You may also have problems if the device is not completely erased, a safe
bet is to erase at least twice as long as needed to bring all bits back to
their original condition , that might be the reason why it doesn't want to
program when the window is covered

Peter

Visit my Packard Bell hate site on:
http://www.geocities.com/SiliconValley/Lab/5883

-----Original Message-----
From: Eric Smith <eric@BROUHAHA.COM>
To: PICLIST@MITVMA.MIT.EDU <PICLIST@MITVMA.MIT.EDU>
Date: Monday, February 09, 1998 8:28 PM
Subject: Re: [OT] 8755A-2


>Someone was having trouble with 8755s only working when exposed to light,
>and I suggested that the problem might be that the parts were exposed to
>light while they were programmed, resulting in the devices being programmed
>for too short a time per cell by an intelligent programming algorithm.
>
>Roger Anderson <AndersonR@MAIL.MEDADMIN.UIOWA.EDU> replied:
>> Yes I thought that might be the solution, but the smart programming
>> algorithm couldn't advance past the first byte with the window covered.
>
>Not necessarily.  If the light causes the cell to read correctly after a
>shorter programming cycle, the algorithm will believe the cell is correctly
>programmed and continue to the next address.
>
>Note that EPROM cells can be either postive- or negative-logic, so
>ambient light can change their bias either toward a logic one or zero,
>depending on the specific chip design.
>
>Cheers,
>Eric
>