Hi All.

I'm having an odd problem with MPLAB-ICD (16F877). I seem to be
somehow latching up PORTC, and I can't figure out why ... perhaps
someone can shed some light on it ...

Here's the scoop: if I download a HI-TECH PIC C program like:

main()
{
#if 1
   int counter;
   TRISC = 0x00;
   while (1) PORTC = (counter++) >> 8;
#else
   /* complex program goes here, also writes to all 8 bits of PORTC */
#endif
}

everything in this trivial program works as expected -- the LEDs
reflect the counter's top byte properly.

However, if I then change the #if 1 above to #if 0 and recompile the
program, something weird happens: At the end of the programming
operation, and _before_ I choose Debug->Run in MPLAB, PORTC:0 starts
glowing very faintly. There's no ac signal or waveform on PORTC:0 --
it's just a 1.75V output on PORTC:0. The complex program runs, but
PORTC:0 continues to glow faintly and and PORTC:1 remains off, even
though the complex program writes to all 8 bits of PORTC.

What's also strange is that the complex program's first chore is to
set TRISC to 0x00, yet when I watch it in a Watch window, it's been
set to 0x03. If I step through the program, I see that it's trying to
clear all of TRISC, yet only the upper 6 bits are cleared.

If I then change back to #if 0, recompile and re-program, the PORTC
problem remains. Only if I remove power from the ICD, re-apply power,
re-establish connections, reset and run, does the problem go away.

In other words, just the act of programming the part with the complex
program via the ICD latches up PORTC:0..1, and only powering it down
and re-programming it with another, much simpler program will fix it.

If I don't use the ICD but rather program the 16F877 with a PICSTART
Plus, both programs (#if 0 and #if 1) work fine. So the chip and the
ICD "demo board" appear to be just fine.

I recognize that there may be something in the more complex program
that is unintentionally changing TRISC. But what I don't understand
is what could possibly be happening that effectively locks up PORTC
such that only a POR will clear it.

Any suggestions?

Thanks!
--

  ______________________________________
   Andrew E. Kalman, Ph.D.   aek@pumpkininc.com

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.