Just a reminder the deadline for comments on the latest draft for comment is fast approaching - this Friday the 11th in fact. (Though it is not clear if that is midday or midnight, so probably best to log in and out again before the end of Thursday..)
If you wish to comment and have not yet done so you need to create an account on the BSI website
Click on login/register on the RHS of the BSI welcome page and give them an email you can read without needing to change machine, as you need to click a verify link they send you.
Once logged in search for "BS 7671:2018" oddly searching for BS 7671 without the space between BS and 7671:2018 does not find it...
There are 4 links as the thing is a bit of a monster. PDFs of each part can be downloaded for perusal, but to finally comment you need the on-screen viewer (the Read Draft and comment button)
A test button that checks the software is still operating is not much use Andy, Look at the alleged tests on Youtube, search AFDD. All tests, no fires, but no ARCs either, just a few tiny sparks.
If the test button gets the thing to trip it's probably checked the things that are most likely to fail (software crashed or mechanics jammed). Rather like the RCD test button - it may not spot 100% of possible problems (e.g. slightly delayed action) but in probably 95% of real world problems if the RCD has failed then the test button doesn't cause a trip.
We should also be consistent with other protective devices - we don't disregard the benefits of a fuse or MCB because we can't test them...
I would argue that a fuse or mcb can be tested insomuch that it can be tested for continuity, so 'Go/No Go', plus they should be checked for the correct raitng in any case.
I would imagine that at present at least, any test button on a combined function all-in-one unit will be to test the RCD element only.