Cooling the IMX camera

[tschoenfelder]

Dear all,

last winter/string it became obvious, that my SCT will not work without having a dew heater. But as well it became clear, that in summer the sensor will get hot. The three devices will require some power plug to use one 12V charger only. A first test showed, that the peltiers are generating a lot of temp difference, so a regulation is a must.
So I created a board for an Arduino Nano, reading the dew ring thermistor, a temp and humidity module and steer 3 12 switches.
I added an indi device to control it from EKOS. My IMX219 has no temp sensor, so snooping the camera will not help.

As I had issues with my duck tail hitting the mount I tried to make the 3d printed box as small as possible (a bit of very small ...).
For the IMX219 I imported the layout model of the Raspi V2 cam in to Fusion and created the housing. The heat sink lies on the peltier lies on the heat pad lies on the IMX219 back side.


The internet/YT shows examples of an IMX477 with a copper bad being on front and back side of the IMX477, but I fear the front to collect humidity and by this water drops killing the sensor or generating dew on the front 1,25 inch IR filter ...

Hope, this generates some thoughts ...

Best wishes, Torsten

[scriptorron]

Thank you for sharing.

Last year I tried to build a cooled IMX477 HQ camera. I bought a TE cooler (Peltier), mounted it it on the backside if the camera PCB (with a thermal pad between Peltier and PCB), made a thermal insulation between cold and hot side and added a big heat sink to the hot side (with thermal paste). When powering the Peltier, the cold side got colder than the hot side. But both sides still got warmer than the ambient. I came to the conclusion that I need an even larger heat sink or a water cooling. Both would be too heavy for my mount. Furthermore the power consumption was above my budget (I power the whole telescope with a 12V 25Ah accumulator). So I gave up.

Your solution is very interesting. I should think again about building a cooled camera. Maybe I made something wrong or my TE cooler was damaged.

[tschoenfelder]

@scriptorron : thanks for these insides. My Peltier made at full 12V one side burning hot, but the other felt very cold. It's still high in the Prio list for me ...

[tschoenfelder]

Hello Ronald @scriptorron,

to enable plate solving, I upgraded the seeker scope and added an IMX290 to it:

The IMX290 connects to a Peltier element running at 12V.

You state both sides to get hot. The peltier takes quite a lot of power. Where should it go? If the hot side is not cooled, the 'cool' side will get warmer as well. Don't buy a new Peltier; add a fan to cool the heat spreader!

Additionally, start with PWM on the Peltier's power at 1-2% until you see the sensor warm up again. My heat spreader is very small, but a vibrating fan will not be good either.
Best regards,
Torsten

[tschoenfelder]

Hello Ronald @scriptorron ,

I continued with by cooling of the IMX290.
The peltier is operated at 12V, the typical 40x40 mm one (12706?).
I shot dark images. PWM of the 12V from 0 over 1, 2, 5, 10, 20, 30, 40, 75, 90%.
I got the impression, that the peltier didn't get any power (no amp meter at hand). The heat sink didn't get warm.

I tried on the second channel another peltier, connected to a heat sink only. It got really hot, so I stopped at 90%!

Today I analyzed the FITS darks taken:

I took via EKOS 1 sec exposures at gains 9.77 (10?) right side and 21,62 (22.3?) left side.

By this I see an improvement on the std dev and average of the images. It is quiet visible. Std dev of cooled 21,62 gain doesn't match Std dev of uncooled gain 10, but the avg of cooled 22.3 is close to uncooled 10. I think this to be nice for this minimal effort (clouds, so no chance to see if dew will impact during the night with this ...)

IMX477 will be next based on these results.

Best regards,
Torsten
PS: Why do I see within the FITS gain 21.62? Is this caused by digital gain stepping in?