----- Original Message -----
From: "Bob Turner" <[EMAIL PROTECTED]>

> Where did you pick this one from? I would think that it is for the future,
> and not the near future. As the technology expands we will surely reach a
24
> x 36 CMOS sensor, but not for a couple of years yet.

AFAIK, the problem of making large CMOS sensors isn't in technology, but in
economy. CMOS sensors, like all silicon chips, are manufactured in units
called wafers, and the problem is that there will always be a percentage of
the chips on each wafer that will become useless becuse of dust and other
contaminations.

An (hyptethical) example:
You have two production lines, which both use wafers of the same size. The
first line makes small sensors, let's say 100 of them on each wafer, and the
second line makes larger chips on wafers of 50 each. The manufacturing cost
of100 small chips is the same as 50 large, since the wafers in both lines
are the same size. So, if it would end there the large chips would cost
twice as much. This is where the problem of contamination comes into play.
Lets imagine that each wafer will get an average of 20 grains of dust,
evenly spread across the surface, and the result will be that 80% of the
small chips will be allright, but only 60% of the large ones.

The math part:
Let's say that each wafer costs $10000, then the price of the small sensors
will be 10000/90=~$111, and the large sensors will be 10000/30=~$333.

Conclusion: Large sensors could be sold if people were willing to pay for
them, but since the manufacturers want to sell as many units as possible,
they compromise the sensor size to make the product cheaper and therefore an
option for more customers. I guess that most people think that a few grand
less will make up for the inconviniences of a smaller sensor.

Well, sorry about the long post, at least it killed some time for me.. :o)

/Andy


*
****
*******
***********************************************************
*  For list instructions, including unsubscribe, see:
*    http://www.a1.nl/phomepag/markerink/eos_list.htm
***********************************************************

Reply via email to