All Sky Camera Revisited

[Gina]

I think this dome would be better in having a cylindrical section which allows the camera casing to go up inside, as well as providing more scope for sealing.  The current dome doesn't have room for that with the lens at the centre of the hemisphere.  There's another version with a wider flange but that's out of stock.

Current dome.

Edited June 12, 2020 by Gina

[Gina]

Designed a new camera housing to suit the new taller dome.

[Gina]

Awaiting the arrival of the new taller dome.

[Gina]

The new taller dome has arrived and I've tested it for size on the test camera casing and it seems to fit nicely so I'm now printing the full production version camera casing.

Edited June 15, 2020 by Gina

[Gina]

Just spray painted the printed parts.  Seals the surfaces.

[Gina]

On checking that everything fitted together correctly, I found a couple of problems resulting in a small design change to the camera casing and the plate that surrounds the Peltier TEC and makes the bottom of the camera casing. Now reprinting these parts.

[Gina]

The new parts fitted together fine.  Joined the extra USB cable to the short piece coming out of the camera, insulated all connections with heat shrink sleeving, then tested it but... camera not detected - something wrong with the wiring it seems.  I'm wondering whether to use flat ribbon cable into the camera rather than soldering USB wires to the camera connections.  I can't see that a few inches of unscreened wiring would be a problem.

[Gina]

Since 3 out of 4 wires were still connected to the camera PCB, I decided to stay with the short USB cable to the camera.  Soldered that broken joint and connected the long USB cable to the short one.  Plugged into RPi and powered up.  The camera is now working again.  I think I'll secure the short USB wires to the camera with hot melt glue to stop it being pulled off again.

[Gina]

Having checked that the camera was working I dribbled some hot melt glue over the connections on the PCB which should secure the wires.  Following that, reassembled the camera and checked again that it worked.  I should now be able to put the camera back in its housing.

[Gina]

Camera in its housing, bottom plate attached with 8 screws into the camera casing and sealed with silicone sealant, then I connected the USB cables and checked that the camera still worked.  Next was to insert the Peltier TEC and attach the waterblock cooler with 4 screws.  This has a springy mounting system so that the pressure on the TEC was limited.  I then connected the water tubes, circulating pump and radiator and after putting water in the reservoir jar, dropped the pump into it and attached the cap. 

Initially I ran the pump and Peltier TEC off 5v from the bench PSU to check that everything was functioning correctly before increasing the voltage to my standard 13.8v and watched the camera sensor temperature drop.  Without cooling the camera was running at 32.7°C but with full cooling it went down to -3.7°C - a drop of over 36°C.  This is with the ASC indoors with an ambient temperature of over 20°C.  I am happy that I shall be able to cool the camera plenty well enough when the ASC is outdoors on its mast.

 

[Gina]

One thing that remains to be seen is whether the focus looks alright under a starry night sky.  It looks a bit out when viewing trees on the far hills in spite of being very carefully set and then locked with the screw.  The dome didn't seem to make much difference.

 

[Gina]

Testing again indoors (raining so setting up outdoors is out).  Have camera sitting on a box on the table looking out of the window.  The far hill is obscured by mist but there are nearer trees.  Here are screenshots from the KStars FITS Viewer.  Full FoV and the distant part only.  This doesn't look good.  I thought a warm camera without cooling was causing the problem but this is with cooling.  Around 0 and -6C

 

Edited June 20, 2020 by Gina

[Gina]

Moved box & ASC nearer to the window to get more outdoor view.

 

Edited June 20, 2020 by Gina

[Gina]

This is what the view looked like with the lens in focus.  Or as good as I got it.  There seems to be no doubt that something has changed and need fixing!

[DaveL59]

does the lens tilt slightly when you tighten the lock screw I wonder?

I use a M12 screw collar to clamp mine on the cctv modules, tho that too can shift focus a touch so needs a couple minor tweaks to get just so. The one CS fit I have takes care to lock as its quite easy to shift the focus slightly when tightening.

[Gina]

It doesn't seem to.  I arranged access to the lock screw by drilling through the camera casing.  Then undid the screw and refocused.  Tested adding the dome and it did shift the focus a little so I corrected for that, tightened the screw and replaced dome.  This is the result.  I think this may do.

[Gina]

Now have the cooling on to check for focus change with temperature.  Above was 33°C.

[Gina]

14°C

[Gina]

6°C

[Gina]

3°C

[DaveL59]

looks pretty stable now then

I do find that the IRcut when it switches can give slight out of focus under IR lighting, so I tend to tweak for best focus under IR and accept a slight out-of-focus in daylight, which is hardly noticeable by comparison. Different use case than what you're trying to achieve tho of course.

[Gina]

-0.6°C

[Gina]

-2°C

[Gina]

-3.5°C

[Gina]

-5°C