> >Another thought -- how are you driving the relay?  I assume you have a
> >bipolar or a MOSFET to drive the coil?  You might be pulling too much
> >current thru the PIC pin if you're trying to drive it dirrectly or thru a
> >bipolar transistor.  That's why I prefer MOSFETs, too lazy to compute
> >base current. 8-)
>
> I agree completely on FETs - one thing of note, though, I've seen inductive
> kickback spikes come back across the parasitic gate capacitance and nuke
> less durable gate drivers (like micro pins... :).  So I'd say be sure to
> either make sure you've got any kickback spikes clamped off well or use
> some sort of drivers - I usually use a cheap PNP-NPN totem pole arrangement
> to buffer the PIC output - also provides faster drive to the gate...
>
>
The only problem with the spike theory.... John said the
project worked fine if he reduced the time delay to
minutes rather than hours.

>"when I have shortened the delays to minutes instead of
>hour it works fine."

I suspect the heat theory myself, that or a really obscure
code problem.


_____________________________

Lance Allen
Technical Officer
Uni of Auckland
Psych Dept
New Zealand

www.psych.auckland.ac.nz

_____________________________