Making twisted cable

[Mike Tomlinson]

En el artículo , Dave Liquorice escribió:

So that the twists in the cable are maintained for each wire. The twists help to cancel out interference.

[Brian Gaff]

And if he really wants a project, I've done it with a piece of single core pvc wire, about 2.5 times the run, One end tied to a door handle the other walked down the garden till its taught, and then wind for the twists to suit. Brian

[Stephen]

For info, LSZH is Low Smoke Zero Halogen.... basically its a special polymer that when set on fire does not release much smoke and zero halogenated compounds.

[Stephen]

I believe from my electronic days this is referred to as CMRR Common Mode Rejection Ratio.

Its also used in Ethernet cabling, you will see four twisted pairs in there too,

[alan]

Using an electric drill to twist the cable.?

[John Williamson]

I'd use a hand drill myself.

[Mike Tomlinson]

En el artículo , alan escribió:

'tis how I do it (using a battery drill on its slowest speed).

Also need to bear in mind that a 100m length of wire won't be 100m by the time you've twisted it, it'll be roughly a third shorter depending on how may twists have been applied.

[Mr Sandman]

En el artículo , alan escribió:

'tis how I do it (using a battery drill on its slowest speed).

Also need to bear in mind that a 100m length of wire won't be 100m by the time you've twisted it, it'll be roughly a third shorter depending on how may twists have been applied.

[Mr Sandman]

En el artículo , alan escribió:

'tis how I do it (using a battery drill on its slowest speed).

Also need to bear in mind that a 100m length of wire won't be 100m by the time you've twisted it, it'll be roughly a third shorter depending on how may twists have been applied.

[Dave Liquorice]

But if you take and there are no twists(*) as you have picked a single wire from two pairs.

There could be an argument that spliting the supply across two pairs means that each pair has supply "noise" and "anti-noise" so will tend to cancel out and not radiate. But that only works if the DC supply is being fed from a balanced source, ie both legs floating, most DC supplies have one leg tied to ground/chassis. It's also making the assumption that the "noise" is the same in both legs, with a bit of active kit sucking power I doubt that will be the case, even if you sorted out a balanced feed.

I'm sure those that designed PoE and used both wires of one pair for

+V and both wires of another pair for -V knew what they were doing. B-)

(*) There might be a very long twist due to the lay of the pairs in the cable but being long won't be very effective and each wire would not be closely twisted with its twin.

[meow2222]

You'd have to do an awful lot of twisting to lose 1/3 the length. Twist til l 10% down should be plenty.

NT

[Dave Plowman (News)]

If cost is critical, TLC do twin 1.5mm flex for 40 quid 100 metres.

[dennis]

Its nothing to do with noise though. The pairs can still be used for signals while having power on them. The power induced noise on the pair will be cancelled out by the twisting.

[Dave Liquorice]

doing.

But DC won't be a problem (apart from on off) no matter what wires you use. Only AC can couple from one conductor to another.

This is true, use it all the time in the day job. "Ordinary" PoE uses the two spare pairs in Cat5 to carry +V on one pair and -V on the other.

Ish. The twists are there to try and ensure that any induced noise is identical in each wire of the pair. The receiver is looking for a difference between the wires for a signal. Noise induced equally in both wires (common mode) doesn't produce a difference so isn't "seen" by the reciever.

Anyway this isn't sorting out why Adam suggested / and / for supply and return.

If the two pairs formed a star quad it would be important to select the corect wires. I can't see how spliting the power across two pairs, one wire from each pair per leg, carries any advantage.

[Gazz]

couldn't you just install the conduit for the wires, then leave it empty, fit a trumpet bell on each end and use it like the old voice tubes on ships?

[dennis]

It doesn't use spare pairs as PoE works with gigE and that usually uses all the pairs.

Well with ethernet the two legs of a pair are connected together at each end by some copper wire, the winding in the pulse transformers.

Shoving a power supply across them is going to produce smoke at one or both ends if its plugged into a gigE switch.

I suspect adam doesn't know the pairs are looped when plugged in, or does know and he is being mischeveous.

[ARW]

Plugged in? I never suggested plugging anything in.

[Dave Liquorice]

That's why I said "ordinary". B-)

select

Only on the ones carrying data, the "spare" pairs 10 or 100 Mbps are open circuit. Gigabit is different as that does use all 4 pairs.

And is also talking about this intercom application. TBH I can't see why and intercom would want to draw 200 mA @ 24 V (4.8 W) the lock release is local. Hum, maybe that 200 mA is trickle charge for the lock release battery? Thus not that fussy about volts/current. I wonder what the electronics actually live off?

[ARW]

It's the way you are taught to install CCTV. One pair for the video signal and the other pairs for the power. It is supposed to improve the signal.

[Dave Liquorice]

or

Well a pair for the video yes, with baluns each end. But it's this splitting of the power across two pairs that intrigues me. I can't see any reason for it and have a gut feeling it's not good.

Strikes me as been passed on as this "is how I was taught" and backwards into the mists of time rather than being grounded on what advantages it might have.