Brian Stephenson wrote:
>
> found the problem with flash memory on the Matsonic board.
>
> I was using code to identify the chip only, not program it.
> The datasheet for am29F010 states chip ID code as:
>
> write to address BIOS + 555 with AA
> write to address BIOS + 2AA with 55
> write to address BIOS + 555 with 90
> read address BIOS for manufacturers ID
> read address BIOS + 1 for model number
>
> The addresses are wrong, should be 5555, 2AAA, 5555 which
> is the jedec specification, but I couln't get ID with the jedec code
> either because the writes to the BIOS chip were too quick, so
> I had to place a delay of 1 usecs between each write.
>
> I dare say programming the chip would run into the problem as
> well as it uses same type of address programming, the following
> table gives a brief rundown:
>
> sst39SF020 90nsec needs 1usec delay (out of asus SiS530 board)
> ae49F2008 120nsec needs 1usec delay (out of matsonic 630 board)
> am29F010B 90nsec no delay needed (border line)
>
> I would hazard a guess and say 70nsec is ok, 90nsec needs a delay,
> there you go, how to waste a perfectly good afternoon.
>
No, it is not a waste. It is a nobel expericence for you.
Ollie
