I just checked the boat's RCBO trips with the RCD test button on the socket tester and it's fine. 

 

That was with the generator off and the mains ring powered by the inverter.

A note on RCD test buttons.

They do not test that the RCD is operating within spec. The tripping current used by the button can be much higher than 30ma. So even if the button works the RCD might still not protect you.

The only way to check an RCD is with a proper RCD tester.

1 hour ago, GUMPY said:

A note on RCD test buttons.

They do not test that the RCD is operating within spec. The tripping current used by the button can be much higher than 30ma. So even if the button works the RCD might still not protect you.

The only way to check an RCD is with a proper RCD tester.

 

Well there's only so much I'm prepared to do. How many people on boats or even in houses test their RCDs with proper RCD testers? Those who do are probably in a tiny minority. 

10 hours ago, blackrose said:

 

… but yesterday it tripped when there was mains running to the boat and I switched on the battery charger. I went back outside and the generator had stopped so I restarted the generator, reset the RCD , came back in the boat and switched on the battery charger and then it was fine. 

 

No you've misunderstood. My RCD trips when the power goes off. It requires power to stay on. It was described as "non-latching", which I thought meant it didn't stay on when there was no power?

 

OK fair enough. It’s just that eg your first paragraph that I’ve quoted above seems to imply something abnormal - else why would you have written it. But of course if the generator has stopped for any reason, the inline RCD will trip because it has become unpowered.

10 minutes ago, blackrose said:

 

Well there's only so much I'm prepared to do. How many people on boats or even in houses test their RCDs with proper RCD testers? Those who do are probably in a tiny minority. 

I have a tester but rarely use it. My caravan RCD gets tested as part of the yearly service.

Problem with using a tester without an isolating transformer is  how many RCD will it take out upstream of the one you are testing. 

17 hours ago, blackrose said:

Just because you woke the topic up. [...]

 

Thanks for having interesting problems. 😀

 

16 hours ago, GUMPY said:

A note on RCD test buttons.

They do not test that the RCD is operating within spec. The tripping current used by the button can be much higher than 30ma. So even if the button works the RCD might still not protect you.

The only way to check an RCD is with a proper RCD tester.

 

An important point when fault finding. We RCD users have to hope the differences in performance don't prevent a device doing its most important duty in good time when needed...

I asked about testing the RCBO because it seemed odd (ie. significant for the fault) that it never tripped.
Possibilities -

1. it doesn't trip (broken)

2. it would trip, but the fault is in the sensitivity window below what will trip the RCBO but above the threshold for the genny RCD ~~ per Gumpy's point

3. the fault, or a significant contributor to the total causative fault, is upstream of the RCBO

4. hmm, other? I tried to be exhaustive with these options because Debugging Rules run on logic, but maybe I missed something

 

My survey of kit on my boat showed several devices that leak 0.5mA to earth just being plugged in, unless switched off at the socket switch. These add up, but not necessarily linearly (resistive vs capacitative) and the whole boat gets nowhere near 30mA.

 

17 hours ago, Ronaldo47 said:

I traced an earth leakage fault on a glass washer that had been tripping its elcb,  to its moulded 13A plug.

Plug got wet? Leak goes away if it's baked on a radiator for a week?

 

(After cutting a plug off like that, I like to twist at least one pin about 90° so it can't be plugged in. The pliers are right there and the plug is never going to be useful again anyway...)