All Sky Camera Revisited

[Gina]

The first test got down to -2°C now on the second with more fins but lower ones and it seems to be about done at 0.7°C.   Looks like the new cooler wins contrary to expectations.  Maybe the bigger air space helps cooling.  I'll go with the new taller cooler then and design the ASC casing and air ducts to suit.

Edited April 23, 2019 by Gina

[Gina]

The 60mm fan sends a lot of air through the cooling fins and I'm not sure whether I want all that air directed up to the dome, one side could go straight out to the environment with just a hood to keep rain out.

Edited April 23, 2019 by Gina

[Gina]

Camera with cooler.  Green cone is focus gear (teeth not shown), red is camera casing, sealed and provides thermal insulation

[Gina]

Dome designed and added.  Then there's the focus motor mounting and the main casing.

[Gina]

Preliminary design of the casing with room allowed for the focus motor.  I'll probably add the motor mounting to the camera casing.

.

[Gina]

I may have a new name for my ASC -  "An OriGinal All Sky Camera"  ?

[Gina]

Base added.

[Gina]

Now I need to work out how to vent the warm air to the atmosphere whilst stopping rain coming in when the fan isn't running.  I'm thinking a cone to direct the air into the dome via the camera and out round the edge of the dome and an outer cone (maybe the cone shaped top of the casing with a smaller cone inside. 

With just the 30x30mm TEC and air cooling albeit blown by a fan, I think the cooling may not be enough.  I could use a double stacked Peltier TEC cooler using the TEC1-12706.  This would take more power than water cooling but provides warm air for dew prevention.  If I use water cooling I would need to add a heater which seems daft.  Plus the air cooled system might not be much less efficient once dew heating is taken into account.

Edited April 24, 2019 by Gina

[Gina]

This shows the idea.

 

Edited April 24, 2019 by Gina

[Gina]

Running a test with stacked TECs but it's not good!  First of all the temperature gradually dropped to around 2°C then rose again to 5°C.  With the single 30x30mm TEC it got down to -2°C.  Running at 11v and just over 5A.  This is very disappointing.  Mind you, it is using a lot of power (over 55W) whereas with water cooling the TEC runs on 20W and then there would be some power wanted for a dew heater but I doubt it would want 35W!

[Gina]

Shut down and powered off and removed the TEC1-12706 leaving just the smaller Peltier TEC running at 12v.  Camera temperature has got down to -7.8°C and still slowly going down.  Conclusion - the TEC1-12706 is not suitable for this application (though it worked for the dehumidifier).

Nowhere near the -20°C I got with water cooling but not bad.  Restricting the airflow a bit by directing some up into the dome may reduce the cooling a bit but we'll see.

Edited April 24, 2019 by Gina

[Gina]

-8.5°C and it's been there for a while now. 

Just added a barrier to restrict the airflow from one side of the cooler to see how much it affects the cooling.  Went down a bit more but now on the rise and back to -8.5°C.

A bit later -- Seems steady at -7.6°C.

Edited April 24, 2019 by Gina

[Gina]

3D printing the casing.  About 3 hours left.

Rather than directing all the warm air up and out of the dome, I shall have holes in the base plate and let some escape out the bottom.  The idea is to maintain a reasonable airflow over the heatsink fins for best cooling.

Edited April 24, 2019 by Gina

[Gina]

Print failed.  I'll see what tomorrow brings...

[Dr_Ju_ju]

Another thought on a way to cut down on the internal condensation, would be to fill the internal chamber with an inert gas that doesn't have any water vapour e.g. Argon (https://www.ebay.co.uk/itm/Wine-Food-Preserver-Gas-Pure-Inert-Argon-in-an-Aerosol-Can-by-Preservintage/323711813681?epid=27029742041&hash=item4b5eba4831:g:wNgAAOSwqaNcdWlZ)

As long as the enclosure can be sealed, including sealings around cable entries etc. that should remove one of your issues, ensuring internally the dome remains clear....

[Gina]

It was failed sealing that caused the problem with desiccant to dry ordinary air.  Maybe I should re-examine my sealing method - I relied on silicone bathroom sealant.  For the few days during which the casing stayed sealed and the air dry, the ASC worked perfectly with the water cooling.  Except it needed a bit more than the 2W dew heater I was using - that is not a problem. 

[Gina]

I think proper O rings or 3D printed TPU sealing rings would be a much better bet than silicone sealant.  I need to change hat from botcher to engineer!!

Edited April 24, 2019 by Gina

[JamesF]

I don't understand why silicone bathroom sealant wouldn't work.  Perhaps not enough to get a really good seal?

James

[Gina]

Hmm... that's what I thought and it was oozing out everywhere.  I wondered if it was actually waterproof.

[JamesF]

Well, it seems to work around my bath and shower :)

James

[Gina]

Just tested the top of the case that was sealed and found a leak.  That part was not as well sealed as the other parts.

[Dr_Ju_ju]

Instead of bathroom sealant, use a liquid gasket (https://tinyurl.com/y5wokj3a) They are meant for much harsher environments than we use....

[Gina]

I'll try it - thanks.  Ordered a tube.  Also ordered some ceramic 5W resistors for the dew heater.

I think the seal I made for the cables worked alright but I may put the RPi inside the casing and then I shall only need power, which is easy to seal.  I was worried about high summer temperatures (ha ha) with the unit in full sunshine (if we get any) but I think an external shield spaced away from the main body should take care of some solar radiation.  It would be open top and bottom so convection can take heat away.  Of course, the dome would be in direct sunlight.

Edited April 25, 2019 by Gina

[Gina]

Wearing my Heath-Robinson hat, the question of internal temperature could be overcome with a Peltier TEC and inside/outside heatsinks and fan(s).  Cooling the internal air when too hot and warming it when too cool.  ?

[Dr_Ju_ju]

As you say, how about some tin foil (heavy duty) wrapped around the outside ....