RCD trip times

[stanleyspanner]

Testing an RCD (part of a split-load CU) I found that at the rated current (30mA) it would trip anywhere between 16-17ms to not at all (even with the meter set to the 2s range). Suspecting the RCD I swapped in another from a spare box (same make - Clipsal - but almost certainly different batch as I bought them at different times from different suppliers). However the other RCD showed much the same behaviour. Both tripped reliably at 50mA. Is this typical, or are both faulty? (Is Clipsal crap? :-)

[Andrew Gabriel]

It's not typical -- I normally find they trip within

20ms at the rated trip current (not used Clipsal though).

Is your RCD tester calibration out of spec?

[BigWallop]

When was the test meter last calibrated? Is the battery still good, or is it a mains operated test meter? It could be your meter and not the RCD's you know. The most reliable test gear is connected across the supply and doesn't rely on a battery.

The mains operated test meters have crimp ends which connect with main supply leads to the RCD switching. Then the test leads are connected across the consumer side of the same switch. It then slowly increases the current to a point where it causes the RCD to trip out. They then store the info' on the display for easy reading after they're disconnected. They do have a battery to make the display work and keep the NVM chip working after disconnection, but they don't rely solely on a battery for the full test.

They're also very expensive, especially if you have to buy three like we did.

[stanleyspanner]

swapped

certainly

Each & every time? It tripped within 16 or 17ms first couple of times, but trying repeatedly I got non-operations. Maybe the RCD was getting tired? ;-) As I say though, I did test on different samples.

Whitfield says "100% of rated ... must ... trip within 200ms" and "150mA ... 40ms" which suggested to me that a (say) 30mA 30ms RCD should trip at 30mA, but doesn't have to do so within 30ms, but would trip within this time at higher currents.

Just recently been calibrated so I'd hope it'd be OK. Will try again on other RCDs (like this house cct, after I've logged off :-)

[ARWadsworth]

It did trip within 20 ms though!

Are you doing the test on both AC cycles? You need to do the test time on both cycles and record the readings with the tester set a 1/2 X, 1X and 5X settings. In all 6 tests.

The not tripping at all should be correct at 1/2X setting

Adam

[stanleyspanner]

OK, got < 20ms each and every time (about a dozen times) on my ancient Wylex 30mA whole-house[1] RCD

Seems like: No, Yes, and (Yes) then :-( [1] yes, I know. The cobbler's children ...

[Stephen Dawson]

The question is very loose.

Climbing up on to high horse :-)

If you do not know how to carry out the test that you are doing, then you will have no idea how to understand to results and whether they are good or bad.

This is why you should employ a competent electrician to carry out this kind of work. If you were competent to do this work you would not be asking this question.

Climbs off high horse.

Stephen Dawson

[Andrew Gabriel]

Yes.

Yes, they aren't required to trip in 20ms at the rated trip current, but normally they do. However, they are required to trip every time!

[John Rumm]

Which one? The OPs description seemed to contain all the detail required.

What a spectacularly useless comment.

ISTM that the OP had the right test gear, knew how to use it, and knew what results to expect.

Looks a bit like an ass from this angle....