cat 5 cable

[ban-all-sheds]

1) Strictly speaking, "Ethernet" is a proprietary Xerox term. IEE802.3 is what you should be using. But nobody does.

2) Twisted pair cable was originally defined for 100Mb 802.3

3) 100BASE-TX does still use CSMA/CD

 

[NotSureAboutDIY]

1) Strictly speaking, "Ethernet" is a proprietary Xerox term. IEE802.3 is what you should be using. But nobody does.

2) Twisted pair cable was originally defined for 100Mb 802.3

3) 100BASE-TX does still use CSMA/CD

I thought telephone systems had been using twisted pair cables since the dinosaurs time do you mean unsheilded twisted pairs, either way thats news to me.

if you're going to say 100base-tx uses csma/cd then surely so does 10base-t and 1000base-t? we only really got shot of collisions and the need for csma/cd (as a protocol and interface) due to microsegmentation via mac address switching didnt we?


I'm interested to know if utp is an effective transmission media for television from the aerial down?

I've seen utp used in video/cctv distribution already but I was assuming that from the aerial (UHF) was too high frequency?

anyone with experience let me know, I've got to re sight my aerial off my roof in the loft and as Im digital any minor improvements I can scrape will probably be required...

 

[Softus]

its not so much that the second wire has to be an inverse (in ethernet i belive the signal lines are are actually completely isolated from the system power rails so they probablly do end up as an inverse at least for the high frequency components) as that the receiver discards all common mode changes and only looks at the difference between the two wires.

and yes both ethernet and twisted pair predate ethernet over twisted pair though i belive cat5 was more or less designed specifically for fast ethernet.

Can we start getting the terminolgy right? Ethernet is not a word that, on its own, defines a signal set or physical medium.

When you refer to Ethernet over UTP you're talking about 10baseT, whereas Ethernet over multidrop coaxial cable is 10base2 or 10base5.

Cat 5 cable evolved from cables with a lower bit rate capability, e.g. Cat 3.

I don't know what "system" power rails you're referring to, but you're right that UTP carries no power. In terms of being isolated, this is the interferece problem that twisted pairs are designed to help solve. So, turning back to the inverse signal question, I note that you use the word probably, so I surmise that you're guessing.

Does anyone know the answer to this? Otherwise I'm gonna have to look it up.

 

[Pensdown]

As far as I remember structured cabling was introduced in the UK in the late 80's using CW1700 (CAT3ish) which at that time was a high grade voice cable. The concept was to wire a building with a cable which most voice & data services could use. BT had one of the first UTP systems called OSCA which could run Token Ring, RS232 Ethernet Asynchronous etc.

UTP can carry power if required (some IP voice systems do just that) and the reason that 4 pairs were chosen, (most data services only require 2 pairs) was for that reason. UTP is all about balanced pairs which is why CAT6 CAT6a has a center core which the pairs are bonded to. In simplistic terms this ensures that the twist remains equal over the full length of the cable allowing speeds up to 10gig@100mtrs.

There is absolutely no connection between any type of transmission service/protocol and CAT** cable and there never has been.

 

[Softus]

Certainly a factual and interesting post Pensdown, some of which I didn't know, but what about the "inverse signal" postulate?

 

[Pensdown]

Sorry, I can't help. I know a man who should know but he's out on the lash while I'm working in my hotel room.

 

[plugwash]

I don't know what "system" power rails you're referring to, but you're right that UTP carries no power. In terms of being isolated, this is the interferece problem that twisted pairs are designed to help solve. So, turning back to the inverse signal question, I note that you use the word probably, so I surmise that you're guessing.

Its transformer isolated at both ends (and yes i have looked at schematics of ethernet interfaces) so all that is strongly controled is the differential mode signal, what levels either of those wires are at relative to any external reference is not well defined.

 

[Softus]

...not well defined.

Well, yes, fine - but I (for one) wasn't asking for them to be well defined relative to an external reference, merely for the signals within one pair to be defined relative to each other.

Do you know if the second wire signal is the inverse of the first? Does anyone else know?

 

[plugwash]

Do you know if the second wire is the inverse of the other? Does anyone else know?

there is a voltage between them that varies, with two wires in isolation that is all that can be defined.

 

[Softus]

Do you know if the second wire is the inverse of the other? Does anyone else know?

there is a voltage between them that varies.

Eh? Surely the answer to "Do you know" can nothing other than "yes" or "no" or "I don't know".

And if it's a "yes", then is signal 2 the inverse of signal 1 (another "yes" possibility) or not (another "no" possibility)?

 

[plugwash]

(note, i've moved this to a new post as it was actually written in edits the first one of which was made after softus posted but before i realised he had done so)

to put it another way from the perspective of a receiver that only looks at the difference between two signals and doesn't compare them to any external reference there is no difference whether one line stays static and the other moves or both move provided that the difference changes as expected.

to put it in terms more familiar to the average sparky, suppose you take a isolation transformer and connect a device to its output. It makes no difference to the device if you leave the output floating, tie one end of it to ground or tie its center tap to ground.

now continueing with that analogy, with one end of the output tied down you would say that the other end is moving, with the middle tied down you would say that both are moving as the inverse of each other but with nothing tied down all you can say (without getting into the realm of electrostatics, parasitic capacatances and the like) is that they are moving relative to each other.

 

[Softus]

cat5 requires the inverse of the signal to be carried in the second wire in the pair.

I deduce from your answer that we're using different definitions of the words "Do", "you", "know", "if", and "inverse".

 

[ban-all-sheds]

Can we start getting the terminolgy right? Ethernet is not a word that, on its own, defines a signal set or physical medium.

[pedant or Xerox lawyer]Yes it is [/pedant or Xerox lawyer]

When you refer to Ethernet over UTP you're talking about 10baseT,

Or 100BASE-TX or 1000BASE-T or 10GBASE-T.

Cat 5 cable evolved from cables with a lower bit rate capability, e.g. Cat 3.

Yes, it was extraordinarily sloppy of me to say "Twisted pair cable was originally defined for 100Mb 802.3". I meant Cat5.

I don't know what "system" power rails you're referring to, but you're right that UTP carries no power.

http://en.wikipedia.org/wiki/IEEE_802.3af

In terms of being isolated, this is the interferece problem that twisted pairs are designed to help solve. So, turning back to the inverse signal question, I note that you use the word probably, so I surmise that you're guessing.

Does anyone know the answer to this? Otherwise I'm gonna have to look it up.

It's a differential signal. Whether one core carries the inverse of another is a largely philosophical argument.

 

[Softus]

Ethernet is not a word that, on its own, defines a signal set or physical medium.

[pedant or Xerox lawyer]Yes it is [/pedant or Xerox lawyer]

You're right, of course, but in this context an insistance on using the original definition of the word makes quite a lot of the discussion more confusing that it already is, so I implied, but should have explicitly written, in contemporary common parlance Ethernet is not a word that <bla bla bla>.

When you refer to Ethernet over UTP you're talking about 10baseT,

Or 100BASE-TX or 1000BASE-T or 10GBASE-T.

Quite so. I stand corrected.

I don't know what "system" power rails you're referring to, but you're right that UTP carries no power.

http://en.wikipedia.org/wiki/IEEE_802.3af[/QUOTE]
I don't believe that PoE techniques are what plugwash was referring to when he wrote "system power rails". I have a feeling that he was talking b*llocks.

Does anyone know the answer to this? Otherwise I'm gonna have to look it up.

It's a differential signal. Whether one core carries the inverse of another is a largely philosophical argument.

I must be being thick because I can't see the philosophical argument here.
Differential is term I'm quite happy with, and I agree that it describes the signal carried by each pair in UTP cable, but inverse? I can see a reason for doing it, being to increase the signal to noise ratio on the wire, but I didn't know that this was a technique used with 10xyBASE-Tz.

And I still don't know...

 

[NotSureAboutDIY]

Can we start getting the terminolgy right? Ethernet is not a word that, on its own, defines a signal set or physical medium.

[pedant or Xerox lawyer]Yes it is [/pedant or Xerox lawyer]

well if we're in pedant territory and we are if we're on an internet forum

wasnt it xerox, dec and intel that combined effort to put together the first open ethernet standard (thats not to say xerox didnt have lans based on similar technology first)
to add to the pedantry IEEE802.3 is virtually identical and was really just to put the standard in the hands of a standards body.