We took over an existing job and noticed that some cables were terminated 568 a and some b. Network was working fine that way. Mos. later, we can't get on internet. Cable modem is working perfect, once we plug network into modem, can't get on internet. Could it be the mixed a and b config. All cables test out fine. Don't think this could crash network all of a sudden.

If connected devices use a single pair of wires (Blue/White), mixed cabling won't manifest itself as that pair lands on the same positions with both conventions. Once you add more pairs, then the trouble begins. Either way, pick a convention and use it from end to end.

568B is most common in the US as it was developed by AT&T (Western Electric).

If terminated the same on both ends it won't matter.

"Cable modem is working perfect, once we plug network into modem, can't get on internet."

You will need to explain a few things, using industry-standard terms. First, I suspect that this not a wiring problem at all, since wires do not stay up late at night to conspire to drive us crazy. Unless a human has gotten his fingers into the situation, wires that worked in the past will work in the present. You are correct when you say "Don't think this could crash network all of a sudden."

That said, please explain what you mean by "cable modem is working perfect." How did you test it? What did you plug into it to determine that it is working "perfect."

Next, what do you mean by "we plug network into modem"? A network, which generally consists of many wires going to many locations, needs a router/switch to create the ports necessary. How did you plug a network into a modem? Have you by-passed the router (assuming there is one) and patched directly from the modem output jack to a single computer's input jack?

Finally, as the other guys have stated, you can use whatever color code you like, as long as the same code is used at both ends of a particular cable.

A (the old Merlin wiring standard) is awful, b is better, just to keep it straight as to which one to use the most.

We plugged laptop directly into modem, Time Warner cable did this, nothing else was plugged into modem. Was able to get onto internet.

Then came out of modem,into Araknis switch, came out of switch and into rest of network. We do Control 4 and Sonos Audio.

Have feeling, someone plugged a wrong device into an Ethernet port, causing this problem.

If it's just a switch, then you need a router, if attaching more than 1 cable, THEN use a switch for any extra ports.

EDIT: Forgot to mention, up here when we swap out or add a a router, we have to reboot the modem on a cable connection so that it sees it. Being on cable down there it might be the same as well.

B was the old Definity/Merlin standard.

-Hal

Time Warner cable modems are all in one. Modem, Router, Switch, and AP.

Does the network work with the modem unplugged? Meaning can you print to a network printer or such.

Not sure about printing. I bet network does not work. I will be on site this week

It was the 258a standard that became 568b my fault, got them confused with 568a.

If the system was working with the mixed wiring and then suddenly stopped working, it is obviously not the wiring.

You need to some basic network troubleshooting.

1. With everything disconnected from the switch but the switch connected to the router, connect your laptop to the switch and see if you get on the internet. If you do, then start adding connections to the switch ports until you add one and you lose internet connection. Then troubleshoot that cable and the device or devices attached to it.

2. Be aware, (and this is the bane of network topology), one device that is throwing garbage on the network can bring down the whole network and make all devices on that network inoperative. It's called denial of service or DOS and can affect VoIP networks as well. Use process of elimination to detect the problem.

Rcaman

And once you identify the cable, make sure you test everything attached to it for the entire length such as supplementary switches and the attached devices. May be a hidden switch in a closet, under a desk, etc.

Rcaman is on track with this one. We once had the cleaning guys come in and do the floors strip, wax, polish and etc... They got into computer lab, disconnected all pc's, did what they had to with the floor, here came interesting part, they took initiative to hook everything back up. Needless to say they had one patch cord plugged into two network jacks and that's when the fun has begun. Following Monday we sure were in for a surprise, I got woken up bright and early and did not appreciated too much when someone was screaming in panic that there was no internet in the whole building.

It's called a 'Broadcast Storm'. A Denial of Service attack comes from the outside.

A simple cable plugged in between two switches on a network causing a loop to happen will cause a broadcast storm and will crash your network. There are several other simple things that can cause it too.


You have to find out if the network is working without the modem plugged in first. Otherwise, you'll just be chasing your tail.

Technically, a broadcast storm effectively causes a denial of service. The point here, is, that troubleshooting the OP's problem will require breaking down the system into segments that can be tested and verified.

Rcaman

Ah, the old cleaning lady syndrome. In my book, I promise to devote an entire chapter to that peculiar breed of trouble-maker. One or two of the stories will be X-rated, of course, but those were extremely rare occurrences. Right, Sam?... (nudge nudge wink wink)

Indeed they were, Arthur.

More on this later.

Sam

Arthur, I will be first in line to buy your book.

Having spent a couple days in Arthur's company at this year's TCI show in Lancaster, I can say I don't think I've ever been as awed by sheer telephonic knowledge and entertained by stories by anyone as much as I was with him.

Thanks, Mom.

Modern network interfaces and switches will adapt to a crossover cable. Older network interfaces will not.

My personal preference (or policy) for equipment that requires a cross-over situation (channel banks, etc) is always to wire a pair of jacks for the "crossover" and use 2 standard A>A or B>B patch cords. In that way, thinking like a repairman rather than an IT geek, when the weakest link, the patch cord, goes wonky, it can be replaced quickly by any old body at the site, and there is no need for an expert to go out and create a crossover cord in the field.

The term for this is "auto=sensing" Only one ethernet interface needs to be auto-sensing. So if the switch has auto-sensing ports, you can plug old equipment into it, and the switch will figure out what it needs to do.

Maybe what you have is crossed pairs. This is a little harder to overcome. For instance, I once saw a cable wired white-blue, blue, white-orange, orange, white-green, green, white-blue, blue, white-brown, brown. This is entirely logical but may not work very well. The reason is that ethernet will use white-orange and blue for one pair (pins 3 and 6). This is not a twisted pair.

The only way to check this out is to open up some wall plates. Expensive cable certifiers can detect this. But a simple cheap wiremap tester will not.

You may have shorts in the cables. You can test this with a cheap wiremap tester. Shorting problems tend to be intermittent, and can be caused by someone tugging on a cable that has snagged on some metal somewhere.

I have seen a single computer bring down an entire network. Try unplugging some computers and see if the network problem goes away.

Twisted pair wiring was invented by Alexander Graham Bell in the nineteenth century to reduce electromagnetic interference from electric trams and power lines, according to Wikipedia. The encyclopedia that anybody can edit also says that twisted pair cables reduce cross-talk between pairs in a cable.



.

The first telephones used a single open wire, with earth return, and it worked because no one else was "using" the earth for a ground return. Then the electrical industry started using the earth as a safety return, and electrical railways used the rails (bonded to the earth) for their return circuits. This caused loud noises and humming, to the detriment of telephony.

So, the original single strand open-wire was improved by creating a balanced pair of wires, with two strands of open-wire to every subscriber. It became apparent that parallel runs of open wire, over substantial distances, created crosstalk and other interference, so the "transposition" scheme was introduced. There are several good descriptions, with mileage calculations, available in BSP's, BSTJ articles, etc.

Once we had the capability to produce multi-conductor cables, the wires were laid up in pairs, since this was already an accepted and established engineering practice in open-wire construction.

Neither Mr. Bell, the non-inventor, nor Mr. Meucci, the actual inventor, had much to do with twisted pair cable, since it appeared later in the history of telephony.

Arthur -

How about "Phantom Circuits"? If I remember correctly these were deliberately induced circuits caused by long runs of side-by-side wire (where you got an extra circuit with fewer pairs.

I never saw one in real life, but I do remember hearing about them in class at some point (generations ago).

Sam

If someone told you that the circuits were created by some sort of induction, they were lying... er..."laboring under a mis-apprehension". There would be no way to create or control such induced signals.

The term "phantom circuit" refers to the scheme whereby two "side" circuits, each comprised of a balanced pair, and each fed via a center-tapped repeat coil, could create a third "phantom" circuit. Since the two side circuits are balanced, there is no cross talk from the derived circuit.

https://en.wikipedia.org/wiki/Phantom_circuit

See page 148 in this book:

https://long-lines.net/sources/att_principles_ocr.pdf

There is a lot of amateur wiring out there. Here are some possible problems:

More than 1/2 inch of wiring was untwisted at the terminals.

A cable was tugged too hard during installation.

A cable is not at least 6 inches from any electrical power cable or does not cross the cable at right angles.

A cable has a bend with less than a 3 inch radius.

The jacks, punchdown blocks, patch panels and patch cables are not of the same category as the cable.

Mice and rats have been chewing on the cables. Yes this really happens.

All of these things can cause intermittent problems. Just because it worked yesterday does not mean it has to work today.

Aha! Thank you, Arthur. i seem to remember the instructor sketching the cable runs and the two center tapped relays.

Aeons ago.

Sam

Once upon a time, the TV station I work for needed to extend some police scanner feeds from a remote site. They phantomed a third circuit using the two pairs they already had and a handful of 111C's.

Back to the OT, I did find at one of my older office sites, an attempt to use some CAT3 cables where the blue pair was split off for phones on a 4P4C jack, and the other pairs were landed on the 8P8C jack, in 568B style (missing the blue pair, of course.)They were wondering why they couldn't get gigabit speeds. This was good enough 20 years ago for 10-base-T...

I have a WE 111C repeat coil sitting here, came from the radio station at my college...now back to your regularly scheduled programming

Back when "Interconnect" first happened (for those kids out there, that's 1968) it was customary to have VCA (Voice Connecting Arrangements) installed and maintained by the telephone company between the CPE and the telephone company's services.

This became a "Wild West Rube Goldberg" when it came to tie lines. It was common to have 2 wire circuits morphed into 4 wire circuits and vice versa using those A B transformers and a "derived" third pair that became the talk pair. If the tech wasn't aware that there was no "hard" wire way to test the talk path, end to end, through the VCA the tech would spend a lot of time on the phone with a CO switchman trying to trouble shoot a problem through the VCA that could not be tested. When that became a nagging headache, the switchmen began to instruct the techs to test on the telco side of the VCA instead of the customer side. If the end to end tested good from site to CO to CO to site, then the tech was to replace the VCA, run to their truck and drive away, quickly.

Being on both sides of the early "Interconnect" years, I understood the frustration the telco techs had trying to deal with the VCAs and, especially, little trained or not trained "Interconnect" techs that didn't know a tie line from a shoelace. I am guessing it was around 1974 when Bell Telephone finally gave up on VCAs.

Rcaman

I started installing a NEAX 31 in the Bronx VA in 1978. I set up the Anti-interface and was ready for NY Tel to install the VCAs. When the installer finally came out (in 1980! [The job got shut down for a year while they installed sprinklers that were left off the drawings!]) he let me know that because my PBX had already been installed on the Network without a VCA (GTE at Tampa International Airport) they were not going to be installing one for us.

I was ecstatic. They caused nothing but trouble. I think it was a little after that, that NY Tel did away with them completely.

Sam

Here's a funny note about the VCAs. The Bell Telephone techs were NOT trained to install or option the VCAs. I actually provided the Bell Techs a cheat sheet for them to option the cards correctly. After a dozen or so installs, the Bell techs just left the VCA cabinet and cards on site and never installed them. Like you, they were fed up with trouble calls because those cards were so badly made. If you carried around a fist full of 120 ohm 1 watt resistors, you could fix most of them in about 5 minutes.

Rcaman

Having come from the other side (hmmm...that sounds like the punchline to a dirty joke) that is, having been the BOC tech who was dispatched to correct the problems caused by the VCA's, I can tell you that the most common troubles were fixed by simply by-passing the VCA whenever possible. The repairmen knew that, the foremen knew that, and apparently everyone all the way up the chain of command knew that. The VCA's were good for a lot of overtime, though.

I remember doing a small key system in a real estate office (this was in 1975). It was a Friday night; I had completed the installation, NY Tel had cut out their equipment and was about to throw over the interface equipment. All was well until they went to plug it in. They had the wrong AC cord! (They had the one designed for a shoe box, not the one designed for the standard power pack).

The locker was closed, the next day was Saturday, the office was set to open - with no phone service. I said we'd have to wire around the VCA. The installer called his foreman to explain and the foreman said, NO. The customer was going to have to do without phone service till Monday.

The installer, frustrated and embarrassed at this point said something like: "Goddamnit! They've got an ITT 584C panel, Stromberg Carlson phones an Elgin power supply and Northern Electric Speakerphones! What the hell kind of damage do you think they'll do to the network?"

We hardwired the VCAs and I'm not sure of they ever got turned on. The job was out in Brooklyn (interestingly, just down the block from where I live now) and I'm pretty sure I never went back there....

Sam

Ditto. There were many new installs that had equipment that never got powered up because it was a huge waste of time. I'll never forget installing an ITT 501. Bell showed up with a cabinet the size of a large ComKey system. There wasn't any room for the VCAs. The tech called his foreman who called, I think, the bartender at the local pub. Anyway, after several hours of trying to figure out where to put the thing, the customer said, "I have an idea, why not take it back and see if it can be installed in your garage." We all laughed, but the thing was never installed. It sat at the client's basement for years.

Bell Telephone techs were reasonable. The other operating companies were NOT! More times than not, they arrived with a Pulsecom card rack and a tote full of modules. No one had a clue what or how to install and the job usually fell to me after many hours of frustration. The stories are many, but the same outcome. The VCAs seldom worked correctly and just made for customer frustration and, as Arthur mentioned, many hours of overtime. That's one page of telephone history that is infamous.

Rcaman

I only encountered Bell's VCA cabinets once (yes, they were on the same chassis as a ComKey 718 KSU), as in huge. My parents bought a TIE 1A2 system for their real estate company in New Jersey in 1976. The TIE 1A2 KSU was about half the size of the VCA cabinet, if not smaller. One day, I was moving some wiring around and accidentally shorted one of the CO lines, which blew one of the 70- type fuses in their cabinet. It was obvious which fuse was blown, but I didn't have any. There was a NJ Bell technician there working on an unrelated issue, so I asked him for a fuse. He got ugly about it, citing that I had no business inside their sacred cabinet, and instructed me to call their repair service.

Not to be outdone by this ass, I just took one of the fuses for line 7 and put it in place of the blown one for line 1. I then called repair and reported NDT on line 7. Line 1 was back to working within minutes and nobody even cared about the last line being dead for a few hours. The SAME NJB technician arrived later that afternoon and replaced the fuse, and then gave me the remainder of the box of fuses for future use. I guess his attitude was the result of the fact that he was a repair technician and needed to justify his existence when so many customers were jumping ship to interconnects at the time.

Around here, (Western PA) most of the Key System techs and PBX techs were doing VCA installs and service. NONE were happy about it. In fact, most were hostile. After one or two encounters, I just walked out of the room and did not come back until they were stymied about how to option out the cards. I never was condescending to them. I was one of them a few months earlier. So, after 50 or so installs with the same techs, we became friends. The funny thing is, almost all of the PBX techs became our employees after they took early retirement. At one time, we had 18 ex-Bell or AT&T techs working for us.

They were all good techs. They just thought, as everyone else that worked for Mother, that they had a job for life. All of the original 18 have since retired or died. We still hire Verizon techs, but they have to be older techs that actually know something. I have copies of Lee's ABC telephone course and test. When a Verizon tech asks for a job, I hand them the test. The older techs snicker, fill it out in 10 minutes and I hire them. The younger techs end up not finishing the test and usually just leave. Sure, the questions are dated, but I want techs working for us that know the difference between Bridle wire and a beer can.

I know Arthur...there is NO difference. LOL

Rcaman

...or who know "the difference between a diode and a Buick" as a foreman of mine used to say.

(cross-posted)
Is this gray unit a VCA?

No, it is just a channel bank manufactured by Pulsecom, a division or subsidiary of Hubbell. Nothing out of the ordinary. T1 circuit in and 24 POTS lines out.

LOL...you could tell if it was a VCA. There would be bullet holes in the cabinet.

Rcaman

OK so I got a couple more pics of the mystery Pulsecom box: