Eric, would you be kind enough to explain in a little more details exactly what you mean by those points? I want to be certain that I do not mislead myself as to your menings, cheers mate.
Is this article a Red Haddock ???
- Thread starter ebee
- Start date
Yes, I also 'wondered'!
I can believe that a force applied to the terminals of an RCD/RCBO large enough to mechanically distort the device might possibly impair the function of the device, but I find it pretty hard to believe that any (copper) conductor of a size that would be connected to a domestic device would be capable of exerting enough force to do that. I therefore wonder whether eric is perhaps talking/thinking of some massive devices and cables in commercial/industrial installations?
As for the problem in testing RCBOs that eric mentions, that problem presumably only exists if there are some loads which cannot be unplugged or otherwise 'isolated'?
Kind Regards, John
The problem described is usually more to do with the bussbar and the way the clamp arrangement twists, making the devices sit at an angle rather than erect, the pressure against the adjacent devices can and does distort the plastic cases to the point the mechanical performance can be impaired.
I had one where a fault had 'tripped the MCB' but it was jammed on resulting on a burnt out 1mm² T&E. The cable remains was still live until I removed the CU cover and the MCB released. It took me some convincing myself that it really was what I observed. Others have said similar things.
Yes, again, I can believe that that can happen - but it's obviously different from what eric was describing, which appeared to be specifically related to conductors connected to the 'output' terminals of an RCBO [he did not suggest disconnecting an RCBO from the busbar in order to 'relieve stress' for testing].
- Joined
- 27 Jan 2008
- Messages
- 27,843
- Reaction score
- 3,373
- Location
- Llanfair Caereinion, Nr Welshpool
- Country
Spot on, but also in the early days I was using RCDs as the front end, and SWA cores can also exert some strain, the problem was testing the RCD it showed faulty, return it as faulty, and no fault found, and then started to look at the cables, bending and twisting them to remove any strain often resulted in the RCD now testing as A1, and it did not seem much strain had been removed, at least not enough to see any distortion, so did not want to remove and refit any wires, as could not be sure the wires were not causing a strain, likely 95% of the time no problem, but to be sure, needs testing after installing.
Thanks for that Eric.
Firstly my intention of checking a reading obtained from any meter (not only the RCD trip time tests) is in the hope any differences large enough to cause concern with each test - including both dead testing and live testing.
With an RCD test then both with and without the circuits connected would hopefully yield very similar results unless some characteristics of that circuit were effecting it, so provided all insulation and continuity tests along with all loads removed we might be looking at circuit capacitance or inductances causing an effect.
Better than a RCD test(s) on its own but not considering the circuit wiring (if possible to test adequately).
That way we know how the RCD itself might react but also how the RCD plus circuit added might react and any expected deviations might be dependant only on which load(s) might skew the results (we do not really have much control, if any, over that feature.
Some of the less modern loop testers required a cool down interval between tests.
Firstly my intention of checking a reading obtained from any meter (not only the RCD trip time tests) is in the hope any differences large enough to cause concern with each test - including both dead testing and live testing.
With an RCD test then both with and without the circuits connected would hopefully yield very similar results unless some characteristics of that circuit were effecting it, so provided all insulation and continuity tests along with all loads removed we might be looking at circuit capacitance or inductances causing an effect.
Better than a RCD test(s) on its own but not considering the circuit wiring (if possible to test adequately).
That way we know how the RCD itself might react but also how the RCD plus circuit added might react and any expected deviations might be dependant only on which load(s) might skew the results (we do not really have much control, if any, over that feature.
Some of the less modern loop testers required a cool down interval between tests.
Yes, one can confirm that the cabling of the circuit is fine before undertaking RCD tests. However, I think eric's concern was about the potential L-E leakage presented by connected loads.
As he implied, if, for example, there were multiple connected loads which had up-front L-E filtering capacitors, they might together show themselves as quite significant 'residual' current. If, to take a fairly bad' example, loads were connected and there were 20 mA of such leakage and a ramp test showed, say, a trip threshold of 25 mA - although that would be interpreted as 'OK', it will have taken 45 mA total to tripo - which is 'not OK'. That is assuming that, as eric suggested, the displayed result of the ramp test is the additional current the tester has had to supply to result in a trip - do you know if that is the case?
Kind Regards, John
That's why RCDs are - should be - tested with the circuit disconnected.
It is indeed - which is why I added my agreement to what eric had said.
- Joined
- 27 Jan 2008
- Messages
- 27,843
- Reaction score
- 3,373
- Location
- Llanfair Caereinion, Nr Welshpool
- Country
I can measure total leakage with ease,
but leakage for each RCBO would require me to remove the cover from the consumer unit, with a test of all 14 would need to do that anyway, but that only happens on the odd time, the testing of socket outlets which are more likely to need RCD protection is more regular.
But to take away the leakage requires a ramp test, and with a ramp test, one is not testing the time to trip. But if doing a ramp test, one does not really need the ½ test, but the schedule of test results should have leakage, ramp mA, and time for 30 mA test to see if anything is degrading, and the standard schedule does not have the option to record these results.
The same applies to line - neutral loop tests, we know it needs to be taken to work out volt drop, but since not recorded one can't see if A always been too high, B some degrading, C the loop impedance tester is likely faulty. If recorded, it would alert one to a change, and one would look at why. But if vital information missing from the form, why bother recording anything, may as well simple write pass or fail? Not talking about why test, talking about why record silly info like the maximum tripping time, we know the maximum permitted is 40 mS, what needs recording is the highest of + or - and now don't need to measure both so just the tripping time.
But to take away the leakage requires a ramp test, and with a ramp test, one is not testing the time to trip. But if doing a ramp test, one does not really need the ½ test, but the schedule of test results should have leakage, ramp mA, and time for 30 mA test to see if anything is degrading, and the standard schedule does not have the option to record these results.
The same applies to line - neutral loop tests, we know it needs to be taken to work out volt drop, but since not recorded one can't see if A always been too high, B some degrading, C the loop impedance tester is likely faulty. If recorded, it would alert one to a change, and one would look at why. But if vital information missing from the form, why bother recording anything, may as well simple write pass or fail? Not talking about why test, talking about why record silly info like the maximum tripping time, we know the maximum permitted is 40 mS, what needs recording is the highest of + or - and now don't need to measure both so just the tripping time.
Agreed. No problem, even though onr Tutor insisted it should be taken from on or near the midpoint of a ring and the end of any radial as they might be the most eronous.
I decided they must always be done with no circuit connected but better still, if the situation permits, to repeat the test with the circuit itself but no load connected , should not do any harm but might reveal a question about how it might react in real life , notwithstanding loads effecting.
Say a degree of confidence testing a new circuit.
No argument with that. It's testing with loads connected which is (in my opinion) of very questionable usefulness - since, as has been said, unless one has been able to measure the leakage due to loads, one cannot really know how to interpret the result.
It`s not long since that you could consider testing earth loops with 25A, pretty good at testing metal conduit joints and sometimes effectively welding poor joints as a by product of the test - find and fix at the press of a button???
Often produces sparks on the poor joint as a dead giveway too
Often produces sparks on the poor joint as a dead giveway too
DIYnot Local
Staff member
If you need to find a tradesperson to get your job done, please try our local search below, or if you are doing it yourself you can find suppliers local to you.
Select the supplier or trade you require, enter your location to begin your search.
Are you a trade or supplier? You can create your listing free at DIYnot Local