Wall plug a fire hazard

[SeanB]

Shoot, then you should not see the stuff that passes for wiring here. No plug, no problem, we have matches! Wiring using single core phone wire, or steel fencing wire. Glow in the dark illegal connections, complete with the earth return, and the tree insulators at chest hight using now bare wire.

There is a manufacturer of extension cords that has a reel that is rated for 15A fully wound, using a special cable with a spiral ridge on it to allow heat to dissipate and to allow air in to keep the cable cool. Comes with a 15A thermal breaker as well, and well regarded by the tradespeople, even if it is more expensive.

As to plugtops, there are cheap and nasty ones ( made from rolled sheet nickel plated brass) that I do not trust at over 2A, and other cheap ones that will happily handle 15A, even though they are rolled sheet as well, but thicker and with a better cable clamp and screw terminals. Others are solid brass ( drilled in the centre to save cost) with a decent screw connection. The moulded ones are notorious for cooking internally, and some of the chinese cords are made using 1.5mm wire for the supply leads and 1mm for the PE conductor ( 1.5mm on a good day, but generally 1mm plus a hair) that break after a short period of use.

I tend to make my own, using cable, a plugtop and an industrial socket outlet in a steel box, with a proper cable gland ( not that funny style the USA uses, these are actually waterproof to a degree) to keep the cable neat.

[Zero999]

Extension cords are rated for operation in open air, not coiled up in a bucket. That decreases cooling enough to be dangerous.

That's pure assumption; you'd be hard pressed to find any sort of derating specification on any commercially available extension based on physical orientation. Suppose I decided to use an outdoor extension cord fully stretched out on 125F pavement, let alone the sub-freezing ambient temperature when the electrical fire occurred.

It's no assumption, it's true and PVC cable is normally rated at something like 85^oC with a typical extension lead only being rated to its full load at 30^oC.

[PetrosA]

Wind a cable up and it becomes a transformer. Simple.

[slateraptor]

Wind a cable up and it becomes a transformer. Simple.

I was thinking more inductor, but yeah, makes sense how there's no possible way an extension cable's rating can hold in both wound and unwound unless worst case was taken into account...although I still maintain that I've never seen a derating spec. My reasoning was def out of line.

[alm]

I believe the inductance is pretty insignificant at 50/60 Hz (feel free to do the math or measure it if you're curious). It's all about the lack of dissipation.

[IanB]

Since the cable has out and back conductors attached side by side most of the inductive effects would cancel out anyway.

[SeanB]

If you think it is inductance then take a reel of cable ( a full reel of 1mm wire still in the shrinkwrap) and pass 5A through it from a 12V battery. Use a 55W headlamp bulb as a load, and leave it for a hour before measuring the temperature rise inside the bundle. That will show you just how much heat is generated in wiring in normal use.

[Zero999]

If you think it's the inductance you need to brush up on your electrical principles: remember inductance doesn't dissipate any power, it only stores energy. If the cable were a superconductor it wouldn't make any difference to the temperature rise whether it's tightly coiled or unrolled. It' only heats up more when tightly coiled because there's nowhere for the heat to escape to.

[SeanB]

Precisely, it is heat that is generated from I2R losses that is the problem, the inductance will be negligible, as the 2 cores are pretty much doing common mode rejection so the only inductance would be a very low leakage inductance. That is why I said use a 12V DC source, as the inductance will be of no consequence with a long time of constant current flow. resistive losses will predominate.

[G7PSK]

There is another source of heating with a coiled wire, if the reel or container is metal you could get eddy currents.

[GeoffS]

our gpo's are rated for 10A, but any good sparky would run wire rated to 13A, then again, they are hard to find today, heck its hard to even find ones with a valid or non expired one these days,

I have seen some of the dodgier ones running 7.5A rated for GPO's, so yeah, if you need 3KW, use a 15A gpo, worst case they run 13A rated wire, and that still covers 3KW

GPO's in Australia  are required to have the same size cable whether they are 10A, 15A or 20A. Cable size is 2.5mm²

[IanB]

All this talk of GPO's is confusing me. Back in the UK that always used to stand for General Post Office 

[GeoffS]

It  means the same things in Australia. Say GPO to most people and they think you mean the post office. 
If on the other hand you say power point, they think you're talking about a software package.

[IanB]

In Britain we just call them "sockets".

[vk6zgo]

The name used by the licencing authorities in Australia is GPO,which is an acronym of "General Purpose Outlet".
There are some strange names around;The company I worked for as a lad called the ordinary mains plugs that you wire onto appliance cords ,"Plug Tops"!

I don't know how many younger Australians do know the other meaning of GPO,as the post office has been "Australia Post" for nearly 40 years,& that organisation hasn't placed much emphasis on anything that sounds like "post office"----they're "Australia Post shops"!

Oddly enough,the term GPO in Oz meant a building which was the central post office for each Capital City,rather than the name of the whole organisation which ran the post offices  & the 'phones,as in the UK.
The name of the organisation in this country was the Postmaster General's Dept ,& the acronym was the "PMG".
OK,I know, "Postmaster" is one word,I didn't come up with it!
The PMG was split into "Australia Post" & "Telecom Australia" back in the 1970s

[ciccio]

Back to the original post..
Once I was testing a "medium size" power amplifier, about  800 VA consumption, when I notice a strange smell: burnt plastic, and what we call "cooked copper".
I switched it off, and checked every component, but everything was OK.
The smell came from the mains cable, a standard pre-made molded IEC socket labeled at 10A, with a cable labeled as 3 x 0.75 mm2, and standard Italian 10 A  plug.
The  cable's outer insulation was melting and it damaged even the bench-top (a black scar is still there).
I disassembled the cord and checked  the cable conductors' copper: it  was not 0.75 mm2, but less than 0.20..
The cord came with a PC power supply, and maybe it would operate correctly with a 400 watt load,  but it was a real fire danger,  being it interchangeable with other devices.
This is not the first time I encounter Asian-made mains cables where the manufacturer was saving on copper size: copper is expensive.

[Zero999]

Precisely, it is heat that is generated from I2R losses that is the problem, the inductance will be negligible, as the 2 cores are pretty much doing common mode rejection so the only inductance would be a very low leakage inductance. That is why I said use a 12V DC source, as the inductance will be of no consequence with a long time of constant current flow. resistive losses will predominate.

It woudn't make any difference if you used one reel for the live and one for the neutral, the extra inductance would not increase the power dissipation at all.

There is another source of heating with a coiled wire, if the reel or container is metal you could get eddy currents.

No the common mode rejection would prevent that, the live and neutral conductors are close enough to cause the magnetic fields to cancel one another out so the eddy current will be negligible.

Back to the original post..
Once I was testing a "medium size" power amplifier, about  800 VA consumption, when I notice a strange smell: burnt plastic, and what we call "cooked copper".
I switched it off, and checked every component, but everything was OK.
The smell came from the mains cable, a standard pre-made molded IEC socket labeled at 10A, with a cable labeled as 3 x 0.75 mm2, and standard Italian 10 A  plug.
The  cable's outer insulation was melting and it damaged even the bench-top (a black scar is still there).
I disassembled the cord and checked  the cable conductors' copper: it  was not 0.75 mm2, but less than 0.20..
The cord came with a PC power supply, and maybe it would operate correctly with a 400 watt load,  but it was a real fire danger,  being it interchangeable with other devices.
This is not the first time I encounter Asian-made mains cables where the manufacturer was saving on copper size: copper is expensive.

Even 0.75mm² would be too small, for 10A you need at least 1mm² cable.

[alm]

The standard C13 IEC power connector is rated for 10 A, so a mains lead with an IEC connector should have at least 1 mm² conductors. I don't believe there's a grounded connector rated for 6 A in that series, which is why you see plenty of equipment with very modest power draw equipped with these sockets. You sometimes see the C6 connectors (rated for 2.5 A) for portable equipment, but most scopes, for example, draw way less than 10 A but still ship with C13 sockets.

I'm not surprised that companies ship cords with thinner conductors with these products, although it makes it very confusing for the consumer who assumes that the cords should be interchangeable.

[IanB]

My Rigol came with a very fat power cord, making it look like it was designed for some huge and power hungry piece of equipment. However, I suspect mechanical strength and integrity is part of the safety equation. It is not only about conductor size.