All Sky Camera cooling

[happy-kat]

Makes me wonder/hopr if I can fix the neximage to being usable if I can cool it.

[Gina]

I don't see why not.

[Gina]

Hmm...  Can't get it any cooler on the table.  I guess 20°C below ambient isn't too bad as I haven't got any thermal insulation round the camera so it's taking all the water of or the atmosphere.

[Gina]

To go lower in temperature I shall have to make a proper rig.  A casing for the camera to provide thermal insulation and desiccant to remove the moisture plus proper pressure to hold the camera body to TEC and TEC to cooler.  There is still significant noise at 0°C so it will want the temperature taking down lower.  That could prove difficult in a warm summer should we get one again.

[Gina]

I just have to modify the design of my present camera housing a bit to accommodate the Fujinon zoom lens and 3D print it.  I can then provide a thermally insulated and sealed camera enclosure with the passive cooler.  Last night I was using an 80mm fan to blow cooling air through the fins of the passive cooler as well as running the Peltier TEC at it's most efficient.  This doesn't bode well for using the passive cooler without fan in the summer so water cooling is looking more relevant and I shall be doing some testing.

[Gina]

When trying to attach the 3D printed base plate to the passive cooler I found a problem with the threaded holes.  I thought they were M3 but they aren't.  Not M4 either but something in between.  Then it occurred to me that they might be Imperial, maybe 4BA with a size of 3.6mm.  I used to have 4BA screws but don't know if I can find any now.  I do have the original supplied screws so will probably use those and modify some of the design to suit.  Or I might leave that for now and try some water cooling experiments.

[Gina]

I have round waterblocks with pipe connections on the back which I think will be easier to use in the ASC than the square one with connections on the side.  The size is about the same as the camera.  This photo shows camera and waterblock upside down.  The tubing shown is rather small and I have ordered some larger size silicone rubber tubing (8mm bore 12mm OD x 5m).

[Gina]

Now to design some 3D printed parts for it...

[Fraunhoffer]

17 hours ago, Gina said:

15°C

10°C

2.7°C

0°C.

That's interesting - you have given my an idea to take apart an old 90's PC from the loft to extract the (large) processor fan and use it for cooling my camera.

 

[Gina]

While the new camera casing bottom part is printing I have set up the waterblock on top of my anglepoise lamp connected by silicon rubber tubing to pump and reservoir.  There doesn't seem to be any problem with the waterblock a lot higher than the reservoir and pump - ATM a couple of feet but I'm hoping it will be alright with several times that.

Edited March 4, 2019 by Gina

[Gina]

Hmmppfff...  the little water pump has gone wrong - about every 5s it draws 200mA for about a second and is not pumping!  Had to order another but this time a slightly bigger one (5W rather than 3.6W).

Edited March 4, 2019 by Gina

[Gina]

I'm getting a bit fed up with the way things are going wrong these days.  Whatever I touch seems to break!

[Stargazer33]

1 hour ago, Gina said:

Hmmppfff...  the little water pump has gone wrong - about every 5s it draws 200mA for about a second and is not pumping!  Had to order another but this time a slightly bigger one (5W rather than 3.6W).

It could be the negative header you have on the reservoir. Is there any way you can mount the reservoir at the same height as the pump?

Edited March 4, 2019 by Stargazer33

[Gina]

I've tried that but no joy.  I even tried with the waterblock below the reservoir.  Tried tapping the pump, forcing water through it in the normal direction but nothing worked.  I know these pumps can fail if there's no water in them when switched on but I haven't let that happen.  I checked the spec of the pump i was using and it can pump water up 2m.  The new one I've ordered can lift 3m.  Actually, a couple of feet will be sufficient on the observatory.

On re-reading your post, the bottom of the reservoir is at the same level as the pump with the reservoir having at least 5" depth of water.  It's the waterblock (that cools the hot side of the Peltier TEC  below the camera) that has to be higher than the reservoir.  The pump never works with negative pressure.

Edited March 4, 2019 by Gina

[Gina]

Guess what???  I've fixed the seemingly duff pump.  Having relaxed watching a bit of TV I decided to take the pump apart and see if I could see anything.  I did, a couple of little bits of what looked like plastic in the impeller.  Removed them, put the pump back together and resisting the temptation to test dry, I connected it to the reservoir and half filled the reservoir with water.  Put the pipe from the pump into the reservoir and applied power - pump worked fine ?.  Oh well I shall have a spare tomorrow!

Connected the water tubes to the waterblock and mounted it high up.  Applied power to the pump and it forced water through all the pipework and waterblock flushing out all the air.  I believe this is going to work.  From previous experience, water cooling is an extremely efficient method of getting rid of heat.

Edited March 4, 2019 by Gina

[Stargazer33]

It's always good to step away from a problem and think about something else. It's amazing how often an answer or idea just pops into your mind!

I used to scuba dive and I remember from my training that objects cool 25 times quicker in water than air, for any given temperature.

[Gina]

I'm definitely going for water cooling for the ASC - the passive cooler is inadequate and feeding dry air to a fan is more awkward than water cooling.  The waterblock is smaller than any sort of heatsink/fan arrangement.  I'm now working on the design of the ASC parts.

[Gina]

Decided I'd take some darks at various temperatures with the ASI 178MM camera.  A direct comparison is difficult as the pixels of the ASI 120MC-S are 2.4x those of the 178MM though the 120MC has the Bayer layer to reduce sensitivity.  What I think can be said is that the 178MC should be affected in sensitivity much the same as the 120MC.  I think dark noise would be the same for MC as MM given the same sensor otherwise.  So I reckon 30-40s exposures will be wanted for the 178MC for nighttime images and hence I propose taking 30s darks with the 178MM.

[Gina]

When I was using the ASI 185MC camera with Fujinon f1.8 lens I used 60s exposures mostly but I don't know how the ASI 178MC compares with the ASI 185MC for sensitivity except that the 185 has bigger pixels, in fact the same size as the 120.  I'm thinking 60s exposures may be more appropriate for 178MM darks.

[Gina]

Starting at ambient temperature of 20°C with no cooling and then 10°C, 5°C and 0°C - 60s exposures, gain 300, gamma 50.  ASI 178MM.  Water cooling system but no camera thermal lagging.

[Gina]

Temperatures didn't quite wok out as above, turned out to be 20°C, 8°C, 4°C, 0°C and -4°C which was as far down as it would go without lagging the camera casing.  I'll post the darks shortly.

[Gina]

I'm getting some anomalous results from viewing the FITS images with KStars.  I think it's normalising them - not a lot of variation between +20°C and -4°C.  I'll find another FITS viewer.

[Gina]

I need to be sure the data I'm getting is really what I think it is though the only way to be sure will be to wait for a clear night sky and cool the camera until the noise disappears or until it's as cold as it will go.  One thing is certain, I can get to a lower temperature with water cooling than even with a fan blowing air through cooler fins.

[Altocumulus]

Is this project still working Gina?

[Gina]

It has changed somewhat since then but yes the cooling is working fine.