All Sky Camera cooling

[Gina]

Following on from a couple of other threads I thought I would start one on the specific topic of cooling the astro camera in an all sky camera (ASC) to reduce noise for longer exposures for use at night.  An ASC by its nature is isolated up a pole/mast or maybe attached to the highest point of a roof and out in the open.  This means normal cooling methods may not be suitable.

Cooling the astro camera is provided by a Peltier TEC but the problem is getting rid of the heat from the hot side of the TEC.  The standard cooling method is a finned heatsink with a fan to blow cooling air over it but this camera cooling system is outdoors in the often damp/wet atmosphere we have in the UK.  This then needs the question asking "Are cooling fans alright when used in a damp environment as they are designed to be used indoors?".  I don't know - does anyone?  It seems likely that damp may cause rusting of iron rotors or maybe affect the electronics inside the fan. 

There are other ways of dissipating the heat such as a large passive cooler which relies on convection for cooling or even water cooling.  Plus it might be possible to provide a supply of dry air.  All these ideas and maybe others are what I wish to explore in this thread.

[vlaiv]

I would first start by assessing actual need for active cooling?

What sort of exposures are we talking here? I think that for anything 10s or less, for ASC you don't really active cooling. ASC by design uses short focal length lens that is fast (F/1.4-F/2). Let's say that doubling temperature is 6C.

If we assume something like 0.05e/s/pix at 0C of dark current (ASI1600 has ~0.03e/s/pix at 0C), for sensor temperature of 24C we can expect 16 times larger dark current. At 24C that is 0.8e/s/pix, hence for exposures up to 10s you will have something like 8e of dark current - or less than 3e dark current noise per pixel - that is comparable to read noise of camera, and probably less than LP noise with fast lens.

Btw, we can see what sort of LP noise we can expect in very dark skies in 10s exposure. In 21.5 Mag skies, with F/2 18mm lens, with 50% QE camera having 3.75um pixels, one will get ~1.24e/s or 12.4s sky brightness. That gives us 3.5e LP shot noise from sky alone - more than dark current noise.

Is it really worth having actively cooled camera under these circumstances?

[Gina]

Experience has shown that cooling is necessary.  This applies to the ASI 120MC-S, ASI 178MM and ASI 185MC with an f1.8 lens.  I'll see if I have some images.

[Gina]

Some test images at various temperatures and exposures.  Best images are with temperature of <2°C the worst is with no cooling and around 15°C.  Exposures are from 10s last pic @ 15°C up to 60s on the lightest one.  I estimate that 15s exposures with temperature cooled to near freezing would be good with the ASI 120MC-S camera (which is more sensitive then the ASI 185MC that I was using last year having larger pixels).

60s exposure at 2°C

15s exposure at 2°C.

10s at 15°C.  (Focus was well out)

Edited March 3, 2019 by Gina

[vlaiv]

What are you trying to measure / record with ASC?

I can't really tell the difference between all of these frames - different scale / presence of moon, etc ...

[Gina]

Some nice images of the sky including the Milky Way with star colours showing.  I can also see approaching clouds long before I can see them out of the window.  (Or approaching clear sky.)  Sorry, the images aren't better documented though there is info in my ASC Blog.  Also, I haven't had the opportunity to do a more extensive report on noise v temperature.  I just know that the camera needs cooling to somewhere near freezing to get the noise down.

The camera can get down to a few degrees above ambient without cooling so is alright uncooled in winter but temperatures can get to 20°C to even 30°C in summer and the noise would swamp the image.

Edited March 3, 2019 by Gina

[vlaiv]

I have an idea how to handle cooling in adverse atmospheric conditions if you feel that you need it.

It will involve a bit of DIY but it should work without fans that would be susceptible to moisture or seizing up in cold weather.

Liquid cooling. You'll need a pump, reservoir, some tubing and cooling block. All of this can be purchased. Pumps are used for fish tanks and cooling blocks are used by overclockers to cool computer components. With enough tank capacity you won't need to dissipate temperature actively - tank will help cool liquid to ambient (liquid is just distilled water with addition of antifreeze). Just use metal tank holding couple of liters of liquid. You can mount aluminum finned radiator to side of tank to help dissipate heat from tank.

[Gina]

Yes, water cooling is one of the ideas mentioned above but is not so easy with the waterblock so high up.  I have used water cooling in the past for camera cooling and use it currently in one of my 3D printers.  I am well familiar with it.

Edited March 3, 2019 by Gina

[lenscap]

Auto-engine electric cooling fans have to operate for years in wet conditions , they run on 12V & might be scavenged from a scrapyard for next to nothing.

Quote

 

 

[Gina]

But they're huge!

[lenscap]

How about an auto heater fan. Similar environment, and they come with a heat exchanger attached?

[Gina]

Possibly.

[vlaiv]

52 minutes ago, Gina said:

Yes, water cooling is one of the ideas mentioned above but is not so easy with the waterblock so high up.  I have used water cooling in the past for camera cooling and use it currently in one of my 3D printers.  I am well familiar with it.

I obviously did not read your post careful enough - you mentioned water cooling indeed.

Most of camera coolers use very smooth running - electronics cooling fans, like ones for computers. This is because of noise and shake - as they are mounted on cameras.

There are industrial grade fans that are designed to be resilient - for example ones used in kitchen extractor fans and such - can operate under steam and higher temperatures (but I'm sure you can find ones that work in lower temperatures as well).

These usually have higher level of vibration and noise due to slack - design decision to prevent them from seizing in higher temps. Maybe you can use one of those - but it would involve quite a bit of DIY.

There is also option to go for heat pipes - you can source them from aftermarket CPU coolers - can be used for both air and water cooling based solution.

[Gina]

These photos show last year's ASC before it was destroyed in a fierce storm.  This uses a passive cooler about 4" square and 3" high.  The cooler survived the storm and is a contender for the next build.

[Gina]

7 minutes ago, vlaiv said:

Most of camera coolers use very smooth running - electronics cooling fans, like ones for computers. This is because of noise and shake - as they are mounted on cameras.

There are industrial grade fans that are designed to be resilient - for example ones used in kitchen extractor fans and such - can operate under steam and higher temperatures (but I'm sure you can find ones that work in lower temperatures as well).

These usually have higher level of vibration and noise due to slack - design decision to prevent them from seizing in higher temps. Maybe you can use one of those - but it would involve quite a bit of DIY.

Vibration may be a problem with cheap fans whilst high quality ones with low vibration probably wouldn't stand up to the environment.  I'm going off fan cooling.

The passive cooler looks like being the easiest though water cooling is interesting and would make for a smaller, more streamlined unit with less windage.  This is a very exposed location particularly from the SE.

[Gina]

The possible problem with water cooling is that the waterblock is going to be the highest part of the system and could suffer from an airlock unless the pump is powerful enough to flush the air out.  I think I shall have to do some experiments.  I do have the necessary parts.  A copper waterblock 40mm x 40mm x 10mm, 12v pump, plastic pipes, etc.

Edited March 3, 2019 by Gina

[vlaiv]

1 minute ago, Gina said:

The possible problem with water cooling is that the waterblock is going to be the highest part of the system and could suffer from an airlock unless the pump is powerful enough to flush the air out.  I think I shall have to do some experiments.

I think that can be easily solved with submerged pump - no way to get air in once you flush all air from tubing.

[Gina]

Air can be dissolved in the water but should be released in the reservoir, again if a sufficient flow to get over the reduced pressure in the waterblock.

[Gina]

Central heating systems have problems with airlocks - radiators need venting occasionally as do the highest parts of the pipework.

[Gina]

Another way of getting relatively dry air up to the ASC for cooling I though of was an air duct from the scope room with a fan at the scope room end but I think this would be quite difficult to implement.

[Gina]

ATM the passive cooler is looking best due to simplicity of use.  Of course, this would work better if the air had an unobstructed path with the cooling face vertical but that is impossible unless I had a big block or angle of copper but the extra thermal resistance might negate the improvement due to orientation.

Edited March 3, 2019 by Gina

[Gina]

I'll set up the ASI 120MC-S with Peltier TEC and passive cooler and run the TEC off a bench PSU to control the cooling and take a series of darks at various temperatures.  I'll use an exposure of 15s as this seems likely to be what I would use for nighttime images.  I'll leave the gain at 50 as that was what I did the sky tests at.

[happy-kat]

An Interesting thread.

[Gina]

Currently taking a set of darks.  15, 10, 5, 2.7, 0 degrees C so far.

[Gina]

15°C

10°C

2.7°C

0°C.