The long exposure modified TouCam Pro II camera may show signs of unacceptable noise in the pictures for exposures of over 30s or so. In order to combat this effect the CCD chip may be cooled.
The cooling device I used was a Peltier module. A second hand device was obtained from eBay at a bargain price of £9.99 + £1.99 P&P. Since new these can cost over £40.00 I was quite pleased.
The supplier I used, Terry Potts has told me that he usually has a supply of Peltiers in stock. Terry also supplies Silver Compound for all thermal surface mating which I used in this project. Terry can be contacted by e-mail at:
I would also like to thank Terry for his suggestions with the design of the 'Cold Finger'.
The CCD chip has to be removed from the printed circuit board of the camera, and placed on the 'Cold Finger' assembly which is thermally coupled to the Peltier module.
Initial testing of just the Peltier module held against a PC heatsink and fan arrangement gave me a reading of +35C on the hot side and +2C on the cold. Not bad considering I simply used pressure to keep the Peltier against the heatsink and no heatsink compound (it does get better).
I decided to use an aluminium enclosure over a plastic one, as I didn't want it to become distorted by the extra weight added by the cooler components.
The heatsink was thermally insulated from the rest of the enclosure with a 1.5mm thick piece of plastic made from an old DVD case (needs must).
Here are the details of the insulator.
A gap of 1mm is allowed around the perimeter of the peltier module. The internal cut-out of 42mm x 46mm is also made in the enclosure. This allows the Peltier module to be seated directly onto the heatsink.
The next drawing shows the assembly of the Peltier module, insulator , heatsink and fan.
The self-tap screws are also insulated from the enclosure by M3 insulators. Here is a picture of the inside of the enclosure showing the four M3 insulators. The screw fixings with the nylon nuts are for the 'Cold Finger' to Peltier fixing later. The grommet is to allow the fan wiring access to the enclosure.
And here is a top view with the plastic insulator fitted.
The heatsink and fan assembly were then attached to the enclosure using four self-tap screws. The Peltier module was placed on the heatsink, with some heatsink compound between them, and temporarily held in place using an old PC slot plate from a video card.
NOTE: The voltage regulator used has been changed and is no longer attached to the case. The modification is described later.
I next made a cut-out for the digital thermometer. The few electronic components required for this module were placed on a small piece of veroboard and mounted to the thermometer. The circuit diagram is show below.
The picture below shows the veroboard circuit in place on the thermometer. I purchased the module from RS Components, part no 337-7864. Click here for the temperature module data sheet.
NOTE: The voltage regulator used has been changed and is no longer attached to the case. The modification is described later.
An LM309K +5V regulator was fixed to the other end of the enclosure to supply the voltage for the thermometer.
The wiring of the regulator, fan and Peltier module is shown below.
Here is a picture of the final assembly so far. The switch on the top is for the thermometer module backlight.
As mentioned earlier, for the first phase of testing of the module, I clamped the Peltier module to the heatsink, with some heatsink compound, using an old PC slot plate from a video card. I then connected to the +12V of a PC power pack.
With the +12V applied the fan started, (always a good sign), and the Peltier module began to get cold on one side and hot on the other. The thermometer worked correctly and the ambient temperature measurement was around +20C.
I held the thermometers external temperature sensor as best I could against the Peltier to obtain the cold reading. The best I got was around -3C.
I then placed the sensor onto the heatsink fins and it settled down at around +35C.
A hole was made in the mating half of the enclosure for the focuser adapter. The original TouCam Pro II focuser adapter was then attached to this using six self-tap screws.
I manufactured the Cold Finger from a piece of block aluminium that I purchased from eBay (seem to be using them a lot lately). The drawing below details the assembly.
The lower part of the assembly is placed on the Peltier module and the CCD is held onto it by a second piece of aluminium. This part has slotted screws to allow the position of the CCD to be adjusted to obtain dead centre of the web-cam adapter.
I was intending to have slotted holes in the case for fixing the 'Cold Finger' lower assembly, which would have allowed for adjusting of the CCD in the other direction, but it was so spot on to centre that I decided to just leave to screw holes as they were. (definitely more luck than judgement there).
M4 screws were used with a nylon washer to fix the 'Cold Finger', which I had tapped to M4 size. The hole in the box was made larger than the screw to accommodate M4 insulators. This means that the screw fixings are completely thermally isolated from the case.
Here is the completed lower part of the cold finger assembled into the camera.
The CCD will be placed on the Cold Finger with the pins protruding down the sloped sides. The TouCam Pro II PCB will be attached near to one side of the Cold Finger so the CCD ribbon cables should be fairly short. The long exposure board and components will be fitted near to the other side of the Cold Finger.
The next phase of testing was to determine the Cold Finger assembly temperature. I applied the +12V and the initial reading of the temperature was around +24C. After about five minutes the temperature had dropped to around +2C. After fifteen minutes the temperature had settled to -1.9C.
The top of the Cold Finger assembly was also made from some aluminium. See the drawing above for its dimensions.
Two M2.5 holes were drilled in the lower part of the Cold Finger and tapped to M3. Two M3x12mm screws and washers were then used to fix the two parts together. Note the cut-out for the CCD window.
Here is a picture taken looking down the adapter that shows the cut-out hole for the CCD window. I was able to line it up squarely in the centre after adjusting the top part of the Cold Finger.
After removing the TouCam Pro PCB from its original camera case, the long exposure modification wires were connected so that the camera would operate normally for test purposes. The camera was then tested to make sure all was well.
For the mounting of the PCB I used another piece of the old DVD box as an insulator, which is fixed to the case with a nylon screw and nut.
The original screws, for mounting the PCB in the camera case, were then used to fix the PCB to cable tie holders. A small pilot hole was drilled in the holders first for easier fixing.
Next the USB cable connector had to be removed. At first I tried to de-solder the connections but in the end I gently bent the connector towards the edge of the PCB and was able to expose the bottom of the connector enough to cut each connection in turn. Note the RED wire is at the bottom of the picture.
Once this was done the five holes for the connector were cleared using a solder sucker to allow easy reconnection of the cable later.
The cable-tie holders were then fixed to the insulator using M3 x 25 countersunk screws a 10mm nylon spacer, M4 and M3 washers and an M3 nut.
The USB connections were then remade, but on the opposite side of the PCB. Don't forget the order of the wires will now be reversed from the original. Note the RED wire is now at the top of the picture. A cable clamp was also used on the nearby fixing to remove any strain on the connections to the PCB.
At this point I inspected my work carefully and the camera was re-tested to make sure I had made no mistakes. Fortunately for me it still worked fine. The USB cable will need to be disconnected later, so it can be fed through a hole in the box, and re-connected once again but I just wanted to check the assembly for fitting purposes.
The two following pictures show the PCB assembly in place in the camera.
When the board is finally fitted, I will place some electrical tape along the top and bottom edges of the PCB so that there is no chance of anything shorting out on the metal case.
Next comes the delicate operation of the CCD removal from the TouCam Pro II PCB. My approach to this part of the modification was to use a craft knife under each CCD pin in turn and, once the soldering iron had melted the solder, gently raise the pin off the PCB. This was not as easy as it sounds and patience is required so as to not break the delicate pins of the CCD. After around twenty minutes the CCD was removed successfully.
Once removed the pins of the CCD were straightened and shaped to fit on the Cold Finger without touching it.
There are seven pins on each side of the CCD. In order to isolate adjacent pins I used a small length of rubber sleeve on alternate pins. These were initially placed on the ribbon cable.
The wires were then carefully connected to the CCD.
And the sleeves pulled up over the pins.
Insulating tape was then used to stop any of the pins from shorting out on the Cold Finger.
Before assembling the CCD on the Cold Finger I disassembled the entire camera so far and replaced the thermal paste I had used earlier with the Silver Compound supplied by Terry. Originally The best temperature I had was -1.9C.
Now I have a reading of -3.3C, a difference of nearly 1.5C. Not bad for just a change of compound.
The CCD now needs to be reconnected to the PCB with the ribbon cables. Care must be taken to reconnect the CCD the correct way around. I checked with an earlier picture I had taken to make sure I did not make a mistake.
I used the top of the Cold Finger assembly to hold the PCB still for ease of connecting. I had removed the PCB from its plastic insulator during this process.
Now that the CCD was fully connected again the camera was tested to make sure all was well.
Unfortunately I had rather bad banding on the image, although at least the image was there. I had read elsewhere that this may happen and could be due to the ribbon cables being too long. I didn't think I could reduce the length of the ribbon much, it was only around three and a half inches long anyway.
However, after a quick scan of the ICX098BQ data sheet I saw that the output from the CCD, on pin 7, was an analogue output. Analogue signals should always be connected using screened cable to stop noise interference. With this in mind I removed the ribbon cable connections on pins 6 and 7, GND and Vout.
I used an old audio cable, that normally goes between a CD drive and a sound card to allow playing of audio CD's on a PC. As the CD audio is an analogue signal the cable used is screened. This piece of screened cable was then connected between the PCB and the CCD using one of the inner cores for Vout (pin7) and the screen for pin 6 (GND).
In the picture below the RED wire is the signal and the BLACK wire the screen.
A second test of the camera was made and, although much reduced, the banding was still apparent.
I did find that by moving the ribbon cable, for pins 8-14, changed the amount of banding that I was seeing. Eventually I found that if I laid the ribbon cable as close to the Cold Finger as I could then the banding would disappear completely. There must be some noise being induced by the ribbon and the Cold Finger helps to shield it in some way. For now I have left it like this.
With the cable in this position I turned the gain of the camera to maximum, the shutter speed to minimum and a frame rate of 15fps. Even with these settings there was no apparent banding to be seen.
I think I will have to rethink the means of connection to the CCD and may use multi cored screened cable to stop any noise induction.
A new long exposure PCB was made from a piece of veroboard. Since I have more room than in the TouCam Pro camera case, the new PCB uses the CMOS 4066 quad analogue switch, which also allows me to remove the need for the long exposure switch of the original modified camera. I also added the components required for the amp-off modification if I decide to try it at a later time.
Here is the circuit diagram of the PCB. It is a slight modification of the SC2 mod diagram on Steve Chambers' site.
The four 10K resistors shown that connect to the 4066 on pins 1, 10, 12 and 13 are in a SIL resistor network package. Here is a picture of the new PCB mounted to another insulator made from the DVD case.
This fits into the camera on the opposite side of the Cold Finger to the camera PCB.
Two weeks later.....
What a wonderful thing a holiday can do for the mind.
Whilst away I decided that the voltage regulator that I was using was a bit over the top, it only supplied the voltage for the digital thermometer which only takes 80mA, and took too much space on the enclosure.
I replaced the regulator with a 78L05 5V 100mA regulator, in a TO-92 style case, small enough to mount on the digital thermometer veroboard. The regulator is shown below in the top left corner of the veroboard.
I also decided to use a simple phono type plug/socket arrangement for the +12V input voltage. The two pictures below show the power input socket and associated wiring.
I am also now using an eight way mini-DIN style plug/socket for the USB and parallel interface connections. This means that I only have a single cable to the PC. The picture below shows the mini-DIN plug and socket for the USB and parallel connections.
The cable that I used for the USB and parallel interface was SCART type cable. This had two individually screened cables, with a further separate single cable and screen, together with an individual four way screened cable.
The individually screened cables were used for the USB data lines, to ensure that noise levels were kept to a minimum. The screens around these two cables and the individual screen were used as the 0V for the USB. The single wire was used as the +5V for the USB.
The four way screened cable was used for the parallel interface which allows me to use up to four of the data lines and the 0V connection. Since I am only using D0, D2 and D3 for this modification, I connected the last wire to D1 which I may use later if I require.
In the picture below you can see the four data lines for the parallel port on the left, the 0V screens for both the USB and parallel ports at the bottom and the two individually screened cables and the single wire for the USB on the right.
The final constructed cable is shown below.
I next made the cut-out holes for the mini-DIN socket in the case of the camera. The original USB cable connected to the camera was cut to length and used as the camera USB internal wiring. The camera parallel port internal wiring was a piece of four wire screened cable.
The screens for the two cables were joined together then connected to a short piece of wire for final connection to the mini-DIN socket 0V connection.
The mini-DIN socket was not fixed to the case, but the wires fed through the case to make the connections easier to make.
Once all of the connections were made, then the socket was fixed to the case using two M3x6 screws, washers and nuts. Here is the socket fitted to the case.
This picture shows the solder side of the mini-DIN connector.
At this point the connections from the camera PCB and the USB plug end of the cable were tested for continuity and shorts. All connections were fine so it was time for another camera test. The USB connection was plugged into the laptop and the mini-DIN plug into the camera.
SUCCESS. The camera still worked fine, (but still with the banding problem present), which was quite a relief.
A small piece of veroboard was next constructed, to connect the fragile wires for the LX mod on the camera PCB, to a piece of stronger ribbon cable.
The red wire is the +5V feed for the LX modification board.
Here is a picture of the board showing some of the LX wiring connections to it. Note that the two connections to the SAA8116 have been removed from their pins, (93 and 97), and connected to via's on the PCB. One of the wires came off and I had trouble re-connecting it. The via's were filled with PCB etch-resist and had to be cleared before they could be used. I did this by placing the very thin wire of the mod onto the hole and heating it with the soldering iron whilst gently pushing it into the pin, (and I mean gently). Eventually the resist gave in and was pushed out by the wire, and the wire could then be soldered into place.
Here is the board mounted with the ribbon cable already soldered to it.
The ribbon cable was routed around the casing to the LX mod board and the joints soldered. The cable from the mini-DIN socket to the LX board was next terminated. The LX board was now complete and fitted into place, shown here at the top of the picture.
All of the wiring between the computer, the LX board and the webcam were now complete so another test of the camera was in order.
Everything worked fine in normal camera mode so I decided to test the LX modification. Once again all was fine, much to my relief.
All I have to do now is sort out the banding problem. One thing I did try on my last camera test was to separate all of the ribbon cable connections to the CCD making the wiring more flexible. Whilst in this state I set the brightness to mid value, the shutter speed to its fastest and the gain to maximum which seemed to show the banding at its worst.
With this setup I then took the bunch of cables on the right, in the picture above, and held them against the cold finger assembly, which is grounded. The banding disappeared completely so I am now sure that using some form of shielding for the CCD cables will resolve the problem.
It became clear after a while that the connections from the mini-DIN socket and the camera and LX boards was a bit cumbersome, so I replaced it using regular ribbon cable. Here are some pictures of the results.
Much neater I think. After refitting the cables they were tested for continuity and all were fine.
I also added some neoprene rubber, to combat any moisture build-up due to the cool temperatures that will be on board the camera, on both of the PCB's before refitting into the case. Some polystyrene was also cut to shape for the ends of the cold finger and put into place to act as thermal insulation.
Another change I decided to make, was the ribbon cable connecting the CCD to the camera PCB, with the coloured style. This would make it easier to identify the correct conductor when re-soldering to the board, since I had decided to split the ribbon into single cores to make them more flexible.
Once the cables were soldered to the CCD I used a small tie-wrap on each bunch to act as a strain relief whilst handling the other end of the cable.
The cables were then soldered back onto the camera PCB, (quite quickly this time I am getting the hang of soldering whilst looking through my eyeglass), and the CCD fitted into its holder. The base of the holder was given a smear of silver compound before being fixed into place on the cold finger.
Note the slits on the neoprene rubber to accommodate the cable. Another piece of polystyrene insulation was then added to the top of the cold finger.
The lid was screwed into place and a final test of the camera made. It worked fine as a normal camera and also in LX mode.
The camera was now ready for its first cooling test. Each of the frames below is a 120 second exposure with the gain and the brightness set to midway.
Temperature 25.2C 0 minutes of cooling
Temperature 6.7C 2 minutes of cooling
Temperature 2.2C 4 minutes of cooling
Temperature -0.5C 20 minutes of cooling
As can be seen the number of 'Hot Pixels' is reduced each time.
This is as far as I have got with the project. I will update the page as I progress further.
Check out my mark II Peltier cooled TouCam Pro II.
Copyright © 2004 - 2007 Phil Davis
Last updated 2nd December 2009