m preparing my 40D for cooling and I measured there is no space between sensor and metal cover on the mainboard. Please how did you solve it, are you spaced board? How thick you used the copper plate? How thick is your peltier? If you space it, how you connect bottom flat connectors? There is no additional length. Hmm and if board is spaced, how to cover it all? Thanks Petr

Hi Peeetr,

Before you mod your camera remember that my mod draws 4 Amps of current for the two peltiers and effectively renders your 40D useless for everything other than astro photography.

The other thing I want to warn you about is that I destroyed 4 PCBs during my adventures in modding the cam, but mostly because I didn't protect the electronic from condensation.

 

I used a copper sheet that is only 1mm thick and it fit in the space between the sensor and the main PCB, it was a tight fit, but it fit. 

Make sure that you insulate your PCB from condensation, fill out the gaps with some foam sheets and insulate as much of the camera entry point as much as possible... I used expanding foam... I'll attach some pictures from my final version of my camera that worked well now for around a year....  

Good luck.... but when you get it working, the difference in subs quality is worth it with almost no noise and much cleaner stacks.

 

[MarsG76]

On 17/03/2021 at 10:29, Peeetr said:

Hi, I´m preparing my 40D for cooling and I measured there is no space between sensor and metal cover on the mainboard. Please how did you solve it, are you spaced board? How thick you used the copper plate? How thick is your peltier? If you space it, how you connect bottom flat connectors? There is no additional length. Hmm and if board is spaced, how to cover it all? Thanks Petr

Some photos of the camera with the copper plate, PCB dew insulation using a hot glue gun, than the final version of the mod...

By the way, don't forget to remove the IR cut filter from m the sensor to increase your Halpha sensitivity...

[Peeetr]

On 18/03/2021 at 10:54, MarsG76 said:

Some photos of the camera with the copper plate, PCB dew insulation using a hot glue gun, than the final version of the mod...

By the way, don't forget to remove the IR cut filter from m the sensor to increase your Halpha sensitivity...

Hi, I thought that the peltier is directly on the CMOS sensor where isn´t enough space. and copper plate conduct out heat. Now it´s clear. Peltiers are outside.  How temperature did you reach on the CMOS?

[MarsG76]

4 hours ago, Peeetr said:

Hi, I thought that the peltier is directly on the CMOS sensor where isn´t enough space. and copper plate conduct out heat. Now it´s clear. Peltiers are outside.  How temperature did you reach on the CMOS?

It would be much neater (and WAY more efficient) to directly place the peltier up against the back of the sensor... but unfortunately this is not an easy option. Doesn't seem impossible but definitely would introduce new problems...

This solution has allowed the camera to work for me for over a year now, and the only thig I had to keep an eye on, is the dew point temperature, and not cool the sensor more than (a MAX of) 4-5°C below that... I found that when I wrapped dew heaters around the front of the camera/filter wheel/OAG join (and had a fan lightly blowing at the camera side), that it didn't allow any dew to settle on the sensor. Now I'm not sure whether the fan actually stopped the dew from forming at a lower temperature by moving the wet air around, but I did have it on.  

Once I did cool down to almost freezing (hovering between 1-2°C)  without the heater straps or the fan, during a night where the dew point was at 10°C and I did get condensation on the sensor bloating the stars.... at this point I turned off the cooling, continued "exposing" and within 30 minutes the sensor heat cleared the sensor. This is when I raised the temperature to be between 7 & 8°C and successfully completed the reminder of the imaging night. 

To answer your question the temperature drop is generally approximately 13°C below ambience (but can be slightly more).

During testing, the two peltiers cooled the copper plate to -18°C on the test bench, but this is here the sensor was sandwitched between the two peltiers.

The copper plate outside the camera reached a temperature of -8.6°C in the field where the ambient temperature was 18.2°C, this is the temperature of the plate right next to the coolers that were running at max. The sensor temperature reached -0.7°C when not exposing and raised to 5.4°C when exposing non stop for 30 minutes... this 13C below ambience was sustained throughout.

The time when I got condensation on the sensor, the ambient temperature was 16°C so that night I was 15°C below ambience... so obviously the less heat that the system has to fight, the greater the potential drop will be from ambient (to a point)...

 

To put into perspective, the uncooled DSLR sensor heats up to around 33C during the long exposures... so I think that it's a considerable improvement.

 

My subs, including the narrowband subs, are virtually noise free... comparing to what it was, it's like night and day... before I had white noise point throughout the whole frame that had to be noise reduced, no doubt losing signal and detail, now I get a white noise point here and there... and frequently the 60-300 second RGB subs have no white spots whatsoever. 

Another benefit to this cooling is that by setting the cooling a few degrees above max cool down level, I'm able to generate and apply dark frames at the imaging temperature, and I did notice an improvement in the stack after I started applying darks... particularly some lighter vertical "slight glow" like pattern which as in all of my stacks in the past.

I'm more than happy with the level of cooling and the results I'm getting with the 40D after cooling it. To the point where I was putting off getting a dedicated Astro cam for a long time.

 

Let me know if you're going to mod cool your 40D as I'd like to see how you go, and see your results.