DSLR cooling, cold finger with air cooled TEC - working version

[Gina]

In view of the length of the development thread on this subject and all the false trails, I'm starting a new thread to describe my working system. I shall be adding to this thread as I collect the details.

The camera used was a canon EOS 1100D. This was chosen as having the best chance of producing good results, having the latest Digic 4 image processor with lower noise and extended light gathering range compared with earlier versions.

This development actually follows on from my experiments with water cooling for the TEC.

[Quatermass]

I have no idea what your talking about gina but it sounds very cool

Sent from my GT-S5670 using Tapatalk 2

[Gina]

Ha ha Mark Yes, it's very cool indeed

[Gina]

Prior to the cooling the camera had been "astronomised" by removing the colour balance (blue colour) filter from in front of the image sensor. This procedure is well covered in other thread by myself and others and based on Gary Honis' web page http://ghonis2.ho8.c...elmod450d1.html I removed just the blue filter, putting the clear filter back. The latter is an IR/UV cut filter which, if replaced, makes it unnecessary to add an IR/UV filter externally for normal astrophotography. But I digress...

Purpose

The purpose of this cooling project was to cool the image sensor within the camera which is more efficient than cooling the whole camera, though a lot more involved. Apart from cooling the sensor to reduce noise, attention was given to keeping the overall weight down to reduce loading on the telescope focuser. In order to make it easier to take dark subs for DSO imaging, set point temperature control was provided using an Arduino microcomputer.

[Gina]

Implementation

This project is in two parts - the control and display is separate from the camera to keep the focuser loading down. So we have two sections :-

1. The camera with cold finger, Peltier TEC, cooler etc.
   
2. The control box with display of sensor temperature and control of set point temperature etc.
   

The Camera Unit

In order to provide lower sensor temperatures without getting condensation, the camera was stripped of it's own case and put in an ABS plastic box. This was sealed to isolate the internal air from the environment. Then the inside air would be dried using a desiccant to reduce the humidity and hence the dew point, thereby reducing the temperature at which condensation would occur. The box would also enclose the cold finger and TEC plus sensors for the cold finger temperature and the humidity within the box.

A heat pipe CPU cooler was used to cool the hot side of the TEC. The TEC was arranged to be level with the box lid such that the cold side was inside the box and the hot side outside, with the TEC sealed to the box. Thus the heat was kept outside the box, allowing the inside to cool and thereby reduce the dew point. The fan on the cooler had speed control and the temperature was monitored and the fan speed controlled by the control unit.

[Gina]

The Control Unit

Connected to the camera unit with a ribbon cable for data and two core 3A power cable, in my case this provides camera cooling control and provides power control and distribution. The camera USB cable is connected directly to the imaging computer through the cable duct from pier to warm room.

As mentioned earlier everything is controlled with an Arduino. Additionally there's an LCD display of 16 characters by 2 lines to show the following data :-

1. The mode - this is the overall control and includes OFF, set point temperature, slow warm up and others (more of later)
   
2. The set temperature
   
3. Temperature of the air in the box
   
4. Humidity in the box
   
5. Dew point
   
6. Cold finger temperature
   
7. Cooler heat sink temperature
   
8. Ambient temperature.
   

Push buttons provide mode control and pots control set point temperature and display backlight brightness.

In addition to camera cooling control the box provides power distribution. Power at 13.8v enters through an XLR connector from a regulated power unit in the warm room via the cable trunking to the pier. The power then goes through a 5A fuse to a digital ammeter which shows the total combined load on the supply. From here power goes to XLR sockets for the mount and power for the TEC in the camera box. An over-voltage crowbar set to fire at 15v will remove all power if the main regulation were to fail. Small switch mode PSUs provide 7.5v for the camera power and 5v for the Arduino and a USB hub. The output voltages are checked for over-voltage as with the main supply to safeguard the electronics and the camera. (In fact cameras, as a second camera power output provides power for a second DSLR used as widefield and finder camera).

[Quatermass]

cricky Gina you must work in a labratory sounds like the work of a genius going on here doing the mod on my little canon 350 was enough for me yet alone adding a cooling system, :-)

Sent from my GT-S5670 using Tapatalk 2

[Gina]

I used to work in a research laboratory many years ago involving electronics and computing. It was my career. So you could say I'm well grounded in this field It was very interesting and I enjoyed my job - I love experimenting and trying new things - it's very rewarding, though sometimes rather frustrating

[Gina]

The Camera Unit - Parts and construction Details

Parts List (apart from the camera) :-

• Black ABS plastic box
  
• Copper sheet 22swg = 0.7mm
  
• Peltier TEC
  
• CPU cooler
  
• Thermal paste
  
• DS18B20 digital temperature sensor
  
• DHT22 digital temperature and humidity sensor
  
• Silica gel desiccant bages and/or modules
  
• Various connectors for power and data
  
• Connecting wire, strip board, solder, and miscellaneous components
  

[Gina]

Here are sources for some of the components :-

• The Peltier TEC I used - http://cgi.ebay.co.u...em=200823177205
  
• The CPU cooler I used - http://www.ebay.co.u...=item256b2a238c
  
• DS18B20 digital temperature sensor - http://www.ebay.co.u...=item3a734d574b or http://www.ebay.co.u...=item2c66e351f6
  
• DHT22 digital temperature and humidity sensor - http://www.ebay.co.u...=item43b30fa816 or for quick delivery http://www.ebay.co.u...=item4ab98a3467
  

[Gina]

Rather than run an extra wire to supply the DHT22 with 5v I used the simplest form of supply - a resistor from the 12v supply and a Zener diode to stabilise the voltage and a capacitor to remove any interference that might be on the supply.

These components were soldered onto a small piece of stripboard together with the DHT22. This assembly was placed as near to the image sensor as possible to measure the humidity and temperature nearby and from those calculate the dew point.

[Gina]

I would have liked to have removed the whole of the camera case but from previous experience, quite a lot of the top part is needed to keep the camera functioning. I did however remove some unneeded parts - and some of the top plastic part with a craft knife. The viewfinder and pentamirror are not required nor, in fact, is any of the back assembly. USB control provides all that is required for AP without needing anything on the back.

Here is a photo showing the butchered camera. I thought I had one without the cooling bits added but it seems not. I shall be describing how the cooling parts were fitted later.

[Gina]

Found a link to the box I used :- http://www.amazon.co.uk/gp/product/B000TLQOBY/ref=oh_details_o00_s00_i00

I'm searching for photos of the image sensor assembly and cold finger now.

[Gina]

I've found what I was looking for These photos show the 1100D partly stripped down. To get to this stage any of the 1100D filter removal/replacement threads can be used as mentioned earlier.

Inside of 1100D (back taken off) swhowing main board and image sensing assembly with metal shield.

This photo shows where the cold finger goes.

Main board removed.

Imaging assembly with metal screen removed. The square chip on the left has pins going through to the other side of the PCB - all the other components are surface mounted.. These pins need insulating from the cold finger. I used a piece of thin hard plastic sheet, more of which later. The 3 screws holding the sensor (and thence PCB) to the frame have been removed to release the sensor and PCB to aid inserting the cold finger.

This photo shows the pins protruding into the space between PCB and sensor.

In this photo the blue line shows where the cold finger will come up between main board and imaging board.

[Gina]

The "cold finger" is a strip of copper sheet just wide enough to cover the width of the sensor and bent into a U shape very carefully to maintain flatness on the legs of the U. These will contact the back of the sensor chip and the cold side of the TEC. In both cases good contact is essential to provide good thermal conductivity. This is aided by the usual thermal grease spread thinly over the surface. The cold finger was insulated from the pins of the image processing chip (which protrude through the imaging board) with a thin strip of hard plastic. This was actually an offcut of secondary double glazing membrane I had used to reduce the heat loss through our windows.

[Gina]

Now looking for the pics of fitting the cold finger and the insulator etc. These must be on my Linux desktop (not on Win 7 box or Photobucket).

Back soon....

[Gina]

This photo shows the sensor and neighbouring parts with the gap where the cold finger fits, the protruding pins and metal frame etc.

[Gina]

This diagram shows how the plastic insulator was shaped to fit between the cold finger and the imaging board. This is fitted into the gap first and then the cold finger slid in between it and the image sensor. The T shape stops the insulator sliding further into the gap than wanted.

[Gina]

I've now found the photos of the cold finger and it's fitting. The earlier cold finger shown in this thread was superseded.

Rather than leave off the shield over the imaging board I decided to "adjust" it to provide clearance for the cold finger. Also one of the locating pegs was carefully cut down (taking care that no metal bits got into the camera - this was made easier having the imaging assembly out of the camera).

[Gina]

Now we've caught up with the photos of the cold finger installation and can now continue...

So here's the carved up camera with cold finger. When I made the cold finger I forgot to allow for the temperature sensor so I made up a thin copper clip for it with a large tab to fit against the cold finger. A piece of rubber was slid between the cold finger and the shield and (with thermal paste applied) the sensor tab was slid in between that and the cold finger. The wires to the sensor were insulated from each other and camera parts with electrical insulation tape. The cold finger is eathed to the frame of the camera (green wire).

The next photo shows the Peltier TEC placed on top of the cold finger.

[Gina]

The next stage was to sort out the fixings for the cooler so that it sat nicely on the TEC with some pressure to maintain thermal conductivity.

In addition to the usual CPU type fixings, there were slots either side of the heat sink the could take plates to secure the cooler. Pieces of aluminium were cut to fit these slots. The first photo shows a piece of plastic glued under the lugs where the viewfinder assembly was fitted then a bolt through this and through the plate secures the top side of the cooler.

For the camera bottom fixing, a bracket was made to fit over the tripod bush and overlap the aluminium plate, with a piece of rubber for springiness and to provide thermal insulation (didn't want heat from the cooler warming the camera frame). This is shown in the second photo.

[Gina]

To provide some thermal insulation between the cooler heat sink and the camera, and also an air seal, a piece of rubber was cut with a square hole to fit the TEC, a hole for the cooler fixing bolt and and a slit for the bracket. (Photo 1.) This was fitted over the TEC and bolt (Photo 2). Next, the cooler was fitted, plus a digital thermometer in a copper clip that reads the cooler heat sink ttemperature. (Photos 3 & 4). (The rubber was from offcuts from my observatory roof.)

[Gina]

Now to the box...

Photo 1 - This is the front of the box with a nice round hole for the lens mount. This was cut with a compasses type craft hole cutter. Took quite a while to cut through the 3mm plastic but made a good job

Photo 2 - This shows the camera (before butchery) placed in the box.

[Gina]

The camera was fixed into the box using the tripod bush on the box side and two screws on the front. These used the screws that fixed the camera front on at the top. I can't find a photo ATM. The lid of the box was cut away to miss the cooling parts and a protrusion on the camera top that stuck out at the back. The rubber gasket sealed down onto the lid.

Next job was the wiring up. Power was connected using a twin screw connector (green) fitted through a side of the box while data used IDC ribbon connectors (black). Using a ribbon cable with alternate conductors earthed provided shielding between signal lines plus an impedance matched cable for the 1-wire network link. Two further connectors (white) were required to make connections from outside to inside. These were the fan connector and the 1-wire thermometer for the heat sink. The connectors were fixed through holes in the box side with hot melt glue providing both fixing and sealing.

I don't claim the wiring (or anything else) to be particularly tidy, but it works

[Gina]

Getting the humidity down in the box has caused a fair bit of head scratching and experimenting. Firstly I tried 3 or 4 10g bags of silica gel. They very very slowly lowered the humidity a bit but not enough. After much discussion in the other thread, I decided to try a capsule of silica gel beads that could be arranged to be removed from the box, dried out and then put back.

This is the item I went for :- http://www.amazon.co...ils_o04_s00_i00

And this first photo shows it placed in the box. This seemed about the best placing as the plastic box has slots front and back.

The second photo shows how I arranged it to be removable. A piece of clear acrylic was glued to the clear plastic box so that the dryness condition could be seen. These gel beads change colour from orange to green as the absorb moisture. The cartridge fits in a rectangular hole in the lid and the acrylic top stops it falling right into the box.