50V AC leaking across a 5-core cable Featured

V=IR

If your measuring device has a very large R (as a multimeter typically does), then a tiny current will cause a large apparent voltage.

As to where the I comes from, you're running one wire parallel with another, so you get capacitive/inductive coupling between the two conductors. Put

However, if you reduce the load resistance (put a resistor between the wire and neutral) the voltage will collapse, because R got smaller and thus so did V. There is no source of power able to drive enough current through the load to hold up the voltage.

TL;DR this voltage is unlikely to be the source of your trip, because this current is likely to be much less than 30mA. But there may be another leakage current from live to earth - I suspect somewhere in the pump, given it gets wet. The 50V is just a red herring.

Theo

You don't give enough info. Most likely there is no leakage, you're measuring capacitively coupled microcurrent with a high impedance meter.

Yes. With my old boiler, isolating the supply (double-pole switched spur) upstairs, left 90V ac on the boiler connections. It was certainly enough to give a good shock when touched, but even a tiny load would pull it right down to zero.

Have you checked the electronics?

I’d be looking for a fault there rather than in the cable - at least unless there are signs of damage etc.

If you can isolate the cable, try testing it with DC. That will rule out coupling but, in 1m, coupling is most unlikely to be enough to be noticeable.

Well, I had a rotator cable that worked at low voltage have issues, but it was all outside and had a bend where it needed o go to the motor. I thin in that case it was heat an cold cycling that dit it it bent around nearly 90 degrees. However on such a short piece, I'd suspect the local rodents. Assuming they managed to find a way into the conduit. Brian

Not so sure about that. It is true that the voltage will vary with load, but does the device that trips itself have sufficient load to avoid this voltage detected? In the case of my rotator, it seemed that the rotation of the aerial would gradually creep out of resolution with the knob. However taking it down and using short bits of wire did not show a fault. I seem to recall that the leak varied as you flexed various bits of the hardened cable.

Brian

Without looking it up how much, there will be some capacitive cross-talk between conductors. A DVM has a very high input impedance, in the M-Ohm region, and so may well see 50VAC.

Was the cable to the boiler bunched together with other live cables? I've certainly seen stray volts using a Hi-Z DVM, but never felt them.

I could certainly feel this. My conclusion was that it was picking up from parallel cables - there is no choice in the run as it has to go between a joist and a wall, under a tiled bathroom floor and then drop vertically in a duct in the corner of the hall below, so a whole host of cables have to run the same route, as there is nowhere else accessible.

Hmm. It seems unreal for a cable barely more than a metre long. And you wouldn't expect to see the same result for the two situations (1) both ends of that 1m long core wire are disconnected and (2) one end is connected to the UV bulb and the pump motor via the terminal block in the junction box?

Of course it's conceivable that both pump and UV bulb are now presenting an open circuit. I should hard-wire the pump to test it.

I have ordered a slightly large junction box - it's a bit cramped anyway and yesterday I accidentally pierced one of the silly membranes on one of the unused inlets, so its waterproofing is now compromised. That's an hour or two of rewiring to begin with.

I will also extract my circuit board from the switch box. The on/off switch for the pump does make use of a terminal block on the circuit board because the (weatherproof but nice-looking) switch has awfully delicate connectors, so I didn't want to connect the 5-core cable directly to it. However, the only track on the circuit board adjacent to the pump switch signal is the neutral. I should still inspect it, though. There's bound to be something obviously I simply haven't noticed yet.

All this faff just for a pond pump!

Michael

That's what I thought! There is bound to be something I've missed that's causing this if it's not the pump itself, and, as per my reply to Theo above, I've now got to change the junction box anyway, so I'll be rewiring the whole damned thing! What a pain.

Michael

Oh, silly me!!

I really should have tested more of the components individually yesterday. I was convinced I must have made a bad job of the wiring or the circuit board - or possibly water ingress to the junction box or damaged cables, etc.

But it turns out it's the UV bulb. The pump is fine (at least briefly).

What happened before I disconnected it some time ago was that it ran for a bit and then tripped the circuit breaker in the consumer unit after a little while. Whatever the problem with the UV unit was, it has clearly degraded, as it trips immediately.

I'll just get rid of the silly thing.

Michael

Sometimes valour is the better part of discretion.

Faced with the need to add some more disco lights to what was a rat's nest of chocolate blocks in a biscuit tin, I horrified my boss by demanding a hacksaw, and removing the cancerous growth surgically.

A new DIN rail box went in and as pairs of wires were identified first with an ohmmeter, and then some live mains applied, the pairs were documented and terminated neatly on the rail connector.

It only took an hour or so.