Network wiring question

CatSnoutSoup

NES Member
Joined
Jul 9, 2015
Messages
7,801
Likes
21,090
Location
Westgardminsterham MA
Feedback: 0 / 0 / 0
Not ham wiring, home wiring. Figured I would ask here as it might focus the question toward being seen by the more technically minded.

After many years without use our home Ethernet wiring is going to be put to use (cable company finally is going to expanding their plant out here into the woods).

So I have been testing all the connections electricians wired when I build the place 15 years ago.

I have a cheap $28 Klein network cable tester, which was good enough to find one RJ45 jack in the garage had been miswired. Guy had crossed the Green and Green/White, but I was able to carefully pull them up swap them and re-punch them. So all the cable runs for networking ( Cat5e BC) are properly terminated at the wall jacks and the distribution panel. The panel runs to a new TP-Link switch that I installed and patched in with some flat Cat7 cables.

So here is my issue all of the network and phone wiring are terminated in RJ45 jacks wired to T568A standard. By design there are no RJ11 phone jacks the RJ45 can accept an RJ11 plug.
The idea was that a cable run dedicated to telephone could be converted to network use just by disconnecting it from the phone panel and terminating it to a network panel that was configured for T568A.

That is what I want to do with one of those telephone cable runs.

If you look at the photo you will see the network distribution panel in the upper left. Notice that there is one open spot #5.

In the lower left you see a bundle of Cat5e cables, striped and then punched onto the telephone punch down block.

I want to take one specific phone cable associated with a jack in our living room and move it to the empty space on the network board, problem is I do not know which one in that bundle it is.

We do not use the phone wiring at all, we do not have landline service but I do not want to disconnect all that work.

I have a very cheap tone line tracer and I have tried plugging the emitter into the jack in the living room and using the probe to check for the wire at the panel.
Trouble is they are all toning at about the same volume. Either the line tracer is not isolating enough or it is simply because all of those phone lines are wired in parallel so they are all seeing the emitter.

OK so how can I deduce which line from that bundle in the lower left leads to the jack in the living room without tearing out all the wiring on that phone punch down block???

Photo is big to allow folks to zoom....

ygyNM2f.jpg



🐯
 
Last edited:
How about removing the living room jack, pick two wires (any two will work) say green and green/white, short them together then go to that punch block and find all the green and green/white pairs and check them with an ohm meter, which ever pair shows a short is the cable that goes to the living room.
 
Borrow a better tone generator and hook it up to the RJ45 jack in question living room. Then you can find the wires you are interested in.

.
 
Get a crappy patch cable that you don't mind sacrificing.

Cut the cable about 1 foot from the RJ45 connector and strip away the jacket and expose some of the wiring. Or make your own from a scrap piece of Cat cable and only put on one connector.

Plug the RJ45 connector in to the jack in the room you want.

Pick a pair of colors that aren't used for the phone line. Clip your tone generator to those bare wires.

Go tone them out in the basement. That should solve your problem of the other wires being connected in parallel. The wires you are testing should not be connected to anything phone related.
 
Last edited:
How about removing the living room jack, pick two wires (any two will work) say green and green/white, short them together then go to that punch block and find all the green and green/white pairs and check them with an ohm meter, which ever pair shows a short is the cable that goes to the living room.

OK except isn't it true that because they are bussed together in parallel a short across one is going to appear as short across them all?

Here is a close up of the phone block. Looking just at the top as an example.

The pairs from the wall sockets come in on the right most posts, alternating Blue and blue/white as they go down the posts.

The connecting post for each wire is the one just to its left.

Those post are taken up by the pair from the phone service provider which has been brought it at the very bottom and serpentine up alternately bridging either all the blue, or all the blue/white. (Note: the arrow I have inserted).

So correct me if I am wrong if I put a meter probe on one blue I am actually make contact with all the blues?
And if I put a meter probe on one blue/white I am actually make contact with all the blue/whites?

If I am looking at this wrong please tell me, because I surely could be.

ijmcnMu.jpg



🐯
 
Get a crappy patch cable that you don't mind sacrificing.

Cut the cable about 1 foot from the RJ45 connector and strip away the jacket and expose some of the wiring. Or make your own from a scrap piece of Cat cable and only put on one connector.

Plug the RJ45 connector in to the jack in the room you want.

Pick a pair of colors that aren't used for the phone line. Clip your tone generator to those bare wires.

Go tone them out in the basement. That should solve your problem of the other wires being connected in parallel. The wires you are not testing should not be connected to anything phone related.

This might be a solution, good looking out.

🐯

P.S.> I started writing the reply above about an hour ago and got distracted, didn't realize there had been other replies.
 
Get a crappy patch cable that you don't mind sacrificing.

Cut the cable about 1 foot from the RJ45 connector and strip away the jacket and expose some of the wiring. Or make your own from a scrap piece of Cat cable and only put on one connector.

Plug the RJ45 connector in to the jack in the room you want.

Pick a pair of colors that aren't used for the phone line. Clip your tone generator to those bare wires.

Go tone them out in the basement. That should solve your problem of the other wires being connected in parallel. The wires you are testing should not be connected to anything phone related.


Found it!

Funny I had just thrown out some ratty old patch cables and I had to go out to the street and rummage through tomorrow's trash to find one to cut up.

Striped the brown and green hook it up to the generator and I was good to go.

Thanks again, that probably never would have dawned on me.


🐯
 
You should re-terminate all the network cabling, it wasn't done right. There is way too much outer insulation stripped back, shouldn't be more than an inch from the 110 blocks (the punch-down location). This will result in dropped frames, but you'll only see it as poor performance. Ethernet, layer 2, is really good at recovering from this kind of thing, and for the most part even network techs will only look at layer 3 drops, a.k.a. packet loss, which won't show until the frame drops are so bad that they can't be compensated for.

You'd be surprised at how often performance problems are a result of layer 2 problems.

You can look for this kind of problem in the cable modem as well, they will show them as unrecoverable errors and the modem will list them per channel. The cable tech will look at a ping and say there are no drops (layer 3 again), but the channel errors are layer 2. Those channel errors get cleared each time the modem is rebooted, some are normal during initial negotiation, but then they should change very little if at all. Knowing this keep the cable company honest and they will fix the issue, otherwise they will just keep telling you it's your equipment.
 
You should re-terminate all the network cabling, it wasn't done right. There is way too much outer insulation stripped back, shouldn't be more than an inch from the 110 blocks (the punch-down location). This will result in dropped frames, but you'll only see it as poor performance.

It certainly would look neater that way and I wish they had done that, but what is the physical reason that this would cause the signal to degrade.

This is Cat5e cable and the jacket is not shielded. It does not have a spline to keep the pairs separated like Cat6, and if anything they are more separated now. The twisting of the pairs is what maintains isolation, and the electrician has maintain the twists right up to the terminals.

So how is the lack of that extra 5-6 inches of PVC jacket going to affect signal quality? Honest question, as I don't know.

🐯
 
Last edited:
It certainly would look neater that way and I wish they had done that, but what is the physical reason that this would cause the signal to degrade.

This is Cat5e cable and the jacket is not shielded. It does not have a spline to keep the pairs separated like Cat6, and if anything they are more separated now. The twisting of the pairs is what maintains isolation, and the electrician has maintain the twists right up to the terminals.

So how is the lack of that extra 5-6 inches of PVC jacket going to affect signal quality? Honest question, as I don't know.

🐯
The cables inside the jacket are twisted in pairs AND those pairs are twisted together all for the same reason, to prevent crosstalk, essentially an induced current bleeding across one wire to another. This is mitigated by the twisting and is enhanced any time cables run in parallel, and it occurs even more in that lovely back and forth tightly wrapped bowtie you've got.

Now we have to go into how ethernet signalling works. It's not the simple on-off you may think. It uses a system call winchester differential, at least baseband over UTP does and that's what you've got. The receiver looks at the signal in very small time slices. (for simplicity's sake I'll stay with 1Mb being equal to 1,000,000 bits and not get into the whole 1024 because it's binary thing). So in a 1Gbps UTP connection the signal is cut up into 1000000000 slices per second. And during that slice of time the voltage can be in any one of 3 states, staying level, going up, or going down. level=0, up=1, and down=1. So it isn't looking for x volts, but rather for a change in volts, it think it a 2v change that triggers, but my memory on this detail isn't clear.
Anything that interferes with clear detection of that change will cause an error and the ethernet frame will be rejected. We'll say that frames are 1500 bits (they aren't but it's close enough), so a 1 in 1500 error causes the entire frame to be rejected. BUT wait, the protocols are very good at fixing this. The receiver will communicate back to the sender to retransmit the corrupt fame, and if part of that communication is corrupt it will try to get that frame resent as well, so one error usually has a cascade effect and creates other errors and more traffic to get it all sorted out.....And none of this is visible at layers 3 and up which is where most techs are looking.
So it's seen as "slow performance" and usually results in ISPs, hardware providers, and software providers all pointing fingers at each other.

Now you will never get a zero error rate outside of a lab, so it's much more about managing your potential for errors, and keeping it as low as possible.

You'll have to trust me on the next part, or not.
Over my 20+ years I've fix many long standing problems by doing nothing more that walking through and putting together a list of sub standard cabling and getting that fixed. Poorly terminated cable, UTP run parallel to power, laying across motors and ballasts. I've flown across the country because some idiot laid a loop of 15" of STP on top of a loop of power cable (ok it wasn't that bad spending 2 days in San Diego on Viacom's dime, for 10 minutes of work). Layers 1 and 2 are you foundation, you can keep fixing you rock foundation or put your house on concrete.
 
Seeing as how Rob bumped it I guess I will add and update. That earlier hookup worked out but only after I connected the Cat5e cable to the garage to the LAN panel and then using a tester found that the electrician’s apprentice had miswired the RJ45 jack out there when he installed it years ago, he swapped a couple wires which I was able gently pull, swap, and re-punch on the jack.

I had not heard anything about the expansion of the cable system since talking to the subcontractor doing the engineering survey shortly after the announcement of the project.
However got the lasted word from Comcast (Xfinity) this past week that approvals have been granted, the system has been design, and materials have been ordered. They say they are now waiting on licensing of the utility poles by National grid. Not sure exactly what that means in this context, I have heard that some new poles were going to be funded through an FCC rural broadband program. I do know US Rep. Lori Trahan was involved and was part of a photo op when they announced the expansion back in October. Frankly I don’t trust Comcast or the government enough to feel comfortable until this is a done deal and they are dropping a signal at the pole in front of my house.

Still not sure at this point if I will have to provide them with a run of RG11 cable from the pole to the house. Don’t know if they will pull their own through my conduit. There is only one conduit and it already has a Verizon phone drop in it, but it is a 2” conduit with lots of extra room and a mule tape installed to pull another cable.

What do any of you guys use for terminating RG6Q. In the past I had used Thomas & Betts “Snap Seal” connectors, but doing a cell phone amp recently I used these Belden/PPC connectors and I really liked them. Liked them enough that I bought a bag of 50 when I saw a deal for $16. That is probably enough to last me forever.


🐯
 
Back
Top Bottom