Why do we have cross-over cables.

[alchazz]

..

For modern ethernet, you don't need cross-over cables. The devices
figure out the pair sorting. All the NICs, routers, and switches I've
come across in the past 3 to 5 years have auto-sorting of the pairs.

When you connect two computers to each other, you use the crossover cable
and you don't need a router.

Al

 

[whit3rd]

Maybe I've missed something, but it's always seemed to me that a logical
approach would be to define some pins as input and some as output, and
for cables to connect input pins to output pins,

That's how the Macintosh serial ports and printers/modems etc. were
wired. It works.

The old RS-232 days were different: it made sense to the original
builders that a data communication device would attach onto a
data terminal device, SO they were specified with mating connectors.
The DCE (data communication equipment :== modem) had a female
plug, the DTE (data terminal equipment :== terminal, computer) had
a male plug. The connection was through simple extension cables,
also convenient for insulation-displacement (ribbon cable) and
for mass production.

I think, historically, the decision to use male and female
connectors (which allows docking modules without a cable)
was the breakpoint.

In the case of UTP wiring, the two wiring schemes are 568A and
568B (and everyone uses 568B, the "crossover cable" is 568A on one
end). Computer-to-hub/router/switch uses straight-through cable,
computer-to-computer uses crossover cable, hub/router/switch to
hub/router/switch uses crossover cable (or a "special" crossover
port).
This is becoming irrelevant, though, with new chipsets: all the
gigabit Ethernet I've seen has logic that corrects the signal for
wire arrangement (the gigabit protocol requires bidirectional use
of all the wire pairs anyhow, so send/receive is just a software
setting).

 

[StickThatInYourPipeAndSmokeIt]

OK, I can accept the field installation error issue, though the
existence of two different standard ways of wiring the plugs seems at
least partly to defeat the goal of reducing errors.

I'd have thought that in a mass-production environment it would be as
simple as having one person do one end and another person do the other end.

Sylvia.

You still do not understand. When these standards were made, and when
such production began "mass production" was individual hand assembly, and
regardless of who does which end (passing it off to another wastes yet
more time), other than pin for pin wiring can and does increase prime
pass yield failure rates.

If the wires change, the error rate goes up. If ALL assemblers wire
ALL cables the same on ALL ends, the likelihood for error decreases
exponentially, because they all learn, and inspect the same pin-out every
time. Errors get made less often, and inspection error get made less
often. With the advent of cable test fixtures, the only place left for
the errors to occur is at the individual cable builder's bench. That
builder's learned mindset, if hard wired to one spec, becomes less prone
to error.

 

[StickThatInYourPipeAndSmokeIt]

TROLLTROLLTROLL
has to be ?
no one could be that fucking stupid

She pops off with a stupid question every couple days. Considering the
level of off topic posts here, it actually causes discussion of actual
electronic issues.

Still, it belongs over in the basic group, as does she... well...
actually, she belongs in the kook group.

 

[FatBytestard]

Rod Speed of old?

If so, he was a goddamned idiot with that moniker as well.

 

[FatBytestard]

Early modems ware loosely speaking analogue filters coupled to serial
line drivers, if you were lucky you could pulse dial by toggling the
DTR line with the correct cadence...

No luck about it. Make-and-break dialing STILL works.

 

[FunkyPunk FieldEffectTrollsistor]

Maybe she is the result of a union between woddles and philthy ?

of course the question is who was the bottom ?

Have you seen her photo? The only sex that ugly ditz gets is from her
fingers. She should do x rated. They'll shoot film of anybody doing
anything.

 

[Sylvia Else]

You still do not understand. When these standards were made, and when
such production began "mass production" was individual hand assembly, and
regardless of who does which end (passing it off to another wastes yet
more time), other than pin for pin wiring can and does increase prime
pass yield failure rates.

Wire coloured x goes to pin y. You just need two groups of people,
who've learned different schemes.

Sylvia.

 

[Capt. Cave Man]

Don't count on it. We found a bunch of 'em that don't work as
advertised. :-(

Best to actually use a known good, proper cable, and not rely on a nic
or router to be your cable tester, or repair facility.

Now tell me that one was wrong too.

 

[FatBytestard]

.

When you connect two computers to each other, you use the crossover cable
and you don't need a router.

Al

And due to your quick scan of the ONE post it appears you read, you
completely missed the point of the entire thread, much less the one post
you responded to.

 

[FatBytestard]

You missed something, and there's a good reason why this condition
persists.

In a minimalist DB25 connection between DTE and DCE we have this:

DTE DCE
DB25M CABLE DB25F
+---------+ +---------+
| TXD 3>>3--------------------3>>3 RXD |
| RXD 2>>2--------------------2>>2 TXD DCE |
| SIG GND 7>>7--------------------7>>7 SIG GND |
+---------+ \ / +---------+
DB25F DB25M

Notice that the cable conductors do not cross and are wired 1:1 from end
to end.

But now assume someone wants two computers to talk to each other through
their serial posts, and we wind up with this:

DTE DTE
DB25M DB25M
+---------+ +---------+
| TXD 3> <3 TXD |
| RXD 2> <2 RXD |
| SIG GND 7> <7 SIG GND |
+---------+ +---------+

How would you suggest wiring up the ports to accomplish that?

JF

The port would have to have an expensive multi pin switch incorporated
into the panel next to it. Very bad.

 

[Sylvia Else]

You missed something, and there's a good reason why this condition
persists.

In a minimalist DB25 connection between DTE and DCE we have this:

DTE DCE
DB25M CABLE DB25F
+---------+ +---------+
| TXD 3>>3--------------------3>>3 RXD |
| RXD 2>>2--------------------2>>2 TXD DCE |
| SIG GND 7>>7--------------------7>>7 SIG GND |
+---------+ \ / +---------+
DB25F DB25M

Notice that the cable conductors do not cross and are wired 1:1 from end
to end.

Yes, clearly with the endpoints wired as they are we need different
cables depending on the kind of endpoints we're connecting. But that
wasn't my quesion.

My question related to why this situation was created in the first
place, since it is not a logical necessity.

Sylvia.

 

[krw]

Best to actually use a known good, proper cable, and not rely on a nic
or router to be your cable tester, or repair facility.

Now tell me that one was wrong too.

You are. If you want to test cables, get a cable tester. If you just
want to hook things up, try it. You might get lucky. You'll always
get it wrong, in any case.

 

[Sylvia Else]

For modern ethernet, you don't need cross-over cables. The devices
figure out the pair sorting. All the NICs, routers, and switches I've
come across in the past 3 to 5 years have auto-sorting of the pairs.

Is there a standard specifying how this works?

I'd have to wonder how two such interfaces manage when they are
connected together.

Sylvia.

 

[StickThatInYourPipeAndSmokeIt]

Wire coloured x goes to pin y. You just need two groups of people,
who've learned different schemes.

Sylvia.

You have no concept of hand assembly processes.

One person wire both ends. That ONE person only needs to examine the
wiring of that one cable, one cable at a time.

Gathering up a bunch of cables to pass them off to yet another
assembler has an attached cost that reduces profit, and cable fab shops
pinch every penny. That's why they hire unskilled workers.

Then there is the error rate thing, which invariably also goes up.

When a single lost cable occurs, it costs the company more than twice
the value of the lost labor. Usually about 5 times when the error is
caught in-house. It can be 16 times the cost if caught in the field.

Rework is a very costly aspect of lean electronics manufacture. If
error rates are high, it quickly becomes the most costly expenditure.

Yes, cables are simple. But that just means that the assemblers *think*
they can be more casual. Otherwise there would be ZERO failures for a
simple 9 pin interconnect. That is only 18 connections, after all, and
that is if all pins were used. Yet still, we have much error.

The cables that have historically shown to be the least error prone
assemblies are, in fact, pin-for-pin matings.

 

[StickThatInYourPipeAndSmokeIt]

Yes, clearly with the endpoints wired as they are we need different
cables depending on the kind of endpoints we're connecting. But that
wasn't my quesion.

My question related to why this situation was created in the first
place, since it is not a logical necessity.

Sylvia.

He already showed how it WAS and still IS a necessity.

 

[Rich Grise]

My question related to why this situation was created in the first
place, since it is not a logical necessity.

Because the modem is "Data Communications Equipment", or DCE, and the
computer/teletype is "Data Terminal Equipment", or DTE. When you use
a computer to simulate a modem (or a server on the other end of the
line) you need to turn the sense of the connector around. This is
why reversing cables used to be called "null modem" cables.

Hope This Helps!
Rich

 

[Sylvia Else]

Because the modem is "Data Communications Equipment", or DCE, and the
computer/teletype is "Data Terminal Equipment", or DTE. When you use
a computer to simulate a modem (or a server on the other end of the
line) you need to turn the sense of the connector around. This is
why reversing cables used to be called "null modem" cables.

Hope This Helps!
Rich

No it doesn't. It entirely misses the point, which is why there is a
distinction betwen DCE and DTE.

Sylvia.

 

[Capt. Cave Man]

You are. If you want to test cables, get a cable tester. If you just
want to hook things up, try it. You might get lucky. You'll always
get it wrong, in any case.

This krw idiot, folks... is a troll. In this case, however, you were
baited into it.

I was, in fact, correct. Note how it was me that stated that a nic is
NOT a cable tester. You are just too retarded to have seen it in your
glaring need to troll me and the group. **** off, KeithTard.

 

[FatBytestard]

Is there a standard specifying how this works?

I'd have to wonder how two such interfaces manage when they are
connected together.

Sylvia.

If a handshake fails, a different connection schema is used (read
switched in) until success is achieved. Pretty simple.