"N_Cook" wrote in message
...
On 12/02/2014 10:20, Gareth Magennis wrote:
http://imageshack.com/a/img845/570/o62e.jpg


The old PCB is the top one.

Note Con15 below the bridge rectifier, it is connected to + and - of
rectifier.
Con 16 is below Con 8 and is just wired in paralell with it.

The old PCB has an LDR, the new one a relay.


There may be more differences, I haven't bothered to check thoroughly.



Gareth.



It would be nice to nail down exactly what the problem was with the pcb
manufacture. As China made, probably affects other makes, which at lower
voltages perhaps would take longer to emerge as a problem. I posited NaCl
salt contamination to the filler material as it seems to give the right
sort of temperature / conduction graph. The problem does not require the
amps to be stored in sheds or garages, ie dampish, to induce the fault
condition.


In this case the problem may have been exacerbated by incorrect value and
possibly temperature co-efficient type of resistors fitted, if Phil's link
is to be believed, so possibly a combination of problems, not a single one.

It would be unwise, IMHO, to assume ALL PCBs in the faulty batches ended up
suffering from this problem. Any contamination/faults may have been quite
clumpy in distribution, and/or down to a single individual during their
shift.

Who knows.




Cheers,


Gareth.