Anybody notice the changes to 643.7.1 and 643.8 when it comes to RCD testing. In the note it says that "Regardless of the RCD type, effectiveness is deemed to have been verified where an RCD disconnects with in the time stated below with an alternating current test rated at residual operating current I△n"
So the minimum requirement is to set your instrument to RCD type AC and carry out the test even if you have a Type A or Type B RCD.
I presume the authors of the GN have not actually done enough testing to encounter that failure mode..
Uncalled for
That is a shame. Most meters handle a non-tripping RCD quite gracefully and just time out after a second or so and issue a warning.
As I said, there's no right answer here.
You failed to respond to "Some devices with RCDs say "Test before use" and the point that you wouldn't introduce a fault current if there were a way to prevent it. Coupled with the fact that the standard for the test instrument is based on standards that BS 7671 is based on (and the RCD product standard) that people are happy to say "only protect some people some of the time" ...
Agree Mike, can't see a real world problem here, only lots of theoreticals,
A bit like the need to test RCDs in the first place? And in saying that, I'm not saying that RCDs don't fail (some mcb's do, and I've seen that also) ... but we're cherry-picking risks I think?
The current is only allowed to exist for a very short period of time, as designed into the meter operation. That is a safety feature.
That's based on a touch-voltage or touch-current for time that's not guaranteed to protect all adults (never mind children).
Isn't it just a simulated, mechanical test rather than an electrical one and not a true trip, in the sense of a truly, operational trip when in service.?
So's pressing the test button.
We're not performing the "type test" of the RCD (which involves so much more, including pre-conditioning and other provisions for the laboratory conditions in which a sample of the RCD for qualification against the standard is subjected ... these are not the conditions in which a real RCD on-site is subjected to, so why would we expect the same performance?
“Tests as required by 643.1 followed by the operation of the test button” is hidden away on page 99 of GN3 (blue band) and was enough for a standard 2391 question as to when the test button should be pressed. The only acceptable answer as far as the examiner was concerned, was after the tests.
Is my comment really uncalled for ? I'm not trying to hurt anyone, and I apologies if you understood it that way. however, I think a serious point has been missed
However, believe me if you are squatting in an under-stairs cupboard when the thing goes bang almost at head level, you will not treat pushing the self test button so lightly in future. If the authors of the GN had been there too, I'm pretty sure that would have been considered.
Personally I think the makers should use fusible resistors for the test loops in their RCDs, but they clearly do not. RCDs that have not been regularly exercised can and do fail on *1 current, and then at least the older more mechanical ones sometimes loosen up and meet spec again after a few hits with a more substantial fault - now would you consider such a 'freed up' device fit to remain in service until next EICR. (if it was jamming brakes on the car would you be so worried ? )
To say 'Test before use' is an entirely different thing to ' Do an un-timed gross over-current test before a properly timed, current limited one' The former is advice to use the test button before plugging in a new load, and suits RCD sockets inline extension leads and so forth.
I agree the test button does not use the CPC, but we are looking at triple fault before we get to danger - the CPC must be very open circuit, not just a few hundred ohms, the RCD must fail to open, and the tester must exceed the danger current by some margin, and probably fail to time out properly either. As most test gear is on some sort of test rosta, this is not so likely.
The fact that domestic RCDs are almost never exercised in reality except when there is a fault is not a reason to not get the best out of testing when the opportunity arises.
We do not test MCBs I think partly as the history of fuses - clearly once you have tested a fuse you need to change it (!), and unlike the RCD they are not a single line of defence, as they are backed by the company fuse in one direction, and for portables by the fuse in the plug in the other - the risk of a single MCB failure is to some extent limited to fire and cable damage rather than an electrocution condition persisting. Indeed from time to time that ultimate company fuse backstop is what saves us, as MCBs do occasionally misbehave too, but the fallout from such a failure is a more limited.
Oddly I have an RCD from Wylex on my lab bench as I type this because we are building one into a piece of equipment that has to travel, and we are putting in a number of RCD failure indications.
FWiW as it is easy to measure it here I have just checked it and the self test resistor seems to be a touch under 5K ohms so about 50mA at 230V. The Robin RCD tester set to 30mA times one mode seems to draw 28mA and in half mode about 12mA. Regardless of long or short button pressing the tester cuts off the test current after 1.3 seconds.
Mike
“Tests as required by 643.1 followed by the operation of the test button” is hidden away on page 99 of GN3 (blue band) and was enough for a standard 2391 question as to when the test button should be pressed. The only acceptable answer as far as the examiner was concerned, was after the tests.
The examiner didn't take into account BS 7671 in their logic, sadly ... the "Tests of 643.1" are "643.2 to 643.11" and therefore include the functional tests of 643.10.
As discussed above there is no indication of the order of tests 643.7 to 643.11, so in fact all the examiner could conclude is that the test button should be pressed twice ... once during the tests of 643.1, and then again afterwards.
To the best of my knowledge, there was never an indication of the order of pushing the test button vs testing the RCD in the 8th Edition.
Is my comment really uncalled for ? I'm not trying to hurt anyone, and I apologies if you understood it that way. however, I think a serious point has been missed
Serious debate goes into this, and as I've said a few times, there are no right answers to all this. However, to assume the decision is based on lack of experience is what's uncalled for.
To apply the 1x test current for 2 s can be a serious risk ... if precautions aren't taken so can a protective device exploding. TBH if going "BANG" is mainly due to test device resistor failure, that can happen at any time the RCD test button is pressed, so are you reducing the likelihood of that happening by pushing the button after using the RCD tester? I'm not convinced.
There are no right answers on this one, as I said earlier.
Personally I think the makers should use fusible resistors for the test loops in their RCDs, but they clearly do not. RCDs that have not been regularly exercised can and do fail on *1 current, and then at least the older more mechanical ones sometimes loosen up and meet spec again after a few hits with a more substantial fault - now would you consider such a 'freed up' device fit to remain in service until next EICR. (if it was jamming brakes on the car would you be so worried ? )
Agreed ... it could also be said that when carrying out checks on live equipment, particularly at or close to the service position in the confined space of "under the stairs", we should follow the DNO's PPE rules? Having said that, the RCDs we are talking about are supposed to be operated (and the test device pushed) by ordinary persons.
I know that I'm always more wary of things going wrong "under the stairs" than in front of DBs with good access and lots of space.
Since it would appear that there is a bone of contention over the test button and when to operate it given the 'risks', perhaps a simple solution might be to delete that part of the standard which actually requires an integrated test button to be included by the manufacturer in the 1st place?
After all, if we are not concerned about diluting the requirements for testing with a test instrument, then why worry about the need to provide a test button which by and large, and in reality does not get used by the consumer?
That would settle the argument once and for all.
We're about to take you to the IET registration website. Don't worry though, you'll be sent straight back to the community after completing the registration.
Continue to the IET registration site
