This is a VGA resolution firewire connected webcam made by Apple in 2006:
End shot showing it is genuine Apple
At the other end is an electronic iris (That can also be manually open and closed) and the camera specs
- 1/4″ CCD
- F2.8 Autofocus 50mm-∞
- 3 Element lens
- 54.3° viewing angle
As you expect with Apple products this is well designed and assembled. Removing a few screws the vented Aluminum tube slides out revealing a two piece plastic gage surrounded with adhesive tape
The Firewire connector is a neat swivel and twist set-up (The camera stand attaches to the Firewire plug)
There are 3 boards including a small vertical board at the rear of the camera which I detached in this image.
On the rear board are a bunch of very small ic’s it is probably related to voltage regulation.
Here is the full teardown showing the main board with a daughter board attached
At one end is the 3 vane iris diaphragm assembly (Which I am not going to teardown as I have a small project that I think I could use this with)
Obviously attached to this is the camera and lens module
The back of the CCD board has a small 14 pin IC and some passives
The daughter board is taped to the main board with double sided sticky tape, and contains a couple of electret microphones and a Cirrus Logic 53L32A a low voltage stereo A/D convertor.
The front of the main board
- Texas Instruments TMS3200DSC25 DSP
- Samsung KS64122H (2006 week 15 date code) DRAM
- Sharp F800BGHB 8Mb (512kb x 16) flash memory
- Texas Instruments TLV990-21 a 3v, 10 bit, 21Msps area CCD analog front end
- Sony D2450R a timing generator for progressive scan CCD
The rear of the board has one die with almost impossible to read markings together with a number of resistors and capacitors.
Turns out the barely legible chip is a Fujitsu MB86614A which is a serial bus interface controller for firewire.
Sony ICX098 CCD
Fortunately a full datasheet is available for the ICX098 CCD. It’s a progressive scan interline transfer CCD, type 1/4″ (4.5mm diagonal) 330k pixel (VGA format) with 5.6μm x 5.6μm pixels.
I wanted to image the die with the colour filters intact. The die is epoxied to the base of the package, so I tried to remove the package frame using a diamond saw blade on my dremel, which was somewhat successful. I was able to get at the chip but in the process I did chip one corner and spread quite a bit of debris onto the die (Which through the plastic cover I could see was pristine.)
Here are the pixels covered with RGB filters. This is the typical Bayer pattern where there are twice as many green pixels as blue or red. The bright pixels are actually termed dark pixels. They are covered with aluminum blocking light (under the colour filters hence the pixels show up brighter). My understanding is they are used to measure and then subtract the dark current (The leakage current when no light is present) from the active pixel measured currents. Ten pixels across measured 56μm confirming the 5.6μm pixel pitch (This slide actually makes a good calibration slide/micrometer for my microscope).
Here is the die photo (With the centre area not imaged, I would not be able to stitch the extremely regular array of pixels , so I have just imaged the periphery.)
The periphery has just a few transistors all with huge gates (5μm and 10μm gates lengths)
The blue filter material extends beyond the array and over the horizontal shift register.
For no reasons that I can think of a couple of the periphery transistors are covered with the red filter material.
There are several articles on the web explaining how interline transfer CCD’s work, I like this one. For each column the photo diode charges are sequentially gated out by a vertical shift register with two rows of polysilicon gates at each pixel. You can see the polysilcon connections (pink/red) to the rows here
And these are the pixels with the filters removed. What is interesting is the photo diode area relative to the pixel area (Fill Factor) is quite low, this is one of the dis-advantages of CCD’s relative to active CMOS image sensors that have replaced CCD’s these days.
There is lots of interesting silicon to look at in this camera, so I will pause here and split this into two posts