> Thanks everyone for your help, it apears adding a base-emitter resistor
did
> the trick. Thanks again everyone for your help.

Then there is something else going on you don't understand yet.  A B-E
resistor should not have mattered unless the PIC power supply is slightly
lower then the emitter.  The PIC contains CMOS totem pole drivers (you
aren't using one of the open drain outputs, are you?).  These drive very
close to the power supply under small load, certainly closer to the power
supply than it takes to turn on the base of a bipolar transistor.  The load
should be essentially zero since there should be no base current in this
case.

I would investigate why the pullup resistor was necessary instead of just
throwing it in there and forgetting about it.  Is the transistor damaged and
leaking?  Is the PIC at a slightly lower supply voltage?  You mentioned that
the fan "came on" as expected, implying that it starts out off.  This means
there is some sort hysterisis going on here, which shouldn't be the case.
Try a different transistor and see if that changes thing.

I also agree with others that your 4.7K base resistor is too high.  You are
asking for a gain of over 100 from the transistor just to get to the 90mA
operating current.  Remember that the startup current is higher.  Your
transistor may not be fully on, which would cause it to heat up, which could
damage it so that it doesn't turn off properly.  I would use a 1K base
resistor and loose the B-E resistor - it doesn't belong there.  That should
give you about 4.3mA of base current.  The PIC can handle this fine, and
only requires a current gain of 23 from the transistor - a much more
comfortable number for a power transistor.


********************************************************************
Olin Lathrop, embedded systems consultant in Littleton Massachusetts
(978) 742-9014, olin@embedinc.com, http://www.embedinc.com

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