We are trying a few things, but one item that sparked my interest is the
fact that the problem is showing up only on the flash parts.  We have some
UV parts and we don't see the problem.
We have run each pic separate, but we have always used the osc.  The osc2
out of the pics the quit working, sometimes will stop.
I am gathering some information, I should know more in the morning.

Best regards,
Giles


-----Original Message-----
Since you have more then one pic connected together and you
do not know witch one is casing the problem or you do not even
 know if it is pic related problem ( 99.9% not ) this is what you
 need to do.

1. separate all 4 pics from each other separate MCR, and use crystals
    for each one or resonator. Use 39-100 pf cap on master clear pin with
10k pull-up
   and .1 mf on power VDD pin forget about using can oscillator for now
    your reset is going to be main power ( if you want to use it ) and  try
    every thing just you ware doing it before. If you couldn't duplicate
    the problem that means what ever you ware doing with can oscillator the
    problem is there. Do not forget that the crystal is analog if you
   are working with wire type of prototyping boards then you should
   use shielded cable for cans only unless they are very close to each
other.
   Do not forget that the scope probe has at list 15pf cap in it
   when you are doing measurement  both caps should have same value
otherwise
   your sine wave will be shifted or no clock at all. I personally use 2
same kind
   of probes and I connect them to both crystal pins unless you can buy a
probe
   that do not have cap in it.

Andre Abelian






> > We are having a little trouble with our PICs.  I have included the
> > information that my co-worker sent to the C compiler
> > listserv.  Anyway, I do
> > not know if it is hardware or software.  Basically I want to know how
> > repeatable a reset via touching the  reset line to GND is.
> > Any suggestions
> > on how to handle the resets is welcome.  I am about to
> > dissect the ASM from
> > the C and verify all possibilities.
> >
> > Information submitted to the C compiler listserv:
> > We are using the CCS PCM C Compiler, Version 2.693, 7652
> >
> > I have been working with 16f876-20 rev B4 chips, and using a
> > simple MCLR
> > circuit:
> > 10k resistor & 1uF cap (and diode to Vcc for power down).
> >
> > I have 3 identical PIC's on this board they share common MCLR and
> > OSC lines (they go through 22 ohm for each PIC)
> >
> > I observe the operation of the PIC's through 5 LEDs all on one port.
> >
> > The problem is that when I short the MCLR pin to ground it
> > doesn't reset
> > as I expect.  What it does do is:
> > Running normal
> > reset
> > no LED's
> > reset
> > Running normal
> > reset
> > no LED's
> > reset
> > ???? no LED's
> > reset
> >
> > It is very repeatable for no LED's to follow Normal running, but
> > it is only 75% for Normal running to follow no LED's.
> >
> > Also since I have enabled a #INT_RTCC and a #INT_SSP function, the
> > processors will occasionally completely lock up, with no recovery
> > unless the power is cycled.
> >
> > I have #fuses HS,NOPROTECT, NOWDT, PUT, NOBROWNOUT and I have added
> > #ZERO_RAM.
> >
> > Does the compiler not identify reset during normal operation
> > correctly?
> > Is this a silicon problem?
> > Is this my problem?
> >
> > I don't want to mess with a WDT, but it looks as if I will
> > have to, unless
> > you
> > can help...
> >
>