Adding cooling and filter wheel to a debayered 450D mono DSLR

[Gina]

Having had problems with the 1100D that I was going to use for this project, I am now looking at doing the same with the 450D.

I have a satisfactorily debayered 450D sensor which I have proved will work at low temperatures and which when installed in a working 450D body, works fine with APT as a long exposure imaging camera.  This was not the case with the 1100D.

The 450D camera innards is the same size as the 1100D and I have also checked that the same parts can be removed from the 450D as the 1100D.  Also that the optical path can be reduced in exactly the same way by cutting down the plastic frame.  I'm therefore using the same hardware as for the 1100D project.

Since this is a new thread for the 450D I will cover some of the same detail as the other (1100D) thread.

[Gina]

This diagram below shows a sideways cross-section of the camera with cold finger (in blue) from sensor to Peltier TEC, the dry chamber (in red), cooled by a PC CPU cooler and the filter wheel.  (This is for the 1100D but the 450D is very similar.)

The top (scope side) of the dry chamber is a 2mm thick aluminium plate which connects the camera frame to the filter wheel and to the scope mounting.  The light path is hermetically sealed by a optical window consisting of an IR/UV cut filter.  All the camera parts are contained within the dry chamber and protected from the dampest of dewy nights.  The cold part is limited to the sensor and imaging board, cold finger and TEC by thermal insulation.  This improves efficiency while at the same time allowing most of the camera to operate at designed temperature.

It is hoped that sufficient heat will be transmitted from the hot side of the TEC round the copper and aluminium dry chamber box to the optical window to stop it misting up.  If not then some extra heating will be applied.

The small black parts with red T inside are digital thermometers (TO99 case like transistors) to measure the cold finger and heat sink temperatures.  The cold finger temperature will be stabilised to the set-point by varying the power applied to the Peltier TEC using Arduino control.  Tests have shown that the temperature given by the EXIF T value in the RAW data is a long way off the actual sensor temperature.  (It is measured in a chip on the imaging board.)

[Gina]

These photos show the casing for the dry chamber.  The components are just placed in their positions - screws into tapped holes will hold them in place.

1. Thin copper sheet bent up to form a box
2. Aluminium plate that will form the top of the box with pieces of angle for attaching the copper box
3. The copper box having been soldered up and put in place

   

[NickK]

How are you keeping the amp/ADC cool?

[Gina]

How are you keeping the amp/ADC cool?

They're on the image processing board, I believe, which is within the cold zone.

[Gina]

I have yet to work out if the 450D will fit in the dry chamber I made for the 1100D.  The length and width are alright but the height depends on how much I cut off the plastic frame and differences in the bits that stick out.  The 450D has a connector on the power board that sticks up - the 1100D doesn't have this.

[Gina]

I've taken a photo of both 450D and 1100D (without sensor etc.) and added layers in Ps to highlight the dry chamber size.  Two identical overlay layers just moved so that each fits over the appropriate camera.  This seems to show that the 450D should fit in the same space

[Gina]

Here's the updated cross-section diagram for the 450D.

[Gina]

I have now removed virtually all the parts not required for astro imaging, added a ready made (for the 1100D) cold finger and Peltier TEC.  I have also checked whether the 450D will fit in the dry chamber made for the 1100D - there is a slight problem as mentioned above.  I have yet to cut the plastic frame at an appropriate place - when I find out what that place is

Sensor with cold finger                                              

Placed on a metal box (for display purposes) with Peltier TEC on cold finger - top view

Bottom view

Placed in copper part of dry chamber and showing the connector protruding

Another view of above

[Gina]

Drawing of filter wheel as seen from scope side.

[SnakeyJ]

Hi Gina - you do all the good/exciting projects and seem to have a near endless supply of cameras to mod!   Looking very nice on this and, fingers crossed, look forward to seeing this imaging with from the new widefield mount.

Can I ask where you are sourcing the TO99 Thermistors and how these connect to the arduino?  I'm after something suitable for my QHY5L-II mod, which is about to be restarted now the lathe is operational

[Gina]

Hi Gina - you do all the good/exciting projects and seem to have a near endless supply of cameras to mod!   Looking very nice on this and, fingers crossed, look forward to seeing this imaging with from the new widefield mount.

Can I ask where you are sourcing the TO99 Thermistors and how these connect to the arduino?  I'm after something suitable for my QHY5L-II mod, which is about to be restarted now the lathe is operational

Thank you Jake

I get the cameras from ebay auctions - yes, that is an endless supply Must limit myself though - it costs money!

The temperature sensors are DS18B20 digital temperature sensor - "1-wire" devices.  They use just one wire and ground and there's virtually no limit to the number you can connect.  Every single device of whatever function ever made has a unique ID code - billions of them (unlike USB) so there is no chance of confusion.   I made a mistake with the case type number - it's TO92 - but the same as the little plastic cased transistors.

There are 2 ways of powering "1-wire" devices

1. Run them off 5v
2. Use what's called "parasitic power" whereby power is obtained from the data line.

More info here :- http://www.maximintegrated.com/datasheet/index.mvp/id/2815

There are various sources - most electronic component distributors seem to have them and, of course, many on ebay.

For using the Arduino there is a "1-wire" library with all that's required.

Later on in the thread I shall be dealing with the sensors and using the Arduino for control.  I shall include all the hardware design and the Arduino sketch.  These will be similar to the design of my 1100D camera set-point cooling covered by earlier threads - but rather simpler.  eg. I do not plan to include a local display and menu control - all controls will be on the computer.  This reduces the hardware and transfers the work to software.

[SnakeyJ]

Gina,

Many thanks for that extremely useful info, I shall order some up tonight.    Really like the idea of being individually addressed, which means I can run two or three (cold side, warm side and ambient) on the same digital input!    I have all the other parts & materials so hope to start turning/milling my new camera body for the QHY very soon (though may just do a run at Jupiter as a last unmodded run, just in case I blow it during the build!)

I think with so much useful information being produced on camera mods at Gina's Labs - you should receive formal sponsorship, though probably unlikely that Cannon will actively support this research.   Perhaps we can get this crowd sourced by the astro community

Thanks again - Jake

[Gina]

What a lovely idea   Thank you Jake

[Gina]

An update to the cross-section drawing with better proportions   It also shows the "bodge" to make room for that connector that sticks out from the power board.

[Gina]

Been sorting out the thermal insulation between imaging assembly metal frame and main metal frame.   Been using an offcut of the rubber roofing which is about a millimetre thick.  It seems to fit in reasonably well.  Now I need to find some thicker insulation for the other gaps around the cold parts.  Maybe some of the rubber matting tile offcuts from the obsy floor or OTOH maybe something like polyether foam would be better - just have to see what I've got.  Plastic foam may be more amenable.

[Gina]

Found just the thing for the other thermal insulation - the bits of polyether foam I took out of my aluminium case to take my lenses and other astro bits and pieces, cut to suitable shapes/thickness.

[SnakeyJ]

Should do the job Gina, and I suspect this would be less prone to break up/turn to dust than the floor tile - very undesirable inside your camera enclosure.   The other approach to mould a really close fit might be to bag/mask the camera body and fill with a gap filling foam - once set you can remove the camera and cut away the excess bag and perhaps seal any cut /rough surfaces.

[Gina]

A couple more photos. 

1. The 18B20 digital thermometer chip in position against the cold finger before being soldered onto a twisted pair cable and covering in thermal paste.
2. Polyether foam around the Peltier TEC and cold components to provide thermal insulation.

 

[Gina]

Should do the job Gina, and I suspect this would be less prone to break up/turn to dust than the floor tile - very undesirable inside your camera enclosure.   The other approach to mould a really close fit might be to bag/mask the camera body and fill with a gap filling foam - once set you can remove the camera and cut away the excess bag and perhaps seal any cut /rough surfaces.

Yes, I think so too.  I did wonder about expanding foam but concerned that it would get where I didn't want it and gum up the works (particularly the shutter).

[Gina]

I think I've got myself a little confused about which way round the camera goes in the copper box   I'll sort it out - never fear

[andyboy1970]

How are you dealing with corrosion between the copper and aluminium boxes?

[Gina]

How are you dealing with corrosion between the copper and aluminium boxes?

It will be painted.

[Gina]

Here are some darks.  Stretched x40 in Ps and the 10% cut off the bottom (if that makes sense ).  ISO 1600 and exposures of 1m, 2m, 3m, 4m, 5m, 10m.

       

[Gina]

I'm working on the cooling part ATM and hope to have some cooling available for tonight so that I can do some cooled darks.  I've worked out which way the camera goes in the copper box   I'm having the filter wheel projecting below the camera as seems to be a common arrangement.

Between times of working on the camera unit and posting here, I'm thinking about how I can arrange guiding.  If I mount this camera on the main mount I can probably use whatever guiding I'm using with that but with a second mount (and SLR lens) I shall need a second guiding system.  One option would be to have imaging camera and guide camera side by side - QHY5 attached to another SLR lens.  I've also been looking into OAG but with the back focus of SLR M42 thread lenses being just 45mm I'm seriously lacking room in the optical path.  No problem on a scope.