> > With PORTB as inputs, and the internal pullups enabled,
> > the current draw is, you guessed it, 825 uA.
> 
> What about if the wpu's are disabled ?

I didn't check; in my experience, leaving an input floating is
an invitation to ridiculously random current draws.  As
*originally* programmed (i.e., with some of the PORTB pins
set as inputs, but with the ICs and pullups that are normally
supposed to keep them from floating not installed) current
draw was ~1.5mA.  And that's with only 3 pins as inputs
and floating!

> > The datasheet tells me that at 5V, I should expect Ipu to
> > be not more than 400 uA.
> 
> I read that value (Ipu) to be *per* pin. If a pin with
> spu enabled is tied to Vss (gnd), you will have aprox 400uA
> (= Ipu) floating through the wpu, out from the pin to gnd.

Interesting interpretation.  The pullup's value would then be
~12.5k.  It's a very logical interpretation as well.

It doesn't really explain why these *unconnected* pins seem
to be adding ~800 uA to the sleeping current of this board.

I must be doing something wrong...<fiddling with code>.

I think I figured it out.  I think having the pins high was
causing the PIC to enter an interrupt handler and get
stuck rather than going to sleep.  Once I disabled GIE,
the current with WPU's on dropped to 30uA.  The added
few uA is because I've been adding stuff to the board
since I ran the initial tests.

Color me chastened.  I was interpreting the Ipu spec
along the same vein as current draws provided for
peripherals (i.e., ADC, BOD, Timer1, etc.).  I also
see now that the spec is stated with Vdd = 5V and
Vpin = Vss.  Sigh.  What an ass I feel like.

Many thanks to you all (especially Bill and Jan-Erik)
for tolerating my...ignorance...politely, and pushing
me in the right direction.

Mike H.
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist