CAT5 Cable Testers = Accurate???

[MulLa]

Hi all,

I've tested a few, well two to be exact, CAT 5 straight cables with a simple CAT 5 cable tester. The tester have two rows of eight lights and they will light up in sync if the 8 wires are in contact. There's also an addition Ground light which only one lights up at a time. Even for cables that are proven to be working.

Now I've just made those cables and tested it with the tester. All the lights go up in sync and things seemed to be fine.

Then I started using those cables and they don't seem to work at all. Tried using a different cable that I know has worked and they worked fine so it's not the config of the systems that's causing the problem.

Just wondering if there's any explaination to this or are cable testers just not that accurate.

 

[spidey07]

Those testers only test continuity and nothing else, meaning they can't tell you if a network cable is good or not.

Are the wires inserted in the correct order on the RJ45 plugs following either 568a or b standards?

There are other testers out there that can accurately test a cable and certify it. They start at a few thousand dollars - fluke is a good place to start.

 

[MulLa]

Thanks Spidey07 for your response.

What are the 568a / b standards? Actually someone told me that as long as the wires on both ends match it would work perfectly as a CAT5 straight cable? Would you care to enlighten me on the 568a / b standards?

What I don't understand is what's the difference between continunity and a cable being good or bad? If it's bad the obviously no signals can get through thus no continunity? Sorry, don't know that much about cables

 

[Tallgeese]

Originally posted by: MulLa
What are the 568a / b standards? Actually someone told me that as long as the wires on both ends match it would work perfectly as a CAT5 straight cable? Would you care to enlighten me on the 568a / b standards? What I don't understand is what's the difference between continunity and a cable being good or bad? If it's bad the obviously no signals can get through thus no continunity?

EIA/TIA 568A/B outline specific wire orders, which is key to maintaining signal integrity.

The misinformation you received is probably the most common problem at the physical layer in networks. Unfortunately, many "professionally" trained IT people have the same misunderstanding. I've had new employees who were self-proclaimed "cable wizards" who insisted that pair order did not matter.

The reason it matters: Ethernet uses two pairs... position 1 & 2, and position 3 & 6. Because of the physical characteristics of unshielded twisted pair wire, if these two position pairs don't share a respective color, then the cable has a physical error known as a "split-pair," and likely won't work, especially at higher speeds.

Here is the pair order for 568B

1 - white/orange
2 - orange
3 - white/green
4 - blue
5 - white/blue
6 - green
7 - white/brown
8 - brown

568A is the same, except that the orange and green color pairs swap positions.

A straight-through cable is one that is either 568A or 568B on both ends. A crossover cable is 568A on one end, and 568B on the other end.

 

[Tallgeese]

BTW: a higher end cable tester would identify this error in a cable, along with a RAFT of other errors that a continuity tester won't.

 

[ScottMac]

Theres more to it than just getting a signal through. The cables are engineered (twists-per-pair, insulators, jacketing) to produce a physical medium that can support high speed digital signalling. They were engineered to the EIA/TIA 568A spec, which include a specific pair-order ("the 568a"), which was later amended ("The 568b").

If you make your cable such that each color pair are together, you end up with a "split pair" on pins 3&6 - you have a pair of wires where one wire is in one twisted pair, and the other wire is in another - so you lose most / all of the noise suppression, and increase the attenuation (loss) and absolutely kill the ability of the cable to reject crosstalk (from pair to pair).

The correct/acceptable pair order for 568a is:

White-green, green, orange-white, blue, blue-white, orange, white-brown, brown

For 568b:

White-orange, orange, white-green, blue, white-blue, green, white-brown, brown

You'll notice the difference is that the green and orange pair have been swapped. An Ethernet crossover cable could be described as bing 568A on one end, and 568b on the other.

As Spidey mentioned, there are inexpensive test units that will tell you that you have end-to-end continuity and polarity, there are more expensive units that will tell you how long the cable is (TFR function), and there some REALLY expensive units that'll tell you if the cable complies with (and is working at) Cat 5, 5e, or 6 specification. The really expensive units (US$3,500.00 and up ) actually send signals at frequency down the cable and measure the crosstalk, (near end / far end), ACR (Attenuation to crosstalk ration), attenuation - the whole banana. With a few situational exceptions, if those scanners tell you the cable is working at certification spec, you know the cable is good enough to sign a legal document saying so.

The other thing to mention here is that every component (cable, panels, inserts, and RJ45 connectors) ALL need to be of the same level spec, you the entire span takes on the level of the lowest component (if you use cheap, unrated RJ45 connectors, you have an unrated cable.

The teltale symptom of a split pair / bad / improper cable is that it may work OK at 10Mbps, but is "slow" at 100Mbps.

That's it in a nutshell, lots of other details, but I gotta run to work.

FWIW

Scott

Edit: 'Geese's flying fingers wins the race, once again!

 

[JackMDS]

A little perspective on Network Cabling.

This is a simplified (to suit all customer version); "Plumbers" please do not get upset.

In most cases of need to transfer electrical information, we use wires that are comprised of copper.

Copper is chosen, because it is inexpensive, mechanically flexible, and a good conductor.

Electrical transfer of electricity is made of rapid transfer of electrons from one molecule of the conductor to the other.

This transfer is generating heath, and electrical field around it. It is also susceptible to be influenced by outside electrical fields (commonly called noise).

The higher the frequency (speed) of the electrical transmission the more it creates noise, or can be affected by outside noise.

In order to suppress the phenomenon the coax cable was invented. The coax cable is build with inner copper conductor that is surrounded by ?plastic? tube, and the plastic tube has a Screen like copper sleeve around it. This copper sleeve is grounded on both sides of the cable, thus creating an electronic shield around the central conductor that carries the electrical information. It blocks sending interference out, and protected the central conductor from out side interference. In the early day of Networking coax cables were used to connect the NICs.

However, coax is expensive and hard to work with, hence comes the CAT5.

The main idea of the Cat5 is twisted pairs of conductors, since a pair of conductor when twisted together will create electrical fileds that can cancel out the noise. (like pushing and pulling the same electrical signal).

The efficacy of the CAT5 cable is a result of the twisted pairs, thus it is very important to build the CAT5 cable correctly, other wise the noise cancellation is reduced, and the Network will stop functioning, or will function at a reduced frequency (speed). Testing, and certifying a cable actually done by transferring high frequency through the cable, and measure the outcome, this is the reason why DC conductance (using regular OHMmeter, or inexpensive CAT5 testers) is not the correct test to insure CAT5 cable performance.


The color scheme is to insure that you are using the Cable in a way that noise cancellation is maximized.

 

[MulLa]

Hi all

Thank you for all the valuable information from everyone. Will be sure to keep the correct cable order in mind next time I make one!!

 

[Rufus210]

Originally posted by: ScottMac
They were engineered to the EIA/TIA 568A spec, which include a specific pair-order ("the 568a"), which was later amended ("The 568b").

ScottMac: this is a question I've wondered about before, and you seem to be the perfect person to answer it: What is the difference between 568a and 568b other than the obvious switching of colors? Is there any particular reason as to why there are these two standards when functionally (as far as I can tell) they are identical? And would it make a difference if I switched lets say orange out with blue and green with brown hence not complying to any color spec, but still having pins 1&2, and 3&5 sharing a correct pair of wires?

 

[ScottMac]

I'm not sure why the 568a spec was amended to include the 568b, I'll ask my friends at the cabling Lab and see what they have to say about it and put up a follow-up, hopefully, this week.

Regarding the use of any color, even though the general scheme is followed:

It's my understanding that pair order, by (proper / spec) color is important, primarily because of the diverse role of structured cabling in general. Remember, the "big picture" reason that structured cabling was promoted: universal functionality over a single medium (copper in this case, fiber would be different, of course).

Category-rated UTP was designed to handle more than a broad scope of data networking protocols. It can also be used for serial applications, 5250, 3270, baseband (composite, S-Video) video, broadband (rf-based signalling), audio (this is a stretch), T1 premeses cable.... anything .... (usually) with the proper "adapter."

Each application can generally be found to use a specific subset of the four available pair in a single Category-rated UTP cable. Baseband video adapters, for example, generally will express the signals to pair three. T1 is put on pins 1&2 + 4&5, ATM/STM on pins 1&2 + 7&8, Ethernet on 1&2 + 3&6, etc.

The manufacturers of the cabling design the cable (generally speaking) such that each pair is twisted at a different rate; this helps to reduce pair-to-pair crosstalk, and offers enough variance to accommodate the various types of signalling that are likely to be applied to those specific pairs.

Since pairs 2 &3 (pins 3&6 + 1&2, respectively) are essentially interchangeble (568A or 568B), and those pairs are commonly used for Ethernet, I'd bet that the specifications for the Orange and Green pairs are also pretty much interchangeble. The blue pair (pins 4&5) are most commonly used for voice, T1, and Token-Ring .... the twist ratio and insulative material used on the blue pair is (many cases, not necessarily universal) different than that of the Orange and Green pair. The Brown pair (pair 4, pins 7&8) are most commonly used for 56K DDS, ATM/STM .... again, the twist ratio, and maybe the jacketing is likely to be different than that of the other three pairs.

The bottom line is, in theory, that each pair or combination of pairs are designed to handle a limited scope of the complete suite of signalling possibilities for that type of cable, within its class of certification (Cat 3, 4, 5, 5e, 6). By deviating from the standard (568A or 568b), you may be applying a signal to a pair (or pairs) that are not designed to accommodate that kind of signal, and you may experience sub-optimal performance.

Certainly, the actual design and capabilities of the cable (per pair or as a group) will vary from manufacturer to manufacturer. It may be possible to use any pair color combination for any function ... but WHY? It takes the same effort to do it "right," and the only way to guarantee proper performance is to follow the specification. Unless you have the Lab sweep/test stats for a given cable, you can't know how the cable will perform on a per-pair basis (fortunately, I can get access to most cables stats ).

I hope this is helpful, I'll try to get an answer on the 568A/B thing this week.

FWIW

Scott

 

[Rufus210]

thanks for the (as always) extremely informative reply. And don't worry, i always follow 586b for my installs, it was just one of those nagging "what if?"s that I've been wondering about for a while.