Insulation resistance failure on rubber flex. Why?

[Mark42]

Two years ago I installed some garden lighting, using u/g SWA to Wiska boxes, then 3-core 0.75mm rubber flex downleads from each LED fitting.

This week the 30mA RCBO kept tripping. I assumed a luminaire fault, or moisture in a Wiska box, but no. It was leakage across all three cores of one 10m length of rubber flex.

There was no visible puncture, and no moisture ingress. Out of curiosity, I kept cutting the length in half to isolate the fault, but every bit was the same! I ended up cutting the whole length into 1m pieces. All read about 0.4 megohm (The originally uncut length was about 0.2 meg).

This was new, modern cable. What's going on?

 

[Gavin John Hyde]

Cutting the cable into 1 metre lengths just seems wrong and i do not know why but i find the image of this funny.
have a read of this
How Humidity Can Affect Testing - https://www.cirris.com/learning-center/general-testing/diagnosing-failures/218-how-humidity-can-affect-testing
You might not to the touch or visually detect moisture in the insulation but if it has penetrated the flex it could be the cause...
Was the flex you used suitable for outdoor use?

 

[KEV 1 N]

Cutting the cable into 1 metre lengths just seems wrong and i do not know why but i find the image of this funny.

How come?

 

[westward10]

Does it clear on your meter when the croc clips are removed.

 

[Octopus]

What type of flex was it?

 

[Mark42]

Cutting the cable into 1 metre lengths just seems wrong and i do not know why but i find the image of this funny.

Yeah, it's daft, but I assumed invisible physical damage or a small puncture somewhere, so I was trying to isolate the fault and investigate.
I'd already replaced the whole drop with new HO7RN; I was curious as to what the fault had been.

 

[Mark42]

Does it clear on your meter when the croc clips are removed.

Good question.
Yes, no meter trouble or dodgy fly leads. Off the scale at 1000Vdc.

 

[Gavin John Hyde]

How come?

reminds me of a comedy sketch where there was a fly between two window pains so they smashed the window to get rid of the fly but then had 100s of pieces of glass and no window. not quite the same i know but my mind works in weird ways, i have a distorted sense of humour.
it is a methodical way to locate a fault though. albeit you would then have to join another piece on or replace once fault is located.

 

[KEV 1 N]

reminds me of a comedy sketch where there was a fly between two window pains so they smashed the window to get rid of the fly but then had 100s of pieces of glass and no window. not quite the same i know but my mind works in weird ways, i have a distorted sense of humour.
it is a methodical way to locate a fault though. albeit you would then have to join another piece on or replace once fault is located.

Haha I get it now, just your first post I replied to didn't have that link on, now it's on. Was going to say it seems like moisture in the insulation.

 

[westward10]

Progressively remove the sheath and see what happens.

 

[Mark42]

Progressively remove the sheath and see what happens.

Yup, yesterday I did mess about with just two small coils of inner core, as I thought I was going mad with these readings. Laid separately on a wooden workbench: passed test, laid touching: failed.
I had no idea that moisture can infuse into rubber insulation. This stuff is used under water for goodness' sake!
I can't remember where the original cable came from. It would have been from an online supplier, probably TLC, but I'm not sure. Looking again it does not have HO7RN printed on the sheath, but the replacement does.
Is there such a thing as rubber cable which is specifically NOT for use in damp environments? That would be spectacularly pointless.
Maybe it was a cheapskate product from EBay. It's possible.
This is a nuisance: I've used it on several LED fittings, by soldering directly to the PCBs internally, to give a longer drop wire than the 30cm or so you get with the fittings, to avoid having Wiska boxes high up in trees.
To get a final diagnosis, I'll bake the coil in the picture in the oven tomorrow, then try again.

 

[PEG]

I have tested lots of Bio-filter pumps,which have failed an IR test,and not had one fail on a cable which has not had damage of some sort.
They are generally of decent quality and come with top spec flexible cable.

If the cable has not suffered damage,stress or overheating of some description,it sounds like the conductor insulation has degraded.

Maybe the maker tried coloured pasta,in lieu of a missing insulation order...

I respect your determination,in chopping it up and keeping on testing

 

[Mark42]

Update: Today I baked the faulty cable at 150C for two hours.

Now it's fine. IR is off the scale.



As many have said it must have been faulty, wet insulation. Does that mean a cr@p batch of cable, or does rubber do this anyway? I've never heard of it and would not have believed it.

As a further experiment I soaked the cable in water for a few hours and tried again. Still off the scale, even at 1000Vdc.



I've now stripped off the sheath and am soaking the inner cores.

I dislike mysteries ...

 

[static zap]

Update: Today I baked the faulty cable at 150C for two hours.

Now it's fine. IR is off the scale.

Was thinking something similar , was it all enclosed or some exposed to sun ?

 

[Mark42]

Was thinking something similar , was it all enclosed or some exposed to sun ?

It was in England. Exposed to some sun, and a lot of rain.

 

[static zap]

Instructions for a mains pond water feature , mentioned keeping cabling screened from sun , though submersible !
...Aerosol cans / holes in ozone layer ...
(or just atmospheric pollution more like )

 

[westward10]

Rubber insulation should not permit water absorption sounds like it is inferior.

 

[Wilko]

Update: Today I baked the faulty cable at 150C for two hours.

Now it's fine. IR is off the scale.

View attachment 38875

As many have said it must have been faulty, wet insulation. Does that mean a cr@p batch of cable, or does rubber do this anyway? I've never heard of it and would not have believed it.

As a further experiment I soaked the cable in water for a few hours and tried again. Still off the scale, even at 1000Vdc.

View attachment 38876

I've now stripped off the sheath and am soaking the inner cores.

I dislike mysteries ...

My take - you have shown that water penetration was the cause. But it will take "some time" for water to penetrate as before, as it took "some time" for the cable to be affected first time round

 

[Octopus]

I can't remember where the original cable came from. It would have been from an online supplier, probably TLC, but I'm not sure. Looking again it does not have HO7RN printed on the sheath, but the replacement does.

I think the answer is wrapped up in this statement ...........

 

[Richard Burns]

Standard rubber cable is not considered suitable for prolonged outside use as water can penetrate the cable over time.
polychloroprene (neoprene) the N in H07RNF sheath is considered waterproof.
but an ethylene propylene rubber sheathed cable (R) is not suitable.
However the water penetration is very slow and as you have found can be removed by baking so a cable out of water and just getting rained on and then heated in the sun should survive for a long time.
Have a look at the cable spces for RR cable and you will see they state not suitable for long term outside use.