Another neutral-earth fault???

[multimapper]

I've just had a fully qualified electrician put in a new CU. It's a split board with MCBs and both sides protected by their own RCB. There was nothing wrong with the old board, but I wanted to add an electric shower.

After installation and having passed all the tests etc we turned on the pre-existing immersion heater. The MCB for that circuit tripped, but the RCB didn't. He said that therefore it must be a neutral-earth fault, probably caused by corrosion of the 20 year old element. (I have purchased a new element and will get it fitted when I find a plumber who can shift the old one.)

I don't understand why a neutral earth fault should trip the MCB but not the RCB (surely if some current is going to earth that should trip the RCB, unless the MCB manages to cut out first), nor why with the old board I didn't get a problem (so there wasn't an overload - my understanding is that an MCB cuts out only on overload). Before the change the immersion worked fine. I'd have quizzed the electrician but it was getting late in the day.

Can anyone explain to me? Please? I only want to understand.

John

PS "Another neutral-earth fault???" - just to show that I have tried finding a solution in this forum.

[Adam_151]

Its not a neutral to earth fault if its tripping the MCB

Its either a fault of negliable impedance between line and either neutral or earth, or possibly (as the immersion heater circuit was fine before and its a new MCB) a faulty MCB

Is the MCB of the correct rating? Does it feel 'right' (ie, not slack)? Does it tripo straight away or is there a delay?

[starspark]

before buying new element did he not think to test the old one??or at least disconnect from switch and check cable to switch wasnt causing fault??probably is element but be best to check first.is there an off peak and boost element?or just off peak?

[multimapper]

Thanks Adam

Quote:

Its not a neutral to earth fault if its tripping the MCB

Thanks for reassuring me on that.

The 16A MCB was supplied as a new one (although my old ones went back into the same bag as the new one came out of!). Its action feels right to me, just like all the others, and it trips immediately. I'm sure that if the electrician had had any doubts about the MCB he'd have changed it.

So do you think that I misheard neutral-earth for line-earth and that there has been a coincidental failure of the element. That seems most probable, if not likely, to me (I won't quote Sherlock Holmes). Assuming that the MCB is working. I have every faith in the electrician, he's well qualified and had years of experience, and his reaction to the fault was immediate and confident. I'm sure that when I manage to fit a new element the problem will disappear. But I won't understand why!

Later ... I have just measured the resistance of the element (totally disconnected) with a Caltex digital multimeter

The live-neutral impedance is about 100K, so has presumably blown. I've checked and rechecked this. May be under load it would be different.

The neutral earth impedance is - well, what would you like? It depends on which way around I use the probes, and which meter range I use. One way around I get o/c. The other way around I can get 25k up to 0.35M and 3.5M on different ranges. Which would sound to me as though there definitely is a neutral-earth connection, but that the corrosion is making it semi-conducting.

So a new element may well resolve it, even if I don't understand what is going on at the CU?

John

[multimapper]

oh, I forgot to add that the meter is coming up with a negative resistance!!! I don't know what that signifies.

[multimapper]

starspark wrote:

before buying new element did he not think to test the old one??or at least disconnect from switch and check cable to switch wasnt causing fault??probably is element but be best to check first.is there an off peak and boost element?or just off peak?

Thanks Starspark

Yes, he disconnected the element and the fault disappeared, so definitely(?) something wrong with the element.

Yes, there is an off peak and a boost element but I wanted to keep the post as simple as possible. The boost seems perfectly okay.

Actually, thinking about it, the off-peak element may have failed some time ago. Neither element has been used for a long time, except for a recent test by another electrician. It's more than likely that it was just the boost element that he tested.

It's one of those control units that is solely powered by the 'normal' suppy and has its own clock synched with the off-peak times so it's difficult to know. In fact, the circuitry won't let you have both elements on at the same time so I was wrong to say that the element worked before, it wasn't tested unless the first electrician changed the time on the clock to switch in the off-peak element, which I doubt.

It amazes me how an apparently simple problem can be so complex!

In an earlier reply I implicitly suggested that an element that was reading a high resistance could short when mains voltage was applied. Is that possible? That would seem to resolve my problem.

Thanks for your interest

John

[ericmark]

The resistance between line and neutral is likely to be around 17.5 ohms the impedance with be that but measuring DC can with some heating elements be different.

The resistance line to earth needs to be higher than 16K ohms not to trip the RCD but should be more like 1M ohms. (assuming over 15ma may trip)

The resistance neutral to earth could be very low and still not trip but should be 1M ohms.

So if line to neutral is showing as 100K ohms (Since cheap meters use DC to measure this is resistance not impedance) and it trips the MCB but not RCD I would guess at a measuring fault somewhere as a 100K ohms should not trip a 16A MCB. May be you measured after the central heating had got the water hot enough and switched the immersion heater off?

To trip immediately one would be looking for short circuit around 1 ohm not K ohm range and I would be looking for a trapped wire.

If you turn off immersion heater local to heater does the MCB still trip? Something does not seem right and I would want to do further tests before changing heater.

[Spark123]

With heating elements the resistance is the same as the impedance as they are resistive devices, however as they have a positive temperature co-efficient as the element warms up the resistance will increase to a point where the heating effect and the resistance ballance each other out and the load stabalizes, i.e. the resistance cold is lower than the resistance hot.

To test insulatation you need to use an Insulation Resistance tester set to 500v to check the wiring in your case. A multimeter will give you some indication but not a terribly accurate one. You can also use the IR tester to check L/N to Earth on the element.

I have seen meters give a negative resistance reading when there is a voltage across the terminals.
Did you disconnect the supply (including the neutral conductor) before conducting the tests?

[multimapper]

ericmark wrote:

So if line to neutral is showing as 100K ohms (Since cheap meters use DC to measure this is resistance not impedance) and it trips the MCB but not RCD I would guess at a measuring fault somewhere as a 100K ohms should not trip a 16A MCB. May be you measured after the central heating had got the water hot enough and switched the immersion heater off?

It is a cheap meter. The water has not been heated. I measured directly across the element terminals, not through the stat.

Quote:

To trip immediately one would be looking for short circuit around 1 ohm not K ohm range and I would be looking for a trapped wire.

no sign of a trapped wire

Quote:

If you turn off immersion heater local to heater does the MCB still trip?

no, it doesn't

Quote:

Something does not seem right and I would want to do further tests before changing heater.

Would be nice to understand better, but replacement is cheap (although not easy) and if it all works okay I'll be happy!
Thank you

[multimapper]

Spark123 wrote:

I have seen meters give a negative resistance reading when there is a voltage across the terminals. Did you disconnect the supply (including the neutral conductor) before conducting the tests?

Yes, totally isolated it.

But ... in this isolated state I get a reading of 600mv across neutral to earth. So I guess that my funny readings are being screwed up by some electrolytic action in the corrosion.

Btw, I just hit 'ignore' when I meant to hit 'thanks'. Is there a way of undoing ignore? Hitting it again doesn't flop it back.

Thanks

John

[Spark123]

I could tell you but you're ignoring me

[multimapper]

Oh no I'm not!

[Spark123]

OKay, you figured it out

Not sure on an electrolytic action causing the PD, it is possible that thermocouple action can cause it but not that likely.

Usually there is voltage between N&E, smallish figures in a TN system unless you disconnect the neutral completely.

[bernardgreen]

multimapper wrote:

Yes, totally isolated it. But ... in this isolated state I get a reading of 600mv across neutral to earth. So I guess that my funny readings are being screwed up by some electrolytic action in the corrosion.

Good quess and correct.

But to trip the MCB there has to be an overload, The MCB trips when you turn the immersion on. Are you sure that nothing was done by the electrician to any of the wiring associated with the immersion. Did he change or inspect or tighten terminals in the switch for the immersion heater or replace any cables.

Does the flexible cable from the heater go direct to the switch or does it go to an outlet next to the tank and the switch is located elsewhere ?

Could a screw or nail have been driven through a cable about the same time the CU was being changed ?

[multimapper]

Spark123 wrote:

OKay, you figured it out Not sure on an electrolytic action causing the PD, it is possible that thermocouple action can cause it but not that likely. Usually there is voltage between N&E, smallish figures in a TN system unless you disconnect the neutral completely.

I've got three elements to play with: a brand new one in its box, the boost one in the tank, completely disconnected, and the one in question likewise. Only that one gives such a (relatively) large PD. To reiterate, no heating has been going on.

Thanks