![](https://seccdn.libravatar.org/avatar/5a9ea428db8a0b542d4ba6de48e846a3.jpg?s=120&d=mm&r=g)
On Friday 18 April 2003 8:32 pm, fsanta wrote:
On Friday 18 April 2003 20:16, Matt wrote:
On Fri, 2003-04-18 at 12:04, fsanta wrote:
I've always done this at *both* ends of a straight through cable:
pin1 blue stripe pin2 blue pin3 orange stripe pin4 orange pin5 green stripe pin6 green pin7 brown stripe pin8 brown
Mate of mine says it's all wrong. Google confuses me still further. Is there anything wrong with my wiring? Thanks, Steve.
Here is a more "official" source:
http://yoda.uvi.edu/InfoTech/rj45.htm
I've known patch cables to work as long as both ends of the cable match.
Matt
Hi and thanks. So it seems that mine is totally wrong and shouldn't work. So why does it? Or maybe it's not working as fast as it should be? Steve.
There are 2 issues here: 1] Simple DC connectivity ========================= The cable carries 4 complete 'go and return' circuits. Without checking the specs, your configuration probably maps each 'go' and 'return' on the near machine to the corresponding 'return' and 'go' on the far machine. 2] AC connectivity ================== Done correctly, the 'go' and 'return' of a single circuit should be in the 2 cores of the _same_ twisted pair. Again, without checking the specs, your scheme is probably putting the 'go' and 'return' in different twisted pairs. We are talking high frequency AC here, not simple DC - ie the capacitor effect between circuits and the magnetic linkage between circuits as the currents change will induce unwanted currents in other circuits. This will increase the 'crosstalk' between circuits - ultimately beyond a certain length of cable, the crosstalk from other circuits will become significant relative to the intended signal on any circuit and reliability will fall off. In the whole tcp/ip scheme of things, packets will be dropped and retransmissions required. Can I suggest you get 500mm of cat 5 cable and strip 300mm - there is nothing like seeing it. The 4 twisted pairs may themselves be very lightly twisted together, but should fall apart into pairs of each colour and its stripe. You will see that the twists in the pairs are of different pitches. The twists ensure that over any substantial number of pitches, both the electric and magnetic fields cancel out on any given side of the pair. The different pitches of the twists on pairs ensures that the benefit of twisting the pairs is not outwitted by the colour core of 1 pair always laying on the colour core of another pair and the stripes of the same respective pairs laying on each other at the half twist points - ie the colour and the stripe of one pair will ideally both have roughly equal contact with the both the colour and stripe of any other pair. So, if you do it to the book, you get the benefit of all that hidden design. HTH Vince Littler