In response to:

Slightly OT.
        I remember from a long time ago, seeing an way of getting low
power
from thin air. You need a circuit resonant at the frequency of your
local AM
transmitter, an antenna and earth and a rectifier tapped into the coil.
Instant free energy. With the low power requirements of a PIC it just
might
be possible!(When I tried it I found I needed a longish wire antenna -
but I
also needed several milliamps)

Anyone else tried this?

Richard

*******************************

I'm hip.  Only I think it was a really long time ago.  It was called the
Arc of the Covenant.  Gold antenna connected to a large capacitor.  The
movie didn't do it justice.  The blue arcs were OK but Harrison Ford in
the snake pit, ick.  Or maybe it was a crystal radio driving headphones.

Anyway.  Dallas Semiconductor communicates between devices such as
temperature sensors, RTC's,  4kEEEproms with 2 wires.  Ground and data.
Each chip on the 1-wire bus (not counting ground) get gets it's power by
half wave rectifying the data and storing energy when the data goes
high.
Check their weather station thing.

Serial and parallel infrared communication devices that don't use an
external power supply work the same way.  As do other "self powered"
RS-232 peripherals like serial mice.

If one PIC is powered (and the bus master) and it sends a constant
carrier signal (like a 50khz switching power supply square wave) but
periodically goes high impedance so the other PIC's can request service
but not so long that the capacitor runs down on the slave PIC, it
works.  Asynchronous bi-directional communication on 2 wires one of
which is ground with one or more slave devices operating on parasitic
power.  The Dallas parts have Time Slots and ID's.

The Dallas DS1820 is a sophisticated temperature sensor that looks like
a small plastic transistor.  The three leads are GND, VDD, and I/O.  In
parasitic power mode, VDD is connected to a capacitor which is charged
through diodes by the 1-wire bus signal.

It's not easy.  But on the other hand, if the 2 bus wires are connected
to a battery to power all the PIC's directly,  you need to source or
sink current to change the voltage measured between the 2 wires for
communication.  That means running down or charging the battery a
little. Or something else has to give like internal battery resistance
or a series resistor at the battery.  You wouldn't want to have much
swing in the data signal.  Which brings on noise problems
(distinguishing signal from noise).  You would also need to sense small
voltage difference maybe with external comparator chips.

Let me know if it works. I might need to do a 2-wire soon.

Thanks,
Don