Ethernet point no internet

[Barnaby Stedman]

i would suggest he explains in more detail what he means by there being no connection

 

[davesparks]

I did do both ends the same wasn't sure what the correct way was as the colours go half and half on two if them. Didn't know you were supposed to not strip too much back or not untwist did by points like that on site for years thought it was neater/easier

Yes, there are the two standards, A and B, that's why there are two different colour schemes labelled. Most people stick to B.

Yes it's easier, whether it's neater or not I'm not sure, but it's wrong.

Did you not take the time to read up on the correct way to terminate the cables before you started installing them? It doesn't take long and it's usually in the instructions that come with modules and patch panels.

 

[robd]

Didn't see any but there may have been some. In fairness I haven't questioned it until now, haven't done ethernet sockets before but hundreds of bt sockets which I hadn't had any problems with. I do take your points on board though, thankyou. To my mind if the colours match at both ends which they do and I have tested to check they shouldn't stop anything working. But do you think the fact they are untwisted would affect the quality of connection ?

And I only meant neater in a sense that they weren't tight without any slack

 

[Strima]

Are the patch leads straight through or crossovers?

 

[Mark42]

That socket is wired wrongly. Swap your white/grn and org and it will be sorted. These sockets are confusing - be sure the follow the 'B' column on both sides.
If you're wrong, but consistently wrong, both ends will of course test OK electrically, but the four balanced twisted pairs will be mixed up, affecting the signal.
Also, as others have said, leave the pairs twisted until the last possible point! (And note the frequency of twist in each pair is different. No one said network cabling, for maximum performance, was easy )

 

[robd]

That socket is wired wrongly. Swap your white/grn and org and it will be sorted. These sockets are confusing - be sure the follow the 'B' column on both sides.
If you're wrong, but consistently wrong, both ends will of course test OK electrically, but the four balanced twisted pairs will be mixed up, affecting the signal.
Also, as others have said, leave the pairs twisted until the last possible point! (And note the frequency of twist in each pair is different. No one said network cabling, for maximum performance, was easy )

Thanks mark that's a great explanation, I hadn't thought of that technical side so thankyou

 

[robd]

Are the patch leads straight through or crossovers?

Not sure Strima, sorry. Customers leads. He is quite a bit cleverer computer wise than me so didn't really question it.

 

[Strima]

Not sure Strima, sorry. Customers leads. He is quite a bit cleverer computer wise than me so didn't really question it.

It's not a common issue but it has been known to happen.

 

[robd]

It's not a common issue but it has been known to happen.

Thanks I will check when I go back

 

[Octopus]

I got a 20m lead with a plug on each end so I could watch them both together, they all flashed through in sequence

That's odd to say the least.

 

[HappyHippyDad]

That socket is wired wrongly. Swap your white/grn and org and it will be sorted. These sockets are confusing - be sure the follow the 'B' column on both sides.
If you're wrong, but consistently wrong, both ends will of course test OK electrically, but the four balanced twisted pairs will be mixed up, affecting the signal.
Also, as others have said, leave the pairs twisted until the last possible point! (And note the frequency of twist in each pair is different. No one said network cabling, for maximum performance, was easy )

What does this do Mark? How does leaving them twisted together help?
This has been an interesting thread!

 

[Spoon]

What does this do Mark? How does leaving them twisted together help?
This has been an interesting thread!

I've been told to leave them twisted as must as possible to reduce noise between pairs.

 

[Mark42]

What does this do Mark? How does leaving them twisted together help?
This has been an interesting thread!

I’m no expert on this, but do like to understand things, and know it’s to do with using a ‘balanced signal’ for low-noise transmission.

On each pair, there is a +ve-going and -ve-going signal, with only the difference them being detected at the far end. If there’s external noise induced on the line, it will usually appear on both wires equally, so remains undetected, as any noise waveform simply provides a signal on both wires equally, not affecting the difference signal. Result: a magically clean signal in a noisy environment!

The twisting reduces cross-talk between pairs in a multi-pair cable. Consider if, due to the extreme proximity of all those little wires, one of a pair might receive a larger crosstalk signal that its partner, it will be unbalanced, affecting the difference signal and so be resolved at the other end as noise, messing up fast data comms.

But if the pairs are twisted, the crosstalk hits one strand first for a bit, then the other for a bit, and so on. Therefore the crosstalk signal is magically applied to each pair’s partner equally. It then ‘looks’ like external noise at the other end and is unresolved.

Twist frequency differs so that different pairs don’t keep crossing at the same points again and again, reducing the cancellation effect.

Clever ennit?

I think I've got that essentially right, but if there are any comms network designers on here who know better, please take over!

(cf. Balanced professional microphone cabling using 3-pin XLR connectors and shielded two-core interconnects).

 

[GeorgeCooke]

This is an excellent example why sparks should not do a network engineers job. (or aerials for that matter).

 

[Charlie_]

You have got the terminations all mixed up