At 09:32 05/05/97 EDT, myke predko wrote:
>>Gerhard Fiedler wrote:
>>> I'm thinking about a thing like this and putting some calibration data
>>> (ADC) into OTPROM, like measure it first and then program the chip with a
>>> patched file. Or better, burn the program, do a calibration run (part of
>>> the program), get some reference values (from the PIC), send these back to
>>> the programmer, and then put these reference values in the program memory
>>> at some unused location. Does anybody have experience with a setup like
>>> this? I.e. is it possible to program just some locations after the "main"
>>> program has already been burned?
>>
>Tjaart wrote:
>>I haven't done it, but I think it would be possible. Program the memory
>>earmarked for the calibration data with retl FF 's (unused). You could
>>also
>>pop the data into external EEPROM if you could do something with the
>>extra
>>memory.
>
>Just to clarify, you should note that the calibration value would later be
>entered into this location.  When I first read this I was confused because
>the Calibration location of the 12C50x is the reset vector.

I wanted to put my _own_ data in OTP EPROM locations (no reset vectors, no
chip configurations or whatever) -- but data I only know after having
loaded the PIC with some program to get me this data. That's why I'd have
to do this in a second programming run. Tjaart confirmed that it might work
as intended.

But probably I won't check it really out, because some further
investigation led me to believe that I need more than 8bit ADC resolution,
so I'll use a 16f84 (with a LTC1288 ADC) instead of the intended 16c711 --
and then I have an on-chip EEPROM.

(But still I think this is something useful and that I'll use it sooner or
later; often you have to calibrate an ADC or DAC, but you just do it
_once_, so that the results can very good be stored in OTP memory. And if
you have a little spare room for the routine in your program space, that's
the ideal place to put it...)

Thanks for the thoughts :-)