Gigabit Ethernet wiring incorrect 2

[RadarGuy1962]

Hi all,
I recently tested a customers Gigabit Ethernet wiring and found that it did not pass NEXT due to split pairs on 3-6. Further inspection showed that it had been wired pin to pin/pair to pair as below.
Pin No. Wire Colour
1 white and green
2 green
3 white and orange
4 orange
5 white and blue
6 blue
7 white and brown
8 brown

Question is what would this do the the Gigabit point to point network?

 

[VinceWhirlwind]

I don't think the ethernet interfaces can see what colour the plastic cladding is around the wires.

The important thing is that it comes out in the correct order at the other end.

 

[RadarGuy1962]

Thanks Vince but doesn't the interfaces in Gigabit use pins 3 and 6 as pair 2 shown below:
Pin No. Wire Colour Name
1 white and green TX_D1+
2 green TX_D1-
3 white and orange RX_D2+
4 blue BI_D3+
5 white and blue BI_D3-
6 orange RX_D2-
7 white and brown BI_D4+
8 brown BI_D4-

 

[cajuntank]

RadarGuy1962, you are correct. You need to split that oranage pair at the termination side to adhere to code. Your current color scheme follows 568A starting with the green pair and should be splitting the orange pair; whereas, 568B starts with the orange pair and then splits the green pair is the more followed standard for commercial installations.

 

[RadarGuy1962]

Thanks for the confirmation cajuntank, that's what I thought.

 

[telcorookie]

White Orange
Orange
White Green
Blue
White Blue
Green
White Brown
Brown

 

[pansophic]

On short runs I would expect that the wiring will not be a problem, but if any of the runs extend to great distances, you will run into issues that will be hard to troubleshoot. The pairs are twisted to reduce cross-talk and to retain the maximum EMF. Every pair in a cable is twisted at a slightly different rate in order to achieve the data rates.

While VinceWhirlwind is technically correct (the interface cannot tell what color the wires are) the signal will degrade more quickly when the interfaces are wired with split pairs.

The best thing that you can do is wire the cables by standards, as it will be a maintenance headache in the future when someone cuts an end off of a cable and recrimps (according to standards).

If that is not possible, grounding the wires that are not in use will reduce the amount of signal degradation that you experience over distance.


pansophic

 

[RadarGuy1962]

Thanks pansophic

I have decided to rewire all of the Gigabit cable for ease of maintenance in the future.

 

[technome]

While your at it see if you can rent a CAT tester or have your telephone guy certify the wires (with a print out). As Pansophic points out you need to carefully maintain the twists, or your wires will not pass on a CAT test. Cheap continuity testers, will only check for bad connections, shorts, and mis-pairing, which create about 75% of the issues with Ethernet wiring.
Good move to re-terminate the wires.



........................................
Chernobyl disaster..a must see pictorial

http://www.kiddofspeed.com/default.htm

 

[Jahed]

hey guys

the most important thing to be noticed is meeting EIA/TIA standards for Structural cabling
got the point ?
basicaly you must act as a good engineer and run your activities scientificaly.
if you have no standard contact me i would send you the pdf file to your Email

have a good time

 

[VinceWhirlwind]

Sure, but if you walk onto a site and the cabling is non-standard but is fully working, it doesn't mean the cost of reterminating the entire site is justified.
In this case the original cabler was clearly incompetent, but there is a very, very low risk from their mnistake.

Here in Australia you would simply tell your customer to ring the original cabler - if they want to keep their cabling licence, they will reterminate it all for nothing otherwise ACMA will send out an inspector and take their licence from them.

 

[daronwilson]

I know it's an old thread, but it baffles me that this still continues to come up AND be confusing to the professionals.

First...while the interface can't tell color of jacket, be aware that there pairs are twisted at different rates, and at high speeds this can make a difference. For example if you were to terminate using the blue and brown pairs instead of the orange and green, it likely would not pass. The 'interfaces' are expected some complex signaling habits, and the manufacturer's make that happen on the pair when they are terminated properly. Don't deviate from the recognized standard.

Second...who is crimping on modular plugs on solid cable and saying that meets the standards? If you have SOLID conductors, it should be terminated on a patch panel or a modular jack rated for the service. The solid cable is transport cable, you shouldn't need to crimp plugs on it, it goes from the patch panel to the jack and doens't move after that. Patch cables are stranded, designed for that movement, and exhibit more attenuation than solid.

Third...why all the bother? Well i go through this with customers all the time. They made their own wizbang patch cable 40 feet long and are really proud of it. We plug it in, the light lights up on the LAN card, and they are pleased. I say "how many packet retries are you getting?" They say "huh, i have no idea". EXACTLY. That is why you don't mess with the physical layer, because it is tough to mearsure. You can put a piece of modular phone cord between the wall and your computer and make the interface light up. Will it transport data at 100mbs without errors? Not likely. Can you tell if it is? again, not likely.

The point is that the physical layer should be rock solid, done with the correct equipment and to the correct standard, including a quality test performed after installation.

Daron J. Wilson, RCDD
Solving 'Real World' problems