Wiring a Wall Type RJ45 Jack

Use a PIC.

Of course. That's rot13 for "xwuls".

John

What's the story of O?

Oy vey! Such a story...

Best regards, Spehro Pefhany

-- "it's the network..." "The Journey is the reward" snipped-for-privacy@interlog.com Info for manufacturers:

Embedded software/hardware/analog Info for designers:

Wiring a Wall Type RJ45 Jack...

for a pass-through into another room.

I presume I should use the "A" terminals, though I have not a clue, I've never done anything but plug CAT-5 cables together before ;-)

Which is correct?

Thanks!

...Jim Thompson

-- | James E.Thompson, P.E. | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

| 1962 |

"When the government fears the People, that is Liberty. When the People fear the Government, that is Tyranny." - attributed to Thomas Jefferson by his contemporaries

"The Constitution is not an instrument for the government to restrain the people, it is an instrument for the people to restrain the government -- lest it come to dominate our lives and interests." - Patrick Henry

They're two different standards, both "correct". I use the B configuration, which AFAIK is more reliable for higher speed ethernet. Check your existing cables and see which they use.

I had one long cable that didn't work with the A wiring (packet loss), but did work with the B wiring. YMMV.

Aha! The cable says CAT5e, which, from the Wiki article, should be the "B" configuration.

Thanks!

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
 Stormy on the East Coast today... due to Bush\'s failed policies.

Buy a wall plate that has a jack on both sides, and wire like you are used to wiring.

they are both correct, depending on the specification. If you are passing it through the wall from an existing connector, just copy what is on the existing connector.

If you've removed the connector and don't know which one to use, you'll need a working device to plug into it. Try the 'A' first, and if it doesn't work, switch it to 'B'. Only the W/G and W/O pairs are being reversed. Normally, TIA 568A is used in buildings I've worked on, but 568B is used sometimes. The TIA standard is specified when the wiring is installed.

If it's for telephone, normally the B/W pair is used and that pair corresponds to the R/G pair on telephone wiring. Green = Tip1 = White/Blue, Red = Ring1 = Blue/white. The second pair on telephone is normally Black/Yellow, so Tip2 = Black = White/Orange and Ring2 = Yellow = Orange/White.

note: white/blue refers to a white wire with a blue stripe and blue/white is a blue wire with a white stripe.

What are you using it for? I doubt if you're using it at 100 Mhz, so the wiring wont be critical. If it is, the specs say no more than 3/4" of twisted pair can be unwound at each punch-down terminal. That's the TIA standard.

T568A is the standard, you shouild use that unless you have a good reason not to.

(electrically it doesn't matter so long as you are consistant about it.)

AFAIK, both are exactly the same electrically, only the polarity of the wires (vis-a-vis color codes) is different. Were you to wire B to B or A to A it is straight through. Wire A to B then it is a cross-over.

Interesting... Actually as I think about this it may be possible, but I'd suspect that there was a hidden flaw in the cable that changing the wiring coincidentially fixed? Might that be possible?

Actually, not at all. Swapping green and orange swap whole pairs, not swap wires within each pair.

I think the difference is in the number of twists per foot in the green and orange pairs (orange has more). If the TPF match the blue or brown pairs, you get more crosstalk, so using the correct pairs may result in less crosstalk.

No question we have to do a full analysis of all factors!

I'm going to be doing a 200+ ft run in a month or so to the new garage, and will try both ways and see if one is more reliable than the other (I've got four cables run, only need one of them...)

Absolutely not. ALL pairs are ALL made EXACTLY the same.

No. Wiring it wrong gets you half duplex operation (sometimes) AND un-terminated lines exhibiting the stray signals that end up getting referred to as cross-talk.

Also how well the terminations are made matter too.

If you have more than say a half inch of untwisted wires exposed at the ends when you terminate, it will fail GbE tests, and you will get 100 base T or less performance.

Terminations should also be of the metallic shielded variety, as opposed to the old plastic plug and crimpers method.

I figured someone would say that, so I grabbed an orange pair and a green pair off my work table and compared them before I posted to make sure. I just re-checked, green is about 3.5 half-twists per inch, orange is about 4.5. It's pretty consistent along the 2.5-foot lengths I happened to have handy.

So no, all the pairs are not all exactly the same.

Only if you wire one end wrong. If you wire both ends the same wrong way, you don't get the same type of problems.

That page disagrees with you:

"One thing that I have noticed though is that in CAT5e, the orange and green pairs are twisted tighter than the blue and brown pairs. So do not expect to get the CAT5e quality on the network drop using the split pairs (brown and blue)."

In my case, blue 3.8 htpi, brown 2.8 htpi.

I guess that's to reduce cross-talk between pairs, at the expense of lower immunity to external interference.

As far as I recall, without sacrificing any long chunks of cable to check, which pairs are twisted more depends where you are in the cable - they use "variable twist rate" (in at least some cat5e). A two-foot sample would not show this.

Happily getting 1000 baseT with no errors from "old fashioned plastic".

--
Cats, coffee, chocolate...vices to live by

You might want to pick up a set of ethernet surge suppressors. In any case make sure that the grounding between the two buildings is good.

--
Cats, coffee, chocolate...vices to live by

They get tighter as you move up, away from the blue/white pair. On a

25 pair CAT 5 arrangement, the last pair, voilet/grey, is twisted so tight it's hard to unravel. At the frequency most establishments run at, the impedance difference between the O/W and G/W doesn't add up to a hill of beans.

I'm not into the arguments used in favour of twisted pair cable. The only reason they are getting away with twisted pair at 100 Mhz is through liberal use of bs. For one, the signals are digital and even a barely legible digital signal can be picked out of background noise with a Schmidt trigger. Try connecting a high-frequency analog signal through twisted pair and see how far you get.

Another matter is the claimed throughput as opposed to the actual throughput. Most telecom signals are regulated to 30 Mhz to prevent broadcasting of signals to adjacent equipment. That means the 100 MHZ claimed for CAT 5 regular is never used at that frequency. It could be, theoretically, but it never is because signals are multiplexed to get that throughput while running at a much slower frequency. A good example of that is the DSL signals sent down a normal telephone twisted pair which is rated at about 10 Mhz on a good day. DSL is accomplished with quadrature modulation, which piggy-backs signals on top of each other.

Talking about 1 Ghz twisted pair is a serious joke. They get that by using all 4 pairs on the cable, plus multiplexing. There's simply no way that twisted pair will ever catch up with coaxial cable and you simply cannot use a twisted pair line at 1 Ghz. The big push on twisted pair is due to how much more easily it can be installed than coax. It makes far more sense to install twisted pair in a hub arrangement than it does coax. It's far more economical.

Now we're seeing a push towards SATA over PATA. Although a hard drive is a serial device, and a PATA signal has to be serialized to write to the hard drive, I don't see what's being accomplished by converting to SATA. Again, the only real advantage is a skinnier cable and the ability to hot-plug the units.

It just makes no sense to push data transfer one bit at a time when you can do it 32 or 64 bits at a time in parallel. Then again, I wont be making the kind of money Intel will by cornering the market with unnecessary SATA technology. We should remember what happened to IBM and OS2 when they tried to foist a technology on a public that did not want it.