Zim,

Viz a viz your explanation, why do you think this translated
into flaky code execution?

I originally got onto the problem because my code wasn't working
"consistently". From one pass of programming the UV chips to the
next, code that worked fine one time would suddenly run flaky
the next. After trying all kinds of things over several weeks,
as descibed in a previous msg, I discovered the MCLR leakage
problem, removed the cap to gnd, and haven't had a problem since.

- Dan Michaels
Oricom Technologies
http://www.sni.net/~oricom
==========================


At 05:12 PM 2/23/00 +1100, you wrote:
>Hi,
>
>> Don's comments on the IR drops on the MCLR pull up resistor suggest this
>> might be a possible cause--or effect from improper/ignorant handling of
>the
>> parts/boards.
>> Any body have a handle on the mechanism of this failure--ie. what is
>really
>> going on inside the pic?
>
>I think that the idea is as follows...
>
>When the Power goes off and Vcc goes down, the cap on MCLR remains charged.
>
>Whenever MCLR is more that 0.6 volts above VCC, the internal Substrate turns
>on (like any diode would) and the MCLR cap is discharged very quickly
>through VCC and the external circuitry.
>
>Depending on the size of the MCLR cap and how quickly the power goes down,
>the internals of the MCLR pin will be thermally stressed and perhaps zapped.
>
>A resistor in series will limit the peak current to a safe figure.
>
>Likewise a diode to VCC.
>
>But I like the "floating input pin" explanation better (for your "creeping
>current" observation)
>
>cheers .............. Zim.
>
>