Maybe OT electrical

[FMurtz]

This is a bit meandery but here goes .

I am getting befuddled, I asked google about current being constant for a given cable for differing voltages it goes on about tables and derating etc Which I can find for myself. It all started at a non electrical NG.. In regards to a question about a 20 metre cable for a 20 watt LED at 12 volts

I said

2.5 sq mm is plenty for one 20 w lamp at that distance. in 240v house wiring 2.5 twin and earth can be be rated at 20 amps.

Next poster.

240v being the key comment there, and one which is completely irrelevant to the question that was asked. 12V changes the landscape completely.

Then I said.

Stick to mechanics, current carrying capacity of cable is the same at different voltages and as long as you comply with ohms law you will be all right.

Then a boffin butted in and said, (at this time he was under the impression that it was 20 amps not watts)

No, it is *not* the same at different voltages, not by a long chalk.

The voltage drop for a given current at a given length *is* the same. Neglecting the issue of waste heat, this means that the voltage drops to zero at a much shorter distance with a lower voltage.

Example : 60A through 65mm^2, you are down to zero before 400m at 12V (ie 100% wasted as heat in the cable), at 220kv, you are less than 3% down at 200 km with the same cable.

the current will be higher at 12 volt but still much much less than

20 amps

2.5mm^2 is unacceptable if someone is paying you to do the job.... The loss exceeds the 3% maximum allowable loss. But it is unlikely to be hazardous.

[Cynic]

Start off by realising that 20 watts at 12 volts requires a current of 1.67 amperes.

[Bob Eager]

and 2.5mm (sq) cable is about 7 ohms per kilometre!

[John Rumm]

It changes it somewhat since voltage drop has a more significant impact on performance. However it does not change the maximum current carrying capacity of the cable itself since that will be dictated by the I^2 R losses.

So 20W at 12V suggests 1.67A. So in terms of the current carrying capacity of the cable, there is nothing to worry about from the cable's prospective.

However, voltage drop on 2.5mm^2 cable will be 18mV/A/m, so for a 20m cable we get 0.018 x 1.67 x 20 = 0.6V, or a 5% drop. For UK wiring regs that would be out of spec since only 3% voltage drop is allowed on lighting circuits.

If you must run the 12V on a long cable like that, then you will need at least 4.0mm^2 (which will be very very slightly out of spec, but not enough to worry about). Note we are also assuming there is zero voltage drop on the mains side of the supply.

(much easier to run mains most of the way, and stick the transformer / PSU nearer the lamp)

[Dave W]

So 100 metres would only waste about 1V of the 12V.

[Dave W]

That's for one conductor; twin would be 14 ohms/km, = 0.014 ohms/metre. 1.67A would drop 0.023 V/m, = 0.46V in 20m, about 11.5V at the load.

[Dave W]

That's for one conductor; twin would be 14 ohms/km, = 0.014 ohms/metre. 1.67A would drop 0.023 V/m, = 0.46V in 20m, about 11.5V at the load.

[Andy Burns]

Plus the same again in the other conductor ...

[Dave W]

Assuming 0.0074 ohms/metre. 1.67A would drop 0.012 V/m, = 0.24V in

20m, i.e. not much drop in 12V.

[Dave W]

20W at 12V takes 1.67A. Assuming 0.0074 ohms/metre wiring, 1.67A would drop 0.012 V/m, = 0.24V in 20m, i.e. not much drop in 12V.

[Andrew Gabriel]

That's fine for DC, and near enough also at 50Hz.

At the 20-50kHz that an electronic transformer puts out, you'll get next to nothing out of the other end, because skin effect becomes very significant (each conductor looks more like a hollow pipe with much small effective cross- sectional area, because the electricity will only flow near the surface at the higher frequency).

I had to use 4mm² to carry 8 amps just 3 metres. In another place, 2.5mm² extra flexible wire (loads of really fine strands) worked too. (In theory, the stands should be insulated from each other in the wire, but in practice the connection between them was poor enough that it worked without worrying about that.)

You might also end up with a radio frequency interference problem if you attach long wires to electronic transformers.

I have also made use of the skin effect in another application. I wanted a trace heating wire against a mains water supply that I couldn't bury deep enough. I taped a

1mm² T&E to it, shorted at the far end. Powering it with a 250W 12V electronic halogen transformer (straight in to the shorted cable) generated a load of 200W in the 17m of the cable, which at something over 10W/metre I thought was a suitable heating power. 17 metres of 1mm² T&E plus an electronic transformer was vastly cheaper than heating cable.

[The Natural Philosopher]

Christ on a bike.

The voltage NEVER drops to zero except on an infinitely long cable...or one of infinitely high restistance

[The Natural Philosopher]

I ran two 50w/12v spots off 2 meters of T & E - ring mains spec shit.

They were noticeably dim, and still are...

[The Natural Philosopher]

Bollocks.

Skin effect at hundreds of Mhz perhaps. Not Khz

[Harry Bloomfield]

FMurtz explained :

Impossible !

The voltage cannot drop to zero, because at zero volts there would be zero current flow - zero current flow means no voltage drop.

[invalid]

With a 0.3V drop the only thing dim is TNP. What makes you think he is making it up?

[John Rumm]

That's just the resistive losses... some of these SMPSU lamp drivers have an output (either AC or rectified DC) of 10s of kHz, and so you can see several additional reactive ohms just from the inductance of the two parallel conductors in the T&E.

[FMurtz]

This poster is fairly technical (or thinks he is) and often buts in with this sort of stuf on very simple discussions and is being used to make me out to be an idiot for my statement, "

2.5 sq mm is plenty for one 20 w lamp at that distance. in 240v house wiring 2.5 twin and earth can be be rated at 20 amps. "

and "Stick to mechanics, current carrying capacity of cable is the same at different voltages and as long as you comply with ohms law you will be all right. "

[Clive Arthur]

Copper, room temp, skin depth 0.3mm at 50kHz.

I don't suppose there's any chance whatsoever that you'd admit to being surprised by that, but it caught me out when developing a power line comms system a while back.

Cheers

[Roger Hayter]

Yes, the Litz wire at longwave story is surprising. Above MF other losses dominate.