From what I remember, the blue pair has the highest twist and is used for the most data intensive communication. The brown pair have the least twist. This is what I've seen in all of the CAT5 cables I've run across and I'm pretty sure I read that in the specs for 568.

CJ

The difference between 568a and 568b is the are pins 1&2 and pins 3&6 are switched(in other words the green and orange pairs are switched).

A crossover cable is a cable that has 568a termination on one end and 568b termination on the other end. These cables are used to connect two computers directly together without a switch or router.

As far as the twists in a Cat5e cable the, they are there for data transmission speed. Each pair has a different number of twists (tighter/looser twists). Then the pairs together are twisted within the shielding. I am not sure if utilizing a particular pair for balun use makes a difference or not but, with data you want as little loss of twists as possible to maintain data tranmission speed. Ernie you said it perfectly!

Cat6 cable has the same number of pairs but the pairs are twisted tighter creating a faster data transmission speed.

Ernie to answer your question we were using 4XEM

No - the green and orange pairs are swapped. They are the only wires used in ethernet. The different twist rates are intentional and are to reduce near end (NeXT)and far end crosstalk (FeXT) between the pairs. Analogous to running LV wire parallel to hi volt wire. The different rates of twist will reduce bleed to the other pairs.

FYI If you terminate one end type A and the other type B you have a crossover cable.

Type B is typical for networking in this day and age.

EDITED: I should have read the entire thread before responding. Sorry for the repeat info. Ernie, I really thought you were pulling a few legs? Are you?

I think its blue, orange, green, then brown. Certain on the blue and brown.

For 10baseT and 100baseT yes, 1000baseT uses all four pairs.

this thread kinds of reminds me of what i will very often tell people when they ask me about how long ive been doing this or what schooling i needed.
Its is not very difficult to learn the hook ups of what we do "A goes to B ,B goes to C" ETC. but what makes the difference is knowing why A goes to B or more importantly "how" A goes to B" i really feel we as an industry dont spend enough time with the nuts and bolts of signal types i would wager that fairly good portion of the posters on this site dont know the difference between RGB and component video. now dont get me wrong it will not come up in most situations but you never know when it will be a day saver.

Proggieus, I can't agree more. It has never ceased to amaze me how many people in this industry never dig past the surface of what they are doing. Sometimes "why?" is a very important question.

Back on topic, this would be what introduces the concept of skew when using standard CAT5, CAT6, etc. in AV applications and why there is UTP that is specifically designed for AV use, such AV UTP by Extron, and Nano-skew by Belden. It is basically CAT5e, but with a standardization of the twist rate between pairs.

It also says right on it, "not recommended for data applications."

I've also kind of thought that that defeated the purpose of AV UTP distribution in the first place, beyond distance issues. The "promise" when the stuff first came out was that you could use one cable for everything or just spec a bunch of CAT5e runs to the tel/data contractor on a union job and just come in and put the widgets on each end, or as much as the AV manufacturers say (to quote George Feldstein at his last major speaking engagement) "I would prefer that AV dealers become IT dealers rather than the other way around." and to some extent believe it, they don't love the idea of lowering the barrier to entry by making all the cable types and terminations the same and using a standard that is the basis of a similar and much larger industry. The more UTP based products become available, the more IT, tel/data companies start entering the AV world, the more UTP based products get sold and so on and so on.

Maybe this is good for the industry, maybe it isn't, but knowing that there is still a difference between what you want in an AV UTP Cable and what you want in a data UTP cable is a great example of the nuances we deal with on a daily basis.

To answer the question more specifically, to the best of my knowledge there is a standard for minimum twist rate, but none for the differences in twist rates between pairs or which pairs where more tightly twisted than other. In the early days of AV UTP transmission, you'd buy a preterminated cable of a certain length and it would come with 5 coax cables, the length of which varied to compensate for the skew in the UTP cable, or you could deal with it in the field. Not exactly the cleanest solution. Today there are active devices that can provide the necessary delay compensation in a box, either standalone or integrated with the receiver. If it was a standard, then the skew would be consistent and predictable by length of cable and you would have to adjust each video component individually, you'd just set it by length.

I wasn't kidding. I have never looked into connecting two computers diretly together, so I have never run across this concept with CAT5. wow, my reputation precedes me to the point that if I say something stupid, people think it's a joke. What a great safety net for the ego! I must cultivate that....

Most of my hands-on computer stuff is done by one of my sons, who used to be in charge of hardware for the administration and staff (not students) of a local college. Now he's in charge of software. So I probably asked him years ago how to directly connect two computers together, he NEVER wanted to accidentally run across a crossover cable and have it muck up his day, so he told me not to bother with it. That's probably what happened.

I describe my approach to this stuff as learning WAY more about these things than I need to solve the problem at hand. This always helps me to make decisions when things don't work right and I literally have to design a solution. For instance, I was on the phone with a tech at Zinwell the other day asking how many mA it takes to run their switch, and LNBs. If I ever need to have an idea how wire length and resistance affect the voltage AT the LNB, this info will help. But it won't be in the manual and it won't find you. You have to seek it out.


Thanks. Only thing is, I have no idea wha 4XEM is. It could be another Star Wars android, for all I know. I'll go look it up.

That is exactly what I have observed. I can't recall the specific brands, most likey ProConnect and Belden, but when terminating on 110 blocks, I know that the twist rates from pair to pair from one manufacturer were different that the others.

This would seem to imply, if the EMI immunity is related to the twist rate, that terminating 568A vs. 568B when using baluns would affect performance, depending on the cable manufacturers' twist scheme.

Oh...I think that was mentioned in this thread earlier.

As "other" says above, that variation in twist from one pair to another is called "skew".

CAT-6 has a different rate than CAT-5E (Big E)...
which is different than CAT-5e (little e)...
which is different than CAT-5 (no E)...
which is different than CAT-3!!

Skew is NEEDED for data runs for the crosstalk reduction reasons outlined in this thread.
Reducing the crosstalk allows for higher usable bandwidth, because less of the cable's capacity is cluttered with errant information from the other pair(s) or from outside.

The skew becomes even more critical as we move forward with Cat cables as a carrier of HDMI.
Twist rate and cable compisition make a BIG difference in high speed distance.
Example:
Cat/5 (2ea) - 100' @1080P
Cat/5E (2ea) - 125' @ 1080P
Cat/6 (2ea) - 200' @ 1080P

(hidden message: install Cat/6)

BTW:
Rick, did you know that AMX is an HDMI adopter/lincensee?
Hmmmmmm...........
(Note-to be Fair: so is Crestron)

Gents, (And the lady or two who are sometimes present.)

As many have noted, the twist rate is different on each pair, and the reason is to provide rejection from other pairs in the bundle.

There is no spec, on what pairs have a specific twist rate. A minimum twist rate is implied by the noise immunity required to meet the specification of the entire cable. But the rates on the individual pairs is manufacturer specific, and not dictated by the networking specifications. The networking specifications don't specify how the wire manufacturers perform the desired specs, they only specify what requirements are in place... dB of rejection on adjacent pairs, dB of rejection of adjacent cable, etc.

Note that using a pair with a higher twist rate will give a higher noise immunity, but this does not necessarily result in a better picture when using balun's. This is because the higher twist rate also results in a higher capacitance for that pair. This can be modeled as each strand of the cable being a series string of low valued resistors, and the node between each resistor having a capacitor connected from one conductor to the other. (Looks like a ladder, with the sides being resistors, and the rungs being capacitors.)

This model is a low pass filter! Above the corner frequency of the RC filter formed by the resistance of the conductor, and the capacitance of the twist, the cable looks like a short to high frequencies. So this twist reduces susceptibility to noise, but does so at the expense of bandwidth. These cables carefully balance the resistance, and capacitance in an effort to provide noise immunity while preserving as much bandwidth as possible, while keeping manufacturing cost reasonable. Like many things in life, this is a balance, and the best overall result is achieved by carefully balancing the behaviour, weighting the factors for what is typically seen in real applications.

So, those who propose that the best performance will be gained with balun's when using the pair with the highest twist rate:

They'd be right if there is lots of noise in the cables environment, and trading off some bandwidth for noise results in a better picture.

They'd be wrong if bandwidth is more important, because in the environment of the cable, noise is not very high.

Note that these cables are designed to be placed in a cable chase, often with 100's of like cables! The way we use these cables in a typical residential installation, they see very little noise compared to the noise seen in a typical commercial environment.

My guess is that better performance would be achieved when using balun's, by using the pairs with the lowest twist rate. But I also think this becomes a almost nonsense point, as with composite you're not going to see the need for bandwidth, and with component, you're going to use 3 out of 4 pair anyway, so you're only dodging the pair with the highest twist rate.

I did quickly try this with my component balun's, and a 1080i picture. I couldn't see the difference with CAT5e cable, between using the lowest 3 twist rates, and the highest 3. Which pretty much fits what I expected... but I did the experiment, because I wanted to make sure that what I expected scientifically fit the real world. Otherwise, I'd have to investigate why... and that would have likely resulted in a new product or technique!

bcf,
Excellent points, all.

Just an addendum -- althought the higher twist rate does cut off higher frequencies more than the lower twist rate, the high frequency performance of the mst tightly twisted pair in any Category Cable will match its spec. As written, it sort of sounds like the most twisty one maybe should be avoided because it's a low pass. Wll, it's a low pass, but it still performs well with video, as shown by your real-life test.

By the way, someone else said skew is the name for the twist. I thought it was the name for the change in the signal from one end of the cable to the other, which is of course directly related to the twist, but I thought it was a signal characteristic, not a twisting characteristic. Comments?

In electrical engineering, when talking of skew, we are always talking about the difference between a clock transition, and the ideal clock transition. Wikipedia says it well at: [Link: en.wikipedia.org]

So... I would agree that skew is a signal characteristic.

But, I thought it could be covered by general useage of the work skew. Looking up skew in an online dictionary I get: [Link: thefreedictionary.com]

None of the definitions really talks about a difference in twist. The closest is talking about "An oblique or slanting movement, position, or direction." I don't really think that seems correct for how skew was being used.

I don't think skew properly relates to the change in the twist rate in cabling.

Skew - Refers to the timing difference which occurs when electrical signals which are traveling over different pairs of cables reach their destinations at different times. The different arrival times of the signals may present a problem when simultaneous arrival with no delay is required.

Skew-Free - A reference to special twisted pair cable in which the length differences between cables reduced to a minimum, thus reducing cable skew

As with the majority of words in the english language there are many definitions in various contexts. In the context of AV over UTP, that is the definition straight from a large and very well-respected manufacturer and educational resource in this industry.

The length differences are caused by the variation in twist rate of multiple pairs of cables in the same jacket, so yes you are correct that the term "skew" does not refer exactly to the variation in twist rate, it is caused by it.

In reference to "skew-free" or "Nano-skew" if you prefer Belden's terminology, which is probably more accurate if we are being 100% literal, it also does not refer to the cable with a lack of varying twist rates, that's just how it's achieved.

I didn't find it important to correct that as in this context the term use is somewhat interchangeable. I'd much rather have someone with a slightly incorrect definition but know what they are looking at when the problem arrises and be able to solve the problem in the field then not now what the heck they are looking at when it happens, as was just clearly illustrated on another thread that just popped up, or better yet to save someone from an expensive mistake by getting them thinking about this at the design stage rather than the trim/finish phase.