Blackened live in switch

[ElwoodP]

Hi, I've just had to move a storage heater temporarily. When removing the wiring from the switch I notice the live on the storage heater cable has blackened and stiff (see photo). Any advice on how to proceed? Can I troubleshoot this as a DIYer or is it for a pro to deal with?

 

[blup]

If the blue wire was in the L terminal, yes get a qualified electrician in.

Blup

 

[Chunky19]

Thats the neutral cable and yes the problem was overheating of the connection or contacts. Buy a new quality DP 20A switch, cut the cables back to good copper and re fit. Make sure your screws are tight!

 

[ElwoodP]

Thats the neutral cable and yes the problem was overheating of the connection or contacts. Buy a new quality DP 20A switch, cut the cables back to good copper and re fit. Make sure your screws are tight!

Oof yes, neutral – brain fart typo. I'm not that stupid honest.

So the problem is likely to be the connection/contact and I can safely replace with new switch.

If I don't have enough existing cable what gauge do I replace with? Heater is 1.5KW. Current cable copper diameter is roughly 1.6mm but obviously stranded so not sure how that converts.

 

[Chunky19]

The incoming cable will most likely be 2.5mm2. The flex poss. 1.5mm2. 1.5mm2 will be fine for a 1.5kw heater. The sizes are the square area of the conductors not measured diameter. If replacing flex best to use heat resisting type, sometimes called butyl.

 

[winston1]

Current cable copper diameter is roughly 1.6mm but obviously stranded so not sure how that converts.

Measure the diameter of one strand, calculate the cross sectional area and multiply by the number of strands.

 

[plugwash]

Measure the diameter of one strand, calculate the cross sectional area and multiply by the number of strands.

This may be a good approach for coarse-stranded installation cables but I doubt it's workable for fine-stranded flex unless you have very good measuring equipment.

 

[plugwash]

If I don't have enough existing cable what gauge do I replace with? Heater is 1.5KW. Current cable copper diameter is roughly 1.6mm but obviously stranded so not sure how that converts.

A circle with a 1.6mm diameter has an area of about 2mm² but not all of that is copper, some will be air, optimally packed circles have a packing density of about 90% but I suspect that strands in a real world cable are lower.

So I would bet that the flex is 1.5mm².

 

[JohnW2]

This may be a good approach for coarse-stranded installation cables but I doubt it's workable for fine-stranded flex unless you have very good measuring equipment.

One obviously needs a micrometer or vernier caliper, but it's something I do successfully all the time (with 'fine-stranded flex' and such tools), and the answer I get is nearly always very close to an 'expected standard value'.

Kind Regards, John

 

[JohnW2]

A circle with a 1.6mm diameter has an area of about 2mm² but not all of that is copper, some will be air, optimally packed circles have a packing density of about 90% but I suspect that strands in a real world cable are lower. So I would bet that the flex is 1.5mm².

I'm not really a betting man but, yes, I would also bet that.

I don't think that, in practice, the figure gets anywhere near 90% for real-world stranded conductors (even though, theoretically, I suppose it would approach 100% as the number of strands approached infinity!). In the example we are looking at, it seems to only be about 75%.

The first table below relating to T+E cables is the one I put in our wiki, and it includes an approximate estimated 'overall diameter' for the stranded live conductors of 4, 6, 10 and 16 mm² conductors (just 7 strands in all cases). As can be seen from the second table (which I've only just produced), the figure seems to be almost exactly 78% in all four cases. One would probably expect (despite the above!) the percentage to be higher (than that 78%) with much larger numbers of strands, such as the typical ~30 strands for 1.5mm² flex.

Kind Regards, John

 

[plugwash]

I don't think that, in practice, the figure gets anywhere near 90% for real-world stranded conductors (even though, theoretically, I suppose it would approach 100% as the number of strands approached infinity!).

Assuming the strands are circular there is a limit of about 90% for perfectly straight cylinders with regular packing, but I doubt the packing in a real conductor is anything like perfectly regular, nor do I suspect it's desirable for them to be. Beyond a certain point I doubt the number of strands makes much difference.

I didn't bother going looking for it before because I was already fairly confident in my conclusion as to the size, but

has a table with the maximum diameters for each size and class of IEC standard conductor (though it doesn't have figures for the oddball 1.25mm² used for UK extension leads but rarely seen anywhere else) . Class 5 and 6 (flexible and extra flexible) cables are allowed (but afaik not required) to have a larger diameter than class 2* (coarse stranded) It lists the following maximum diameters for class 5 and 6 conductors, I would expect typical diameters to be a bit smaller.

1.0 mm² -> 1.5 mm
1.5 mm² -> 1.8 mm
2.5 mm² -> 2.4 mm

So out of the standard IEC sizes a class 5 conductor with a diameter of 1.6mm basically has to be 1.5 mm², It's too big to be 1.0 mm² and too small to be 2.5mm² (even if it was 100% copper it would only be 2 mm²). I guess it's possible it's the oddball UK 1.25 mm² size but even if it is it doesn't really matter replacing it with 1.5 mm² would be fine.

* I suspect this is why "tri-rated" cables are fine stranded, there is enough slop in the rules for fine-standard conductors to make a cable that complies with the rules for both a US standard size and an IEC standard size.

 

[Harry Bloomfield]

Strip enough of the wire, to enable the end to be easily folded over doubled, so as to almost fill the terminals, make sure the screw tightens onto the copper - what I do then, is give the wire a waggle and give a final tighten to be sure they are firmly settled in the terminal.

 

[JohnW2]

Assuming the strands are circular there is a limit of about 90% for perfectly straight cylinders with regular packing, but I doubt the packing in a real conductor is anything like perfectly regular ...

The maths is simple enough for straight, circular cross-section, strands (as you say, 'cylinders'), side-by-side but the reality is that they are usually 'twisted' ('spiral'). Whilst that theoretically makes no difference to the maths as applied to any ('infinitely thin') cross-section, I think (but may well be wrong!) that the effect of the (3-dimensional) spiralling will, in practice, usually be such as to increase 'wasted space' between strands. Indeed, I'm not yet convinced that one actually can 'spiral' a strand without distorting its cross-section such that it becomes something other than 'perfectly circular'.

Beyond a certain point I doubt the number of strands makes much difference.

I need to think a bit more about that. As I said, my initial 'intuitive' (often dangerous!) feeling was that as the strands approached being infinitely numerous and infinitely small, the amount of space occupied by copper would approach 100%. However, on reflection, I don't think that's true - so I'll see if I can determine the answer for a 'general case'.

[whilst I say 'general case' there are presumably serious constraints as regards the actual numbers of strands, since only certain numbers will allow 'optimal packing' ]

Kind Regards, John

 

[Harry Bloomfield]

Whilst that theoretically makes no difference to the maths as applied to any ('infinitely thin') cross-section, I think (but may well be wrong!) that the effect of the (3-dimensional) spiralling will, in practice, usually be such as to increase 'wasted space' between strands.

I would suggest the opposite, but only by a very tiny amount and due to deformation of the copper [1]. What the spiral will do, is to increase the CSA - the tighter the spiral, the wider will be the CSA, but only up to a point.

[1] Notice how the individual strands of the copper are deliberately deformed on much larger, less flexible cables.

 

[JohnW2]

I would suggest the opposite, but only by a very tiny amount and due to deformation of the copper [1]. What the spiral will do, is to increase the CSA - the tighter the spiral, the wider will be the CSA, but only up to a point.

I think it's hard to say, so I wouldn't be at all assertive about any possibility.

I would think that it's probably impossible to take a straight cylinder of copper and 'bend' it in any way (as is required for spiralling) without 'distorting it in some manner and I suppose I was thinking that 'theoretical maths' then becomes essentially impossible because one (at least, I !) does not know exactly what distortion has taken place. I definitely need to think a bit more about this!

Kind Regards, John