Making twisted cable

[Mike Tomlinson]

In article , [e-mail] writes

Try it sometime, you'll see what I mean

[Mike Tomlinson]

In article , Dave Plowman (News) writes

Saw an ad, it's on special offer in Wickes for 22.50

[Dave Plowman (News)]

Even better. ;-)

[Dave Plowman (News)]

Not on their website. They do have 1.5mm TW&E for 22.49 - but for 50 metres.

[tony sayer]

Cable resistance Dave...

Some POE equipment is quite hungry thats why they tend to use high "ish" voltages like 48 volts well some of the POE supplied Microwave units we install are ....

[Rick Hughes]

Nope ... the twists are there to cancel out induced noise. Why not just use Cat 5 or Cat 6 ... cheap as chips.

Each pair in these cable also has a different twist so pairs don't interfere ... so if you don't get good results on first pair, try another.

[Dave Liquorice]

Except they don't, they are there to get as equal as possible induced noise in both wires. It is the receiver looking only for a difference between the wires that doesn't "see" the equally induced noise.

I've got to admit being a bit skeptical about the makers cable specification. Now if it wanted a couple of amps @ 24 at the remote end I could understand it but 200 mA?

Wanders of for Cat5 spec. wonkypedia: DC-Loop resistance 0.188 ohms/m. 0.205 mm^2 CSA, max current 0.577 A.

DC Loop:

0.188 * 100 = 18.8 ohms but that is loop, ie there and back. So one wire is 18.8 / 2 = 9.4 ohms. We would parallel the two wires for each supply leg so total resistance is 9.4 / 2 = 4.7 ohms. So at 200 mA there is 0.94 V drop per rail giving 22.12 V at the far end. Is it really that fussy?

CSA: Doubling up so 0.410 mm^2. What was the makers spec?

And yes, just twisting up some twin isn't likely to work. It would be difficult to get a short enough twist length for the twists to be effective.

[Dave Liquorice]

In this context irelevent. My badly worded sentance try:

"But it's this splitting of the power legs across two pairs that intrigues me."

That is a pair carries +V on one wire and -V on the other. Rather than +V or -V on both wires of a given pair.

Yep, Isn't the top power for PoE 25 W and that should be available on the end of maximum segment length of 305 m?. As you say you up the voltage, to lower the current and thus resistance based losses.

[meow2222]

8><

Why would they spec low R cable if it works with high?

I cant see why. Bell wire would be a cheap candidate, or better low v singles.

NT

[Rick Hughes]

CAT 5 /6 certainly does reduce induced noise ... I was intimately involved in UTP purchasing .... it's not perfect but it does improve s/n ratio.

Just run out a piece of CAT5 and try it ... or CAT6 if you have it an earth screen at one end only.

[meow2222]

or connect it with a resistor to try it

NT

[tony sayer]

I might be wrong but AFAIK it does do that with some equipment's..

i.e. one pair carried the Pos volts the other pair the Neg usually Brown and Blue. Green and Orange the data..

I doubt you'll get that much thruput over the 305 metres. We had one the other year that ran outa puff around 80 metres and dint need that much data thruput . We since replaced it with Fibre fantastic stuff that:)..

Yep..

[Mike Tomlinson]

En el artículo , Dave Liquorice escribió:

Maximum segment length of twisted pair Ethernet is 100m.

[Mike Tomlinson]

En el artículo , Dave Plowman (News) escribió:

Yes, I missed the magic word 'flex'. Apologies if I led anyone up the garden path.

[Mike Tomlinson]

En el artículo , Dave Liquorice escribió:

You're just being pedantic now. Rick's meaning was clear.

[tony sayer]

In article , Mike Tomlinson scribeth thus

Just to be more pedantic different pairs are twisted at different rates to reduce pair to pair crosstalk;!..

[Dave Liquorice]

How many feet in 100m? 328 bother...

Why does Cat 5 come in 305 m boxes?

[Dave Liquorice]

'cause they is ignorant CCTV/alarm "engineers" who don't understand what they are doing? The otherwise very competent and knowledgable spark who installed the CCS "CW1308" here presumably bought it as it was cheap and "CW1308". But that doesn't stop it seriously breaking ADSL.

"twin" was refering to 1.5 mm T&E. Bell wire won't have much more copper in it than the paralleled up wires of cat 5 and in this application would reuire three lengths making up, intercom, +V and -V supply. Singles have relatively thick insulation, preventing the conductors being twisted closely together and it will also inhibit getting enough twists per unit lenght to be effective and again you'd need to make up three lenghts.

[Mike Tomlinson]

En el artículo , Dave Liquorice escribió:

You didn't write the software for the Mars orbiter, did you? :)

It's always been 300m afaik. Maybe you got one with a bit extra for luck?

[Andy Burns]

If only it was just one and a 64th of an inch out

Not IME, 1000ft or ~305m