Making my own patch cords

[tpman]

I made my own patch cords, first timer. I made 4 cords that go from my router to my home network. I have plugs that go to 4 different locations in my house. I tested the 4 cords that I made and they each worked just fine. I am only using these two foot cords to go from the router to the plugins for my network. Heres the deal, I made them 'straight-thru' and they work, but I don't have them color coded correctly. I just went Light/dark thru the whole plug, I didn't cross the light orange and light blue wires. If all I am using them for is to go from my router to my plugs for my home network, does it really matter? I am not going thru a switch or another router to get to my home network. The wires are all the same gauge and offer no difference that I can see as far as speed/connection is concerned. Correct?

I have this posted in the Linksys forumn as well, sorry.

 

[franklin97355]

Yes it will work, YES it makes a difference. Don't post with any problems you have communicating before you make (or buy) properly constructed cables.

The answer has always been 42

 

[cyberspace]

I must disagree that it makes any noticible difference for this sort of use, surely all the wires are identical, merely the colour of their jackets differ.

A friend of mine managed a fairly large network and he insisted (much to my dismay i must add) on never following the conventions for straight-through cables - never once were there any issues of performance or connectivity.

It will work just fine, no problems at all, but I must say it's a bad habit to get into and I would suggest following the conventions for 'correct' cabling.

I would be keen to know what difference (if there is actually any) it makes? Obviously in a company someone else may come along and wonder what the heck is going on with the cables but in actual terms of passing voltage down the line...surely there is no impact.

'When all else fails.......read the manual'

 

[magistertc]

Cat 5? the wires are twisted pairs, this provides some shielding from signal interference - maybe also a little help with impedance matching. If you have two signal lines in one twisted pair, you will reduce the threshhold for interference and be more likely to send bad/unreceivable packets, the connection speed and quality will be less than optimal. In marginal situations, the cable may not work at all.

 

[cyberspace]

Thats a fair point I guess - but to me it's more of a theoretical issue than one that will actually come into practice and present a real problem.

The twisted pairs are still twisted...just not at the very ends (as with any cable!), and so long as there is still a good termination, that again will surely not cause a noticible problem.

Still, I prefer to stick to the conventions, they are there for a reason after all

'When all else fails.......read the manual'

 

[ImpetusEra]

Run pings and see how many packets are lost and what the latency is. You really should redo the ends properly.

 

[edfair]

I can't quote a source but evidently there are differences in the twist pitch of the various pairs. Not that it will make a difference in this application.
Of more significance is the type of cable. Ethernet specs call for a maximum length and within that maximum length only a portion is to be of patchcord. I'm assuming that this relates to multistrand being larger in diameter and thus affecting intercable capacitence.
And another issue is the RJ45 connector itself. There are multiple types and using the wrong type leads to failure.
I would assume that the ultimate answer would be a lan tester that could measure the cable characteristics. The last time I did any research on availability they were in the 2K plus range.

Ed Fair
Give the wrong symptoms, get the wrong solutions.

 

[cyberspace]

Well im sure for home use, if it's working, then theres nothing to worry about.

'When all else fails.......read the manual'

 

[tpman]

Thanks for the help! I did redo the ends just to cover my bases and to keep the little woman happy. She was more concerned about it than I was.

 

[mikeydidit]

This should help clear up franklin97355 post. In the cabling forum, this comes up from time to time. The main point being that patch cables are cheap to buy , are certified, and home made patch cables are one of the single most points of failure. Again with one or two devices on a home network it should be fine.

thread575-1082060




Mikey

 

[pansophic]

The wires are twisted to provide protection from common mode interference and to take advantage of electro-motive force (EMF). Over short distances, this is generally not a problem, and where the cables are longer, it can usually be averted by grounding the non-conducting wires.

But it is important to keep to standards generally, even if we don't understand WHY the standard exists.


pansophic

 

[BirkenVogt]

I have been told the difference in twist is to keep the pairs from "nesting" or locking into one another so that they are spaced more uniformly. So if you were to use the brown/brown-white pair in place of the blue that would be OK electrically but I agree that it would be stupid to do so and there is no reason. Just thought I would throw that tidbit in there

Birken

 

[tviman]

You won't notice any difference in the "side-by-side" method as long as you're using a HUB (not a switch). If you connect to a switch you'll be OK as well AS LONG AS the length of the cable is less than about 35 feet. Any longer than that and connnectivity becomes a gamble and definately not at 100Mbps! You might be able to force a connection by configuring your TCP/IP to connect at 10Mbps ONLY but again, it's a crap shoot for connectivity. It's just as easy to do it correctly as it is to just lay the pairs side-by-side so do it right the first time and you'll save yourself a lot of grief later.

There's always a better way. The fun is trying to find it!