On 5/12/2016 6:04 PM, Dave Plowman (News) wrote:
In article ,
Brian Gaff wrote:
I'd just not mention it and assume it was fine. After all all the pcbs
have to be the same whether repaired or new, if the circuit was
dangerous then the device would be dangerous no matter what pcb was in
it.

ISTR Geoff, who owns a company which repairs such things, saying that over
a period of time you get to know what things fail on a particular make,
but is not allowed to improve the design by using uprated parts. It has to
be repaired like for like.



Actually the definition of "like for like" is not so straightforward. To
take a slightly silly example for convenience, you can get an unbranded
13A socket from B&Q to a given BS, or you can get an MK one meeting the
same BS. We all *know* that the MK one is likely to be better made. So
if the OEM uses unbranded, but Geoff uses MK he is actually making an
expert judgement being made that he has met the "like for like"
criterion, and has improved on the original. If it went to court, I am
sure he would not have much difficulty finding an FIEE to support his
decision, and no insurance company would in practice ever try to fight it.

In my industry (nuclear power) it is frequently necessary, when a piece
of now obsolete hardware is being replaced, to obtain a statement from a
professional engineer that the replacement is "like for like" to avoid
having to go back through all the system proving tests which might have
been carried out decades ago.

There is always an element of risk in such exercises. For example, going
back to 13 amp sockets, irrespective of the BS requirements the
flashover voltage of a B&Q socket *might* be greater than the MK one.
So, if the fault case involved overvoltage then the B&Q socket might
actually be better than the MK even though they both meet the same BS.
But by using a chartered engineer to review the case, you should
normally go past the "box ticking" stage and confirm whether the correct
selection criteria are being used.