Peter P. wrote:

>>Even easier is to use a bunch of spring fingers to read the
>>conduct/no conduction to a piece of PCB. No optical biasing issues.
> 
> 
> That only works with gold contacts and they have to be debounced properly. Paper
> tape does not have error correction ...
>  
> 
>>Someone could probably write code to use a web cam to
>>track the holes as they moved past a backlight.
> 
> 
> Contrast is much better if they are pulled in front of a flat black platen, with
> front light. The camera does not take kindly to direct light sources pointing
> into it. Also the software to handle the output already exists, it is the same
> as used with the scanner. Please see the bottom of my previous posting.

OK, I looked. I didn't see any URLs for this 'scanner' based software.
Am I missing a post?

A lot of the paper tape in my junk pile is BLACK, so black on black
gives lousy contrast. (PDP 11 & PDP 8 diagnostics tapes).
It also would be difficult to correctly light aluminum (metal) tape 
(used for bootloaders and other high repeated use stuff).

Backlighting works well if you set the camera gain manually.
With correct lighting intensity you get nearly 100% modulation
of the video signal, making it easy to discriminate.

And you wouldn't need a PC to do it.
Just pull the tape horizontally, and have the PIC A/D sample a few
microseconds into the scan line for whatever scan lines have
the holes. Use the sprocket hole as your clock for 'data valid'.

Or pull the tape vertically and dissect a single scan line into
the bits. Could even be done with hardware shift register and
comparator + PLL. Just need to have a stable optical setup.

How about a 'Rube Goldberg' style virtual design contest to come
up with the most unusual way to read a paper tape?

R




-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist