MAKING
THE CAMERAS
|
|
|
Raw CCDs are
received from the wafer fabrication department, the
Micro Technology Division (MTD). Here they are
attached and critically aligned to a metal plate,
which is the key reference point which determines
the accurate positioning of the CCD in the film
plane.
|
|
The CCD is
aligned precisely in seven axes -- left/right,
up/down, top left, top right, bottom left, bottom
right and skew -- before being tacked in position
with an adhesive which cures in 45 seconds. This
fixes the CCD sufficiently to proceed to the next
step but not enought to withstand shock in the
field.
|
|
Static-free
tape is used to create a dam, between the metal
plate and the imager ceramic, into which viscous
Stycast epoxy is flowed. The tape prevents the
epoxy from spilling out the other side. The Stycast
is cured for eight hours. This not only fixes the
CCD permanently but also prevents dust getting onto
the imager face.
|
|
(X)
|
|
Once the the
Stycast has been flowed in and cured, the space
between the CCD and the reference plate is
completely filled. The relationship of the plate
and CCD is fixed forever.
|
|
(X)
|
|
The imager
plate is now mounted onto the main camera board
|
|
through which the CCD pins protrude. These are now
soldered into position.
|
|
The mounted
imager now moves to the Defect Mapping area
|
|
where
it is first cleaned with alcohol
|
|
then
placed on an exploded camera body mounted on a test
fixture.
|
|
(X)
|
|
This rig
makes a series of images of a 1,000 foot candle
light source. These are evaluated by a computer
program on a pixel by pixel basis, checking that
all the pixels and columns are present . The output
of each pixel is plotted, any differences being
written to a file which will be stored permanently
in the camera and used to even out these small
variations on a shot-by-shot basis.
|
|
The imaging
assembly then moves on to this station for Colour
Calibration.
|
|
This is
achieved by exposing a colour chip chart
illuminated by daylight, tungsten and flourescent
colour temperature light and comparing the values
obtained by the imager again known fixed
values.
|
|
(X)
|
|
The image on
the LCD at first looks quite a bit different from
the chart. If the red is too bright, or the blue
too dark, another computer program writes an
alogithm which will correct those deviations. This
is written to the image file created in the Defect
Mapping area.
|
|
(X)
|
|
The imager is
now ready to be installed in the image plane of a
camera body, assembled in another part of the
facility.
|
|
(X)
|
|
The imager is
now ready to be installed in the image plane of a
camera body
|
|
(X)
|
|
where
it is fixed in place.
|
|
(X)
|
|
Next, the
camera's shutter is locked open and a probe is
fixed into position in the lens mount.
|
|
This measures
the position of the imager relative to the lens
mount. This is the final check that all the
tolerances in the Kodak components, ensuring that
these stack up in the right direction. To ensure
compatibility between components, no adjustment is
made to the lens mount.
|
|
Although
Kodak DCS cameras are based on established 35mm
film camera bodies, new parts have to be made,
either by the camera manufacturer or by Kodak -- as
in the case of these magnesium castings for the
Nikon F5 based DCS620 and 660 range.
|
|
In the case
of the taller DCS6XX cameras, the main board is
mounted inside the new Kodak casting.
|
|
()
|
|
The design of
the more squat Canon EOS-1n bodied DCS520 and 560
is somewhat different. These are the three boards
as manufactured, held together by a (green) frame
so that the (brown) inteconnecting cables do not
get damaged. The PCMCIA card drive is the silver
rectangle in the middle. This will lie across the
base of the camera.
|
|
The other
side of the board shows the two LCD panels (colour
and monochrome) at the top.
|
|
The boards
are broken out of the frame by hand
|
|
leaving the three boards connected by the
interconnecting cables. These are hard-wired, so
they do not have any connectors to cause
problems.
|
|
The boards
are mounted into a jig (blue) which holds them in
position for testing. Care must be taken to avoid
damage to the delicate cables.
|
|
This is
mounted onto a camera body test fixture and
plugged-in to simulate the cables in the
camera.
|
|
Finally it is
hooked up to a computer and the FireWire connection
is made.
|
|
Once checked,
the boards are wrapped around the base of the
camera on a jig. to form a 'brick'. Note the way
that the cable wraps around to connect the
boards.
|
|
Here we see
the PCMCIA card slot, the tripod bush and, at the
bottom of the picture, the cable curving onwards to
the third (concealed) board. the FireWire and power
connection are at the opposite end of the base of
the camera.
|
|
We now reach
'camera complete' where the 'brick' is installed in
the camera body behind the imager board.
|
|
The back
cover of the camera is secured in place by nine
screws.
|
|
It's now time
to make more checks. First, focus alignment. The
targets in this station are captured in various
positions in the viewfinder to check focus across
the field.
|
|
System
verication for image quality to verify exposure and
linearity of colour balance.
|
|
When
everything has been checked once again, it's time
for "button up", the final step in the assembly
process. It's here that the final screws are put in
place
|
|
the
strap is laced into place
|
|
and
the Kodak pedigree applied.
|
|
But still
there are more checks, this time as 'real' shots
|
|
pointing at these charts and colourful
packs.
|
|
()
|
|
With
everything assembled, checked, double-and
triple-checked it's time for a loving final clean
and polish by Blanche. (I wish she was constantly
available to clean my nose prints off the LCD
screens!)
|
|
before
the cameras are packed, together with instruction
booklet, software, batteries, charger, power supply
and cables.
|
|
Finally, the
cameras are ready for shipping. The yellow boxes
are for Kodak customers, the others contain the
D2000 variant of the DCS520 for delivery to
Canon.
|