Shahid Sheikh wrote:
> Hi all,
>
> I was looking at the programming specifications for the PIC12C5xx chip
... snip...

Shahid:
        I made my own programmer about a year ago, and it's still working
very well for me, so I likely understand the specs correctly...

-> The recommended minimum programming pulse width is 100 microseconds, as
stated in section 2.2, and is controlled by software, NOT by any logic
level applied to the chip.

-> The 100 nanosecond pulse you refer to in the timing diagram is the
minimum pulse (logic level) that will be recognized at a 12C5XX input pin
when you are transferring data or commands to it. Don't transfer data any
faster than this.

-> EACH TIME you write to a memory location with a 100 microsecond pulse,
you must issue a Load Data command, send 16 bits of data ('0'+14 bits+'0'),
send a Begin Programming command, wait enough time to give 100us
programming time (100us minus the time to transfer the begin programming
command), then send an End programming command.

-> Each 100us programming 'exercise' is followed by a Read data command to
see if the location has received enough 100us pulses yet or not. DO NOT
issue an Increment address command. Once the data reads back correctly, it
is considered to be 'marginally' programmed, and 3 times as many pulses as
it took to get this far are then applied to over-program the location. I
actually repeat the 100us pulse subroutine 3N times - I've never tried
actually issuing one 'long command'.

-> If the chip ever needs more than 25*100us pulses to get to the
'marginal' state then it is faulty and should be failed and tossed out with
the trash.

-> Once that location has been programmed, the Increment Address command is
issued to step to the next sequential location. Everything is repeated for
the next word... Make sure you have a big return stack, as you'll be doing
a lot of looping!

-> P5 and P6 are correctly given as 1 microsecond minimum in the table. The
timing diagrams incorrectly show 1 millisecond.

-> Note 2 does not contradict section 2.1.1. Section 2.1.1 says that you
must use ~13V when writing locations in the chip. Note 2 is telling you
that you can reduce the Vpp supply and still remain in the programming
mode, where normal program execution is inhibited, and the pins are
assigned their programming functions. You better get it back to 13V before
programming the next word, however.

-> Section 2.1.1 part 5 incorrectly refers to a speed verify mode.
Microchip tells me this is wrong - the 12C5XXs do not support this mode.

Hopefully this is of help to you. If you read the docs carefully, I think
you can trust them. The main problem that Microchip seems to have with
their docs seems to be with their symbols for micro and ohm, not with the
content. If only they used a Macintosh like so many of us...

        Jim