> We replaced the PIC in the first faulty unit and it is behaving=20
> correctly now

Do you mean you have two problems - a faulty chip and a power-
up issue ?

> The PIC has a 10K resistor from MCLR to Vdd, but no capacitor
> to Vss. If I reset it by shorting to Vss, the unit works fine again.
> It's only when the board is first powered up that the fault occurs

If you change nothing else, a cap would ensure that Vdd gets up
before /mclr, holding the chip in reset. However, PWRTE is meant
to perform that task
=20
> It's almost as though the UART is not initialising correctly. I have=20
> PWRTE enabled and the watchdog disabled

If the receiver is powering-up before the PIC then the PIC will be
seeing noise from the Data output, and that could be going into the
UART buffer. I wonder if this noise is getting to the PIC's Vdd
internally via the protection diodes before the PIC has powered
up properly from the supply rail

In projects of mine that use a receiver, the receiver's Vdd is sourced
from the PIC, and is supplied only after the PIC has initialised. Done
that way not with power-up in mind, but for power-saving during
normal operation

> If I power it down and back up again however, it fails to respond
> to commands via the UART

Powering down may not allow enough energy to dissipate to cause
a reset before powering up again, especially now that micros are such
low power devices.. If you try one of those power downs and ground
/mclr you may find it does reset properly

*
*
**********
Quality PIC programmers
http://www.embedinc.com/products/index.htm
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.