V G wrote:
> That would not be practical. The chambers themselves are quite large,
> and the room that the chambers are stored in is itself quite humid.

As bad as the inside of the chambers though?

> It
> would not be practical to run long wires into the chambers (especially
> those for SPI, I2C, and 1-wire communications intended for short
> distances).

Not necessarily, although you may have to reduce the bit rate.  The things
you are measuring are very slow compared to even a "slow" bit rate in
microcontroller terms, so this should not be a issue.

Also, there must be a wall somewhere that separates the inside and outside
of the chamber, and I doubt that wall is so thick as to make wires from one
side to the other "long".

> Would coating the PCB and electronics with some sort of spray help to
> prevent problems caused by water? Any other ideas?

I think you'll need more than just spray, probably real potting.  Again, I
would try to minimize the circuitry in the harsh environment.  If you reall=
y
really need some electronics near the sensors (you still haven't provided
convincing evidence that you do), then make it a minimal PIC system that
reads the sensors on the board and passes the aggregated data back over a
single serial stream that can handle the run to outside the box.  You send
in 5.7V and ground, and the board returns a UART output at some slow baud
rate, like 1200.  The purpose of 5.7V is that you can put a LDO on the boar=
d
to make locally clean 5V power.  If you are really worried about noise, the
UART output can be differential, but I expect that's a complication not
needed in this case.

Now you can make multiple of these small sensor boards, each being carefull=
y
potted and sealed.  You can even spread around several per chamber, with th=
e
aggregating and control logic happening outside where the environment is a
bit more benign and where it's possible to do maintainence and debugging
while the chamber is running.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.