[vhfpmr (OP)]

Does anyone know why the actual conductor size in electrical flex is a third smaller than the size moulded on the sheath?

I've just bought some 0.75mm^2 flex that's actually 0.5mm^2, and some 1.5mm^2 that's actually 1.0mm^2. I've checked my calliper using feeler gauges, and it's accurate. Co-pilot's suggesting tolerance, and since it's stranded cable,  the packing density of circles, but I don't think that cuts the mustard.

[alancalverd]

I've never understood how these specifications are constructed, and the internet doesn't help much!
It would be interesting to separate the strands and measure them individually, to see whether in principle they could be packed to give the nominal cross section, since the strands tend to unpack when you strip the end of a cable.
 

[Eternal Student]

Hi,

    Various different standards exist even in just one country like Britain.   For example,  BS 6360  is one set of British Standards for electrical cable but there are also others.   So it will depend on the standard they (the wire manufacturer) are claiming for their product.

     One of the key criteria for BS 6360  is electrical resistance per unit length.   Historically, a typical bit of conductor wire had to be of roughly the specified cross-sectional area to achieve a low enough resistance.    However, the standard does allow for a slimmer wire to meet the criteria and it can still be labelled as having properties of a wire of a larger cross-sectional area.   If a high quality conductor is used, then the wire could be one-third smaller in cross-sectional area, exactly as your wire seems to be.

Best Wishes.

[vhfpmr (OP)]

The flex is made to BS EN 50525-2-11 which refers to 'nominal' size, then references EN 60228 for the conductors. That spec lists a maximum resistance of 26Ω/km for 0.75mm^2, and 13.3 for the 1.5. From that, if I calculate the implied minimum size for a copper conductor (ρ=16.8nΩm) it's 0.6mm^2 and 1.3mm^2, so the cables still don't appear to meet the specification, unless the spec has been relaxed since 2004, which is the newest copy I can get free access to.

https://www.jenuincable.com/uploads/file/bsen50525-2-11-2011-cables-for-general-applications-flexible-cables-with-thermoplastic-pvc-insulation.pdf
https://pgparsco.com/wp-content/uploads/2024/12/IEC-60228-English.pdf

This chart suggests that the spec hasn't changed on the latest (2023) spec:

https://www.webro.com/media/pdf/iec_60228_bs_en_60228_stranding_chart.pdf

The twin & earth cable I've measured is the same as its nominal size.

[paul cotter]

I can't imagine relatively pure soft drawn copper varying that much in resistance as Eternal Student has postulated. I am not saying this idea is incorrect, I just don't know. I have never bothered putting the calipers on anything other than very fine enamel wire.

[Eternal Student]

Hi.

I can't imagine relatively pure soft drawn copper varying that much in resistance...

   Some impurities, like oxygen and Sulphur, make a lot of difference.



   Materials & Properties: Thermal & Electrical Characteristics - Scientific Figure on ResearchGate. Available from: https://www.researchgate.net/figure/ncrease-of-the-electrical-resistivity-of-copper-as-a-function-of-the-concentration-of_fig2_341927168 [accessed 15 Jun 2025]

    That's the diagram that came to hand and it's not great because we'd have to assume continued linearity and extrapolate.      They claim very pure copper has a resistivity of 15.5  nΩm.     One-third of that is about 5 nΩm,   extrapolating the graph we see this change in resistivity if O₂ impurities account for only 0.025 %  of the mass  (or 250 parts per million).

    According to the website called  copper.org ,  standard electrical grade copper wire can have O₂ impurities of 0.03 %,   which is the sort of figure we want.     Meanwhile an expensive oxygen-free cable favoured by Hi-Fi buffs for speaker wire can have oxygen impurites as low as 10 ppm (or 0.001 %).

    I agree, it is a bit of stretch for a cable to be one-third less in area but it's not completely out of the running.

Best Wishes.

[Eternal Student]

Hi,

     I've read your ( @vhfpmr )  post about the standards printed on your wire, although I'm not familiar with those standards.   

    I just don't know who made your cables.  It could be that those cables are made in a place we shall call "Wakanda" (from the DC universe Black Panther genre) and do use a high quality conductor  BUT  it's also possible they just didn't actually pass a suitable standard despite having it printed on the sheath.   Basically, if you don't think the cables are sufficient then don't use them.   That's the safest advice I can give.

Best Wishes.     

[vhfpmr (OP)]

It's not an issue, just curiosity, the new cables are both bigger than the ones they're replacing. They're from Time Cables in Milton Keynes, but they're just a distributor, not a manufacturer, so it's anyone's guess where they originated.

The story is that I needed a new lead for the iron after 30 odd years of winding it round the stowage had fractured it, but being as I wasn't sure whether I'd used 0.5mm or 0.75mm I measured it before I went to B&Q, and was surprised to find it was 0.36mm. When I got to B&Q I found the only choice was 0.75 or 1.5 anyway, but after finding the old one was a lot smaller than expected, curiosity made me measure the new one as well.

The 1.5mm was for an extension lead which I thought was 1.0mm, but measured at 0.73mm. Whilst I was about it, I checked another extension lead and found this:


1000000077.jpg (1630.56 kB . 2729x1761 - viewed 757 times)

I've no idea how long I've been using it with no earth on one of the sockets. 

[vhfpmr (OP)]

I have never bothered putting the callipers on anything other than very fine enamel wire.

Which begs another question: how thick is the enamel?

[Bored chemist]

Hi,

    Various different standards exist even in just one country like Britain.   For example,  BS 6360  is one set of British Standards for electrical cable but there are also others.   So it will depend on the standard they (the wire manufacturer) are claiming for their product.

     One of the key criteria for BS 6360  is electrical resistance per unit length.   Historically, a typical bit of conductor wire had to be of roughly the specified cross-sectional area to achieve a low enough resistance.    However, the standard does allow for a slimmer wire to meet the criteria and it can still be labelled as having properties of a wire of a larger cross-sectional area.   If a high quality conductor is used, then the wire could be one-third smaller in cross-sectional area, exactly as your wire seems to be.

Best Wishes.

There are two problems with that.
First, nothing has an electrical resistance that's two thirds that of copper.
Even silver is only about 6% better conductor.

More importantly, imagine that we have a wire with the same resistance per unit length, but half the diameter.
It dissipates the same power (at the  same current).
But it has only half the area to dissipate heat from, so the temperature rise will be roughly twice as big. That's not good for teh insulation.
 
One reasonable answer to "Why does my cable not seem to have as much copper as it should?" is "The manufacturer lied."

[vhfpmr (OP)]

I'm not sure just how much the rating is determined by heat. A 1.5mm cable is rated at 16A, and 0.75mm at 6A, so there's 3.5 times as much heat in the large cable, and twice the thickness of PVC for it to pass through, but only about 50% greater surface area, both at the copper surface, and the PVC-air surface. Something looks strange when you plot the ratings too:


Flex Rating.png (9.8 kB . 492x293 - viewed 714 times)

It's not heading for the origin!

[Bored chemist]

Why  "twice the thickness of PVC " ?

[vhfpmr (OP)]

Because bigger cables have thicker insulation and sheath.

Here's another graph:


Flex Rating.png (14.62 kB . 492x293 - viewed 759 times)

Power per square metre of copper surface area. I've approximated the surface to a circle, I'm not sure I can be bothered estimating the effect of stranded wire on the surface area.

[Eternal Student]

Hi.

...  First, nothing has an electrical resistance that's two thirds that of copper...

   Very pure Copper can have an electrical resistance 2/3 that of the impure copper that was still considered to be a suitable grade for elctrical wire.   See post # 5 for discussion, although it's not important.

...Something looks strange when you plot the ratings too.   It's not heading for the origin!

     That's probably because of some safety margins and approximations used in their formula for assigning the ratings.   For example, there may be a simple subtraction of some fixed amount somewhere in their formula.

- - - - - - - - -
    Just a reminder and because this forum should provide safe advice:   You've now had 2 people suggesting that the wire may not actually be good enough or conform to the information printed on it.   However, you say you bought it from a well-known hardware store in the UK rather than from a minor outlet advertising on ebay or similar.   That helps a bit but even big retail outlets like that do make mistakes sometimes and stop selling a particular product line etc.

Best Wishes.

[Bored chemist]

I'm not sure just how much the rating is determined by heat.

What else?
It can't be the resistance  / voltage drop, because that depends on the length of the run.
The cable specification of (for example) 1mm^2 doesn't know how long the cable will be.
A short enough length of thin cable wouldn't add too much resistance to a high current circuit, but it would still overheat.

(There are, of course, other rules which do regulate permitted voltage drops and thus may require a heavier than usual cable)

[vhfpmr (OP)]

You've now had 2 people suggesting that the wire may not actually be good enough or conform to the information printed on it.   However, you say you bought it from a well-known hardware store in the UK rather than from a minor outlet advertising on ebay or similar.   That helps a bit but even big retail outlets like that do make mistakes sometimes and stop selling a particular product line etc.

I've measured a total of nine cables so far, some as old as 50 years, and the only ones that meet spec are the three T&E ones, even though the maximum resistance spec for T&E is tighter than for the same size flex. All six flexes seem to be undersized (as determined from maximum allowable resistance using the resistivity of pure copper), so I don't think this can be anything to do with just one rogue manufacturer.

What else?
It can't be the resistance  / voltage drop, because that depends on the length of the run.

You would think so, but it just surprises me that the dissipation per unit area varies so much.

[SeanB]

Yes they are generally smaller, mostly because the manufacturer is using less copper, and then aiming to be on the upper edge of the allowed resistance per unit length, which has a rather large variation allowed. Also remember the cable size quoted in the spec is an area, not a diameter, so Pi does rear up in there along with a squared radius, so any small errors do get multiplied a lot.

After that well some manufacturers do not bother, relying on the fact that no checking is done at point of entry, and then after that fixing it is the problem of the distributor and final seller, so they can produce miles of cable which is purported to be 2.5mm square, but which is 2mm square in reality. It might take a few years for the regulators to first find it, then get enough samples to say it is actually deficient, then issue a recall, which the UK resellers will have to pay for. But they made a lot of money off of it in the interim, and simply sell it again to the next sellers. extreme end is the cables used on cheap products, which come with either a pre moulded plug on, or with a detachable lead, like an IEC one. Seen way too many where the lead supplied claims to be rated for 10A, but does not have a protective ground on it, or where measuring the resistance you find a 2m lead coming in at 10R or more, fine for an appliance that draws say 50VA of power, but try using that cable to carry an actual 10A load it turns into a smoking mess. attached leads often you will find the plastic is thicker, and the cores inside are not copper, not even CCA wire (which fractures with a few bends) but are some few strands of a mystery wire that is akin to steel, in that it is ferrous, and which has multiple strands that often pierce the inner insulation, so your wire is in fact lethal to handle.

[Petrochemicals]

May be something to do with the "skin effect" the electricity is carried by the outer of the metal more than the inner, the flex will have a greater surface area ? Begs the question doesn't that make flex cheaper though and why is it not used more. Probably a comparable measurement.

[vhfpmr (OP)]

Skin depth at 50Hz is over 9mm, and even at RF, stranding is only beneficial if strands are insulated individually, and woven in pattern designed to minimise the proximity effect, as in Litz wire.

[Petrochemicals]

Skin depth at 50Hz is over 9mm, and even at RF, stranding is only beneficial if strands are insulated individually, and woven in pattern designed to minimise the proximity effect, as in Litz wire.

I am not sure about that. Skin effect is a known phenomenon in power lines, none of which meet your specifications.