Hi Nigel-

Don't use ANY nickel in an area of flex you intend to bend. Use ImAg or OSP.

ANY plating is vertically grained as opposed to rolled-annealed, so when
you put it over the RA you have wonderful crack initiators that propagate
right through the RA. Nickel is the worst. Yes, ImAg is still plating, but
it appears to be thin enough not to be a problem.

If you can protect any bend areas from the plating, then there is no
problem with ENIG.

Wayne Thayer

On Thu, May 7, 2020 at 5:53 AM Nigel Burtt <[log in to unmask]> wrote:

> After a bit of advice from the TechNet massive.
>
> The nickel layer on an ENIG finished flexi PCB stiffens the copper and can
> lead to cracking on a flexi, or so I am led to believe. Where a flex design
> includes both interconnect functionality and SMT circuitry, how do people
> specify the surface finish? Ideally you want the nickel thickness to err on
> the side of the lowest nickel thickness within IPC-4552 and not the upper
> end of the standard tolerance, don't you? But does that then potentially
> impair the solderability and reliability of the assembly for other reasons?
>
> Can suppliers even adequately control Nickel to say 1-3um, rather than
> 1.3-6um?
>
> I've seen papers that suggest using higher phosphorous levels or using
> columnar rather than laminar nickel deposition, but these seem to be
> proprietary solutions (pardon the pun) and not universal panacea that any
> supplier could offer.
>