Re: CCD

On Mon, 20 Sep 1999, you wrote:

> From: "Arne Chr. Jorgensen" <instel@...>
>
> About the CCD, - yes, I have had the same thoughts. But most of
> the CCD's I know of, has a serial output. I tried something with a
> 1024 pixel area, but before I got any results, - I fried the
> piece. I had planned to get a smaller one, 128 bits - but the
> problem is still the serial interface.

> Process the bits in parallel. Maybe it could be broken down to a
> just some TTL gates, a 16 to 4 decoder, -- you know in what
> direction I am thinking.
>
> There is also some small detectors with 4 integrated sensors, used
> for position lenses in laser heads, etc.
>

Hi Arnie & Jon

I've played around with some ideas on encoders - I was trying to resolve
angular position, but the idea will apply equally to linear sensors. If you
take two films, both with lines at the same pitch and at a slight angle, one
film fixed and the other free to move with the slide. If the lines are arranged
at equal and opposite angles, you will get ' moire fringes ' set up, as the
films move relative to each other so do the moire fringes. All that is needed,
is to count the fringes as they pass a fixed sensor. The resultant number of
pulse are greater than the number of lines on each film, and by using a simple
64 bit CCD ( TI no. TSL214 ) it should be possible to resolve the position to a
few microns. But as the sensor ihas a serial output, the speed would be rather
slow.

Regards, Paul.