Jump to content

Simplified high quality All Sky Camera.


Recommended Posts

Following on from my very complicated "all singing, all dancing, All Sky Camera" with Peltier TEC and water cooling and high resolution camera, this is intended to be almost the opposite.  Not the cheapest as I want decent quality and almost 180 degrees coverage.

 

The earlier version of ASC used the ASI178MC CMOS camera with small pixels and hence high resolution.  This meant a relatively low sensitivity and with the corresponding long exposures, a high noise level.  To combat the noise I employed a high quality Peltier TEC to cool the camera and water cooling to get rid of the heat.  This resulted in a very complicated setup which I don't want to proceed with in this dreadful weather.

 

This new version (i've lost count of the version numbers) will used a new much more sensitive but lower resolution CMOS camera - the ASI385MC.  Much larger pixels and new technology compared with the 178 means much better sensitivity.  Consequently I am hoping to dispense with water cooling and even Peltier TEC and build a much simpler ASC - maybe one others may like to copy.

 

Link to earlier ASC.:-

 

Edited by Gina
Link added
Link to post
Share on other sites

Difficult to decide much until I get the new camera which should be Monday.  However, I can look at possibilities.

 

Convection cooling would be nice - avoiding fans and whether they would stand up to damp weather.  Admittedly the ASC could be run only in dry weather as it wouldn't make sense to run it in the wet!!

 

I'll look at passive coolers first then.  I have tried a few in the past but the cooling was not sufficient for the poor sensitivity of the camera I was using.  I can try these with the much more efficient camera.

  1. ARCTIC Alpine 12 Passive - CPU Cooler for Intel Socket 115x, 92mm PWM Fan, up to 47 Watts Cooling, Pre-applied MX-2 thermal compound, 95 x 96 mm - Black  Except mine is plain aluminium.
  2. Nofan CR-80EH - rather expensive option though should be more efficient.

After that we come to a huge variety of fan based coolers

Link to post
Share on other sites

Found this from when I was testing the cooling a couple of years ago :-

 

Conclusion :-  Without a fan and air duct, the Nofan beats the Artic Alpine 11 Passive cooler into the dust!! 

The Artic passive cooler works alright with active airflow through it but so do much cheaper finned coolers.

Link to post
Share on other sites

The NoFan looks nice but the price has gone up over £20 since I bought mine and there are mounting difficulties for this application.  The ASC would have to be mounted on the side of the mast rather than on top so that the NoFan has clear air.   Sealing to it may also be a problem.

 

270300872_Screenshotfrom2021-02-1222-31-53.png.25419ed7223a55807cc4d9e41986e3d5.png

 

1136817961_Screenshotfrom2021-02-1222-32-52.png.5a33bed330f326e1d07bffd6ee198ae8.png

 

 

Link to post
Share on other sites

With the NoFan seeming too difficult to use we seem to be left with fan blown or water cooling.  Certainly fan blown cooling wind over water cooling to save complexity .

 

OTOH the SX Oculus ASC seems to use neither fan nor water cooling and apparently just using the cylindrical case for ambient cooling, but I've just checked and the Oculus is mono not colour.  That males a big difference.

 

With a heat conducting aluminium tube as the mast one possibility might be to thermally connect the camera casing to the mast with a turned aluminium "plug".

Edited by Gina
Added to text.
Link to post
Share on other sites

Once fan cooling is used the source of the air needs considering as does protecting the fan from damp or even wet.  That brings up the whole question of sealing just the camera or the whole ASC.  One suggestion has been to take air from the relatively dry "warm room" of the observatory and blow it up to the ASC through a pipe with a fan.  This would have the advantage of not needing to seal the ASC as it would have a higher air pressure inside than outside.  Getting more complicated as options are added!!

 

The next step up after natural ambient cooling and fan cooling is to add a Peltier TEC and apply a small amount of extra cooling that way.  All the questions of cooling still apply - just yet another level of complexity added!!

Link to post
Share on other sites

The photos below show an earlier version of my ASC which worked well until a severe storm ripped it off and exposed the camera to rain thereby destroying it.  Later a "fatal" accident befell the lens I was using.  Camera was ASI185MC the forerunner of the ASI385MC camera I have ordered and less sensitive.  The lens was A Fujinon Fisheye Lens f1.18 and 1.4mm FL - type DF!.4HB-L1.  I bought what I thought was a replacement of the lens but it was flawed - bad star shape round the edge.  This ASC version used a Peltier TEC.

 

The lens I'm using now is 1.55mm and f2 so less FoV coverage and a bit less light catching but I think this will be more than offset by the better camera.

 

image.thumb.png.89e58b78c5c2a6356174db6d2a548820.png

 

This shows a wider view.

1715158875_MountedonROR05.thumb.png.5a20c730d9eb55a7182c08eafe5bcfe1.png


Now I'm coming to the conclusion that this could be the answer to a simpler ASC.  Not necessarily on the apex of the Roll-Off-Roof - it could be attached to a mast instead.  Fan cooling has lots of problems and water cooling has other problems.  Both are more complicated than I would like.

 

The Arctic 11 passive cooler is discontinued and replaced by the Arctic 12 with black anodising which they claim provides much better cooling.  It also seems to have a flat top apart from in the middle so should be much easier to seal.

PassiveCooling01.thumb.jpg.31c0476ad51974565d0c76d136b01497.jpg

Arctic Alpine 12 Passive Silent Intel CPU Cooler 47W LGA 1150/1151/1155/1156

 

Edited by Gina
Added bits
Link to post
Share on other sites

Been using CAD to design model of the ASC.

 

Camera, lens and dome.

1997227432_Screenshotfrom2021-02-1510-09-51.png.d62032001c61e9b3f94c1472c937c895.png

 

Making dome more visible.

786941568_Screenshotfrom2021-02-1510-11-09.png.540762a2def8605756981afa08fe0d5a.png

 

Aluminium passive cooler modelled and added to assembly.

187921335_Screenshotfrom2021-02-1511-09-25.thumb.png.3405277586c63ee947a2e872c7c26cd7.png

 

Another view at a different angle showing that this model really is 3D.

1905966452_Screenshotfrom2021-02-1510-59-17.thumb.png.2cacdfc54cfeeb2c3c6224edc4877968.png

 

All these parts were modelled from scratch using AutoDesk Inventor Professional 2008 CAD software.

Edited by Gina
Typos
Link to post
Share on other sites

One thing that is clear from this model is that there wouldn't be room for an RPi without spreading out sideways.

 

If I can get away without Peltier TEC or remote focussing, I plan to have the RPi in the obsy and work out a way of sealing the USB cable.  The dew heater is no problem.

 

I have pretty much decided to mount the ASC on a mast rather than the ROR.  To get up to the apex of the roof would need a stepladder, which has a tendency to sink into soft ground.  Personal safety definitely points to the mast option.

Link to post
Share on other sites

Tried the ASI385MC and lens on the mast using the simple 3D printed bracket shown below.

105341368_Screenshotfrom2021-02-1520-41-54.thumb.png.4cc5575382197952358458d2e5646b20.png

 

Got this FoV with small cut-off at the top which is acceptable.

1190792564_Screenshotfrom2021-02-1620-39-29.thumb.png.48443ba20d6c2879410ba55727cb3bd3.png

 

However, the pixel size of this camera and the sharpness of the lens meant the stars were smaller than a pixel Bayer cluster and would fall on single colour pixels making the stars show as different colours at different times.  These screenshots show the images zoomed in to show Cassiopeia at different times.

1801814988_Screenshotfrom2021-02-1620-15-42.thumb.png.460bdde715c53cc922f8f7588bf8aaf1.png1289154174_Screenshotfrom2021-02-1620-25-31.thumb.png.c01208c3c8af3097e8f02c2db3029424.png

 

This is not acceptable!!

 

Orion.

1395408597_Screenshotfrom2021-02-1619-24-47.thumb.png.aebfdb8f530e81182b6544b3f0eb29f9.png

Edited by Gina
typos
Link to post
Share on other sites

The upshot of this is the the ASI385MC is not suitable for this application.

 

Alternatives are the ASI178MC with a lot of cooling or possibly the mono version which will be more sensitive.  Otherwise, it;s a different colour camera entirely and the Raspberry Pi HQ Camera has been suggested.  This has smaller pixels than the ASI178MC which has pixels small enough not to spoil the star colours, so the Pi Cam should show proper star colours.  How sensitive this camera is remains to be seen and I have bought one for testing.

  • Like 1
Link to post
Share on other sites

I took the innards out of the ASI178MC and separated the boards to show the underside of the sensor board also took a photo with just the top off.  This camera has already been saved form the scrap heap from when the USB socket got broken and I soldered wires onto the board.

ASI178MC01.thumb.JPG.94300d1e9378c119b6f2e86fcd4f687a.JPG

 

Sensor and main boards opened up showing the back of the sensor through that tiny hole in the sensor board.

ASI178MC03.thumb.JPG.eba751758c74f2eb280c637197c1c805.JPG

 

Inside of the case.

ASI178MC04.thumb.JPG.557dab94cfe1bff49ef6007849e84954.JPG

 

It can be seen that there's a small rectangular hole in the sensor board showing the back of the sensor.  So if opened up air could be blown onto the back of the board for cooling.  Of course it still needs the main board to capture images so this would result in a bulky and awkward setup.  It would need a new part made to connect to the lens but I think it would be possible though far from ideal.  I would prefer the Raspberry Pi HQ Camera if I can make it work.

Link to post
Share on other sites

That aperture looks inviting. Just a copper cold finger bonded to the case would wick some local heat away and at night the radiated cooling would be effective I think. A simple sandwich with the lower board.   It takes me back to the days of toucam mods 🙂

  • Like 1
Link to post
Share on other sites

I'm concentrating on the Raspberry Pi High Quality Camera for now - much easier to use.

 

CAD model of camera with just enough detail to avoid parts interfering.  The camera tripod mounting bracket seems attached with screws so may be removable.

1260007574_Screenshotfrom2021-02-2212-52-51.png.f9be09af7ff0a223500db9e40d1b0c62.png

 

Had several thoughts about possible cooling methods. 

  1. The underside if the sensor PCB has lots of components so conduction cooling even with thermal paste is not viable.
  2. A simple chamber with camera at the top and cooler at the bottom - but convection is the wrong way - cooled air doesn't rise!! 
  3. Air could be cooled by a heatsink with base attached to that big passive cooler.  With Peltier TEC in between if needed.  Then the cooled air blown at the sensor PCB with a fan.

A radial fan like this could blow air onto the centre of the sensor PCB under where the sensor is. 

152378057_Screenshotfrom2021-02-2215-42-49.png.3ea2a2f9ba664eecb77e28b2446a87d2.png

 

1468491987_Screenshotfrom2021-02-2215-45-02.png.2d8bf5399afdae44c82619d8ae44b428.png

 

I thought "that's alright, I can have the passive cooler in it's most efficient orientation" but when I modelled this I got :-

791554173_Screenshotfrom2021-02-2219-26-08.png.46371631faffc4aa06e2eb625dc6cde8.png

 

More thinking needed !!!

Edited by Gina
Link to post
Share on other sites

OK I thought - back to having the passive cooler below everything else and use an axial fan with suitable ducting.

 

To start with, from the bottom :-

  1. Passive cooler
  2. Peltier TEC (if wanted)
  3. 40mm x 40mm x 11mm heatsink with fins upwards
  4. Axial fan 40mm x 40mm x 10mm blowing upwards, drawing air through the fins of the heatsink
  5. Not shown yet - air duct to concentrate cooling air into centre of sensor PCB

505133316_Screenshotfrom2021-02-2220-21-16.png.119f634f6d97cbf3e2819970d45aa67c.png

 

Added air duct on top of fan and an inner casing to contain the cooler air and return it to the heatsink.

1920421729_Screenshotfrom2021-02-2222-48-13.thumb.png.257b7014e2eac6d988bd05a6050f31c9.png

 

Main outer casing added and shown in clear green to show internal parts.

967890393_Screenshotfrom2021-02-2311-18-33.png.21eb056c5553695264ba9bdc8a9e3e85.png

 

Another view.

891806477_Screenshotfrom2021-02-2311-20-18.png.ad25555aa07eb0a1dea3e63bdf853a7d.png

 

Casing shown "As Material" rather than transparent.

1334917341_Screenshotfrom2021-02-2311-21-42.png.92de4166744a942f70ccbe3956b1055f.png

 

I suspect the main casing (which will be sealed to keep the weather out) will need extending to accommodate the RPi etc.  I have a pressure equalising device on order which will need a flat surface to fit into so part of the casing will need to be flat.

  • Like 1
Link to post
Share on other sites
  • 2 weeks later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...