Gina's DIY All Sky Cam - Complete Redesign

[Gina]

I have some rather more extreme ideas for the ASC

One thing the SBIG camera has is heating of the lens mounting plate to stop condensation on the optics and this also heats the air to stop dew on the dome.  I could do this if I were to replace the top of the camera casing with an ali plate but I was thinking of eventually having the camera enclosure sealed with a bag of silica gel to absorb any remaining moisture in the air inside.  That would prevent any condensation on either the lens or the sensor (since my sensor is exposed).  At present the dew heater keeps everything warm and above dew point but also heats the image sensor, increasing the noise and hot pixels.  Not the best idea

One point about using ABS plastic for the camera enclosure is the thermal expansion.  I could reduce this to a third by using aluminium.  Also, I could cool an aluminium housing and hence the sensor to reduce dark noise.  With dry air in the enclosure I wouldn't get condensation on the bottom lens and the top one would be in warm air from the dew heater.  The heat from the hot side of a Peltier TEC could be conducted upwards to provide dew heating.

The above are currently just thoughts and I might do better to just buy a better camera when I have the funds    My time might be "free" but it isn't limitless - nor is my energy.

[Gina]

Having got fed up with setting the ASC up on the observatory for testing then taking it down again for mods - repeatedly, I decided to arrange testing out of the living room window with the capture and control laptops indoors.  So I designed a right angle bracket so that I could poke the pole out of the top window and clamp it to the window frame.  The length of the pole helping to get the camera further out from the house.  Then I had another thought that if I made a 45 degree bracket I could get the camera up above the roof and get a wider view

Here are my two designs in SketchUp.  I love the "Follow Me" function   These things look like plumbing fittings   Currently printing the 45 degree version.  This will take just over an hour and use 23g of filament costing about 40p.

[Gina]

I tried the 45 degree bracket and the pole poking out of the window and the bracket used to attach pole to obsy corner post hooked on the guttering but 45 degrees was too much - looked like 30 degrees was nearer, so I now printing a 30 degree bracket

[Gina]

I now have the ASC set up on the house (bungalow actually) with the pole hooked onto the guttering and coming just inside the window.  There is something like a metre or a bit more above the guttering and clear of the roof.  I got the bracket pretty much right as the camera is horizontal judging from the image or maybe slightly lower on the south side.  I currently have no dew heater due to not finding any suitable resistors but I have some on order   I do have the dome on and everything else is much the same except that the dome is mounted lower.

For James I have done a little test - I have exposure on 10s (current max) and set the gain at 50 which produces a grid on the preview image as can be seen below.  BUT saving a FITS file and processing that, doesn't show the grid.  See second pic below.  I seem to have knocked the zoom lever in fiddling as the view is smaller than the frame.  I have cropped the saved image accordingly.

[JamesF]

That's very odd.  I'm intrigued.  Are you reducing the image scale to fit the entire image on the preview screen?  The only thing I can think of at the moment is that it might be some artefact of the image scaling.

James

[Gina]

That's very odd.  I'm intrigued.  Are you reducing the image scale to fit the entire image on the preview screen?  The only thing I can think of at the moment is that it might be some artefact of the image scaling.

James

Almost, zoom adjusted to fit the image height to the window then the horizontal scrollbar adjusted to centre the view circle.  ie. the frame is wider than the window.  I'll try changing the zoom.

Later... yes, it varies with the zoom.

[Gina]

An example with the gain upped to 70 and FITS file processed with curves and levels.  Contrast was too high at this gain setting so I reduced it with curves.

[Gina]

Last oaCapture image for tonight - gain set to 43 exposure 10s.  Capture saved as FITS file, converted to TIFF with FITS-Liberator, opened in Photoshop and level adjusted a bit.  Saved in PNG file format.

The stars of The Plough are nice and sharp as are one or two others but most are very fuzzy   I wonder it this is the dome.  I'll try tomorrow night without the dome (going to bed now).

BTW the orientation is probably about 15 degrees out as I have the camera lined up with the house which faces slightly east of south.  I didn't think of correcting for this in the pole to box bracket.

[Gina]

I ran SharpCap overnight with exposures of 60s and gain of 50.  The dark noise was very much less than before which I put down to no dew heater.  The dome didn't dew up as it happens   This demonstrates that keeping the QHY5 cool definitely helps.  Running at near ambient temperature (plus small amount of heating from the QHY5) the focus didn't seem to vary much at all.  The stars in The Plough stayed sharp all night.  Other stars stayed blurred all night which I'm pretty sure must be down to the dome. 

I could see the forecast bank of mist come in in the video - the stars in an area of the sky gradually disappeared as the mist encroached into the view.  Cloud around the horizon also showed clearly.  The light of dawn started showing around 5am and the sun rose about an hour later.

One problem - the shutter didn't close as the light increased this morning and I had to tap the pole to get it to close gradually.  This could have been due to damp I guess as the camera enclosure is not completely sealed ATM  - damp could get in through various gaps including the USB connection.  I wouldn't have thought there would have been enough to stick the solar film to the sensor though.  I need to investigate.

[Gina]

Camera opened up.  The solar film was well clear of the sensor, in fact I think it was probably catching the lens.  The gap between lens and sensor is only about 5mm.  I'm beginning to think the solar film is not stiff enough.

I can confirm that the sun does not seem to hurt an unprotected sensor though I haven't tested it in the noon-day sun when the atmosphere lets the most sunlight through.

[Gina]

I've stiffened up the solar film with fragments of ABS extruded filament (0.4mm diameter).  But every time I do a mod to the shutter seems to affect the meter movement a bit and I don't think it can stand much more - it's very delicate   I'm thinking now that this arrangement is more trouble than it's worth and that maybe I should just go for a micro servo and plastic framed shutter.  I can make that the right shape and height straight off and I can adjust parameters in the Arduino sketch to put the shutter in the right positions.  I remember someone suggesting a micro servo as simpler overall and I think they were probably right

[Gina]

I've got the micro servo working - sort of   Added the code bits to my control sketch and connected servo to Arduino.  I can set the open and closed positions and sending "0" or "1" through the serial port moves the shutter to the designated positions.  BUT on uploading the sketch the servo moves all over the place as the sketch starts, in three steps, outside of the shutter movement range   If I were to fix the shutter blade to the servo shaft as usual then the shutter would slam into the sides of the casing, probably breaking the gears in the servo   Anyone else had this problem? 

Rotation required is 25 degrees with the range of movement within the casing of about 50 degrees.  The sensor is in the middle and the shutter can go either way to open.  I guess I could arrange a friction drive between servo and shutter blade and provide stops for open and closed positions but I don't like friction drives - they tend to wear loose.  Another possibility might be a spring.

I dunno... problems... problems... problems...

[JamesF]

Is this because the output pin values on the Arduino are undefined until the sketch initialises the hardware, do you think?

James

[Cath]

I dunno... problems... problems... problems...

Indeed

Such is life Gina, don't you think ?

[Gina]

Yes, to both of you   Probably...

[JamesF]

Would it perhaps be possible to use some of the other pins to create an "enable" signal in that case?  To create a control system that does something like "only pay attention to the movement controller pin if some other pin has the opposite value", or "only pay attention to the movement controller pin if two other pins have values opposite to each other"?

James

[Gina]

Yes, I wondered that myself or something similar.  Maybe a relay with make contacts between control pin and servo.  The drive to the relay could come from two other digital output pins.  But would it work if the outputs are undefined?  One could be high and the other low.

Maybe a delay circuit, started by the Arduino and closing the relay after a delay.  

This is ridiculous

[Gina]

I wondered if this problem only arose when uploading the sketch so I unplugged the USB and plugged it back in.  WORSE!!  The shutter firstly went anticlockwise about 30 degrees then clockwise in no less than six steps of similar size until it reached the servo end stop.  Then it moved back to my preset closed position as set up in the sketch

[JamesF]

Does this help? http://forum.arduino.cc/index.php?topic=47878.0

James

[JamesF]

Reading a little more it appears that all the data pins are configured as inputs when the Arduino starts a new sketch and the values float until they're defined as outputs by a sketch, so it looks like some method is required to tie them to a specific value until your sketch gets control.

James

[Gina]

It might   I'm using pin D9 for the servo as shown in the Arduino web site.  I'll try a 1K5 resistor (which I have to hand) to gnd.

Later... NO - no difference - one step back, 6 or 7 steps forward - happens every time I plug it in   Step size of around 10 degrees or so.

[Gina]

The behaviour is different between uploading and re-plugging.  On upload it goes two steps of about 30 degrees forward then a pause of a second or two then another step clockwise to the end stop then back to my preset position.

I'm very puzzled by this - the use of micro servos is well documented in the Arduino web site and I've found the same instructions in other places.  No mention of crazy behaviour on connecting or upload

[Gina]

Looks like this is another "blind alley"     I've gone back to the meter movement - I've got that working again   Threads of ABS filament to stiffen the solar film and keep it clear of sensor and lens and it's moving free and clear

I may well play around with micro servos with the Arduino later as this could be of use in other projects.

[Davey-T]

Are you going to have it working in time for the Lyrids Gina ?

Dave

[Gina]

I doubt it   I was just connecting up for another test when the cat pulled the whole lot off the table   Now I have to strip it down again and see what's broken because it isn't working

May have an early night...