It's a two-step process whereby I've firstly removed the CCD array from the camera's circuit board and will attach it to a heat-sunk and fanned peltier device. I've chosen a 30watt unit which should draw around 4 amps when running. Eek! Big battery required. Second phase will be to carry out the long exposure mod. I plan to re-case the whole shebang in a new plastic waterproof case.
First I've removed the CCD array. Applying quick heat to the pins and a sharp knife I was able to gently pry up each leg and straighten it.
Holding it gingerly in a clamp edge-on, I attached 7-core ribbon to each side so that it can then connect back to the board remotely - allowing me to locate the CCD where the back of it can be cooled - on the peltier device.
Pic 1: Lifting pins
Pic 2: The CCD removed and pins straightened.
Pic 3: Extension ribbon attached to CCD
Pic 4: CCD on peltier next to heatsink and fan assembly.
Tip
When de -soldering small SMD device it is easier when you melt additional solder across the pins (make solder bridge) and then you can lift SMD device straight off without bending pins or damaging PCB pads. Don’t have soldering iron too hot, about 350 degrees is sufficient. Clean up with solder wick.
Thanks Karl - exactly what I found after the first two pins. Only thing then is to make triply sure that you evacuate the extra solder so that no two land patterns on the board end up bridged.
Having done the LX mod to my Toucam, I have some appreciation of the fun and games you are having. Would like to try the cooling mod, but the mechanical aspects deter me, along with the probability of condensation on the cooled surfaces of the chip in our humid NQ climate. Are you planning to change the CCD chip at some stage for a more sensitive version?
I thought I'd start on the case tonight as well. I bought a plastic jiffy case that is 115mm long by 65mm wide by 40mm thick. It has a grometted seal and the two halves are held together by 4 recessed stainless screws.
I drilled a hole in the base - just off centre (to allow for the placement of the CCD) and screwed in my Mogg adapter, which made its own thread. Great fit and very sturdy. Secondly I carved out a square hole to accomodate the fan and heatsink assembly. It protrudes through the case for better cooling. I plan to drill a few neat holes top and bottom to allow for condensation to escape if it can. This is why the switch gear and electronics will remain set up high in the case.
I placed a SPDT switch above the fan to allow me to switch between normal (AVI) and long exposure. I've yet to fit the 25-pin male parallel plug and I'm also going to fit a female USB port so that I can do away with the ToUcam's permanently attached cable.
Tomorrow I'll do the L/E mod wiring - for which I bought a small roll of the old "wire-wrap" wire - which is brilliant for this task. It's very thin, single core and insulated by a very thin and flexible sheath. Can't wait!
Having done the LX mod to my Toucam, I have some appreciation of the fun and games you are having. Would like to try the cooling mod, but the mechanical aspects deter me, along with the probability of condensation on the cooled surfaces of the chip in our humid NQ climate. Are you planning to change the CCD chip at some stage for a more sensitive version?
Regards,
David.
Not at this stage David. If I get some good results with this setup I'll consider it at that point I guess. Do you know of a suitable candidate?
Where things will go apart from the 1/4" aluminium billet which I will cut tomorrow. This aluminium piece will physically house the CCD in a wide groove to allow the wiring to exit either side. It will be bonded to the CCD with thermal paste. The rest will stack over the top of these in the following order:
Its a good idea but you need some finesse here. Extending the wires from the CCD means there will be more electrical noise induced in them, whether it be more voltage drop or induced capacitance hum from surrounds or your body etc.
Maybe you should use twisted pair wires or coaxial wires. I'm in favour of twisted pair for clean signals but in this case because of the signal type I'd go coaxial cable with the shield well earthed.
Maybe even ensuring that it does earth thru a earth stake to dump all EMI.
Also at this stage have you thought about the CCD power supply. I would suspect the one on the cct is cheap and not regulated much (probably its fed by a power pack). Could be a good time to build in a good low ripple supply for the CCD and cctry ....maybe even battery for the CCT and cctry.
Good work, keep it up. I'm dying to see the results.
edit:
forgot to mention, beware moisture build up from the peltier onto the CCD.
Last edited by wasyoungonce; 07-06-2007 at 08:07 AM.
Its a good idea but you need some finesse here. Extending the wires from the CCD means there will be more electrical noise induced in them, whether it be more voltage drop or induced capacitance hum from surrounds or your body etc.
Maybe you should use twisted pair wires or coaxial wires. I'm in favour of twisted pair for clean signals but in this case because of the signal type I'd go coaxial cable with the shield well earthed.
Maybe even ensuring that it does earth thru a earth stake to dump all EMI.
Also at this stage have you thought about the CCD power supply. I would suspect the one on the cct is cheap and not regulated much (probably its fed by a power pack). Could be a good time to build in a good low ripple supply for the CCD and cctry ....maybe even battery for the CCT and cctry.
Good work, keep it up. I'm dying to see the results.
edit:
forgot to mention, beware moisture build up from the peltier onto the CCD.
Thanks for the advice. I have similar concerns, and without having done this before I guess that time will tell.
In reference to the ribbon rather than shielded, I'm firstly going to try what has already worked for other people. I, by nature, tend to disbelieve what I read until I either prove or disprove it myself. In this case I know of others that work well as long as the length is kept to a minimum and they are not routed past anything else on the way to the CCD. Most of their length will be stored in the trench I'm cutting into the cooling billet - so should recieve some modicum of signal privacy.
Power to the CCD is supplied by the regulated 5v line via the USB port, and that is elevated to 15v on-board. If it proves inadequate I'll add a p/s.
Quote:
Originally Posted by http://www.madpc.net/~greg/icx414_amp_off.htm
The power supply works as follows:
The CCD circuitry requires regulated +15V and -5.5V. These need to be generated from the +5V dc USB supply. They are produced using a single switching-mode power supply controlled by the SAA8112 (Vesta) or SAA8116 (ToUcam). The output from the switching transistor feeds a boost circuit to give a raw +18V unregulated supply and a buck circuit to produce the unregulated negative supply. The raw boost and buck supplies feed individual fixed voltage regulators that produce the steady +15V and -5.5V supplies needed by the CCD front end.
*hint*
