Electrical Interference on Cable Internet???

[Felder]

Howdy all. I recently upgraded my internet connection from dial-up to high speed via cable (about time, I know). Since installing the cable I have been having problems getting connected to the internet. If I connected the modem straight to the cable drop coming in the house I would get a good connection, but when I run a cable from the drop to the computer (about 25', no splitters) I would fail to get a connection. Sometimes it would connect, but most the time it would not. So, I got to looking at the cable run (finished basement, drop ceiling) and discovered that in two places the cable was running along side and sitting right on top of some romex. Hmmm, I thought to myself. Could I be getting interference on the cable from the romex? So, I moved the cable several inches away from the romex and low and behold a got a good internet connection.

I have a cable guy coming out later this week to check my signals so I'll ask him if the electric lines could cause enough interference to affect my internet signal. It appears so and makes sense to me especially if the cable is right on top of and running parallel with the romex. Has anyone had any similar experiences?

-Felder

[Steve Barker]

Also, using a good quad shield RG-6 will help the situation. Make sure the shield is grounded somewhere along the line.

[Art Todesco]

[gfretwell]

Call the cable company and get them to fix it. You are paying plenty for the service.

[Felder]

Apparently nobody read my full post. I have no problems with the cable. I did have until I move my cable line away from the electrical lines. I was just curious if anyone has had a similar experience when running a cable internet line too close to an electrical line.

-Felder

[mm]

I thought they did read your post. Why do you think,and why do you mention, that they didn't?

Exactly. You did have problems. I took their posts to say that if one used the better cable, even when the two cables ran together, right next to each other, there likely wouldn't be interference.

[buffalobill]

no. but clarification needed: this sounds like a self-installation that simply needs a cable company technician. our cablemodem runs off basic time warner RG6 cable wire. the cable company in buffalo ny installs and maintains it. does yours? are you speaking of a second type of wire, ethernet [category___] wire fed from a router? here that is maintained by the homeowner.

[gfretwell]

I think you have another problem. The cable runs for miles on a pole in line with 13kv or more on the top strand and usually a 120/240v secondary. The ingress into the coax is negligable and intentionally blocked anyway. My cable splitter/amp is about 2" from my electrical panel with no problems. If I had to bet on what you have going on I would say it was a bad termination on your cable. The cable guy will usually cut off all the ends and put new snap/lock ends on ... just to protect his plant. In the end, the cable company is responsible for preventing ingress. That is why they cringe when homeowners start stringing cable.

[Mark Lloyd]

I have cable internet, which works well. I put the modem as close as I could to where the cable enters the house, and ran ethernet (CAT5) cable the rest or the way. Also, CAT5 is easier to handle than coax.

And (once you get a good connection) there's other things to worry about. An inexpensive router between the cable modem and the computer(s) is a good idea, even if all you have is one computer.

You can ask the cable company about that (SOME are helpful that way), but there's really no need to. It's YOUR network.

[Mark Lloyd]

And the quality of the connectors (and their installation) is important too.

[Felder]

Because I was getting suggestions on how to fix the problem. I said in my post that the problem appears to be fixed. I was only hitting on anyone with similar experiences.

Okay, I see what you mean. I am using thr cable supplied to me by the cable company (Time Warner - Road Runner). It is RG-6. Still no problems since I moved the cable away from the electrical lines. So, to me that implies I was getting interference from the electrical lines. Maybe, maybe not. It's only been 24 hrs since I "fixed" the problem. I'll ask the cable guy when he comes out.

Sorry for any misunderstandings.

-Felder

[BillGill]

This is my second try, but my response from yesterday didn't show up in my list today.

You did the right thing in moving the coax away from the power line.

Generally speaking, it is a bad idea to run communications cables parallel to power cables. There is always a good chance of interference. When I was an engineer at NASA the standard policy was to never run power cables and signal cables in the same raceway. And when signal and power cables have to cross they should cross at right angles.

Even if there is no apparent problem when running power and signal cables together there may be when unusual power conditions occur. I have been told of severe consequences when such events occur.

In the home power and signal cables should always be run in separate wall cavities.

Bill Gill

[gfretwell]

That is certainly a good theroretical design goal but reality is the power company and the telco/cableco share the same poles and underground ducts so their wires do run parallel for hundreds of miles. If you are getting ingress on coax or twisted pair, you have another problem. I bet you could see it using TDR in a second.

[Edwin Pawlowski]

On poles, they are often a few feet a part. What is the range or field that is going to matter? I would imagine the difference between a couple of feet and touching makes a difference in signal, magnetic field or whatever.

[gfretwell]

It will still add up and on a power pole in a neighborhood you usually have a single 13kv unbalanced primary. In your house they use balanced cables with a hot and neutral or 2 balanced hots.

When I was doing communication wiring we set up a torture chamber in the office and did everything you "can't" do like looping LAN cables around florecent fixtures and running along next to the building feeders from the transformer. We found that if everything was terminated properly it had zero effect. It was only when we had other problems that it screwed up the data. This was using test programs that collected LAN statistics with unusually high loading. Again, the TDR flagged these problems immediately. Coax and twisted pair are very good at eliminating outside interferance

[BillGill]

That means that good design rules don't apply because you know that they don't really mean anything. And of course everybody lives in an ideal world where all circuits are properly terminated and there is never any need to follow the design rules.

Power and communications lines do not share conduit in long distance runs. They may be "near" each other, but they are separated not only by space but by earth which makes a pretty good shield.

On the pole out in back of my house the telephone and cable lines are something like 6 to 10 feet from the 220 lines and about another 20 feet from the 7 KV lines. With the square cube drop in level that makes quite a difference in the interference level between say 6 inches and 10 feet.

And power lines are not balanced. One side is grounded, both the neutral and the ground are grounded at the entry point to the house. The closest you can come to a balanced power line is a 220 line. And that may not be completely balanced, because the device connected to it may have a 120 circuit in it that will cause an unbalanced flow in the 2 legs.

The best thing to do, even under ideal conditions is to always follow conservative design rules. That will help to take care of the cases where things don't meet the ideal that you say will take care of all of our problems.

Bill Gill

[gfretwell]

If you don't make good terminations all the routing design rules in the world won't help that much and if the terminations are good it doesn't matter. Coax and twisted pair are very robust transmission methods. Certainly you can go around separatring cables and rearranging them so they cross at perfect 90 degree angles but if that is fixing anything you are just pissing on the fire. That time would be better spent fixing the NeXT problems you have from bad terminations and impedance problems damaged cables. I never designed anything but I fixed them for 40 years. I know what works.

[BillGill]

I've said it before. Good design rules are good at any time. Good design rules do include good terminations, but they aren't everything. Follow good conservative design rules and the chance of ever having a problem is much reduced. Don't throw out the baby with the bathwater just because you have all the answers. Not too long ago I had a cable tech in because I was having an intermittent cold weather outage. He walked in and jumped right on the connectors on my cables and refused to listen when I tried to tell him the problem. The system was working when he got there, because he got there after sun rise and it had, as usual, started working. I admit he got a little better signal on my TV, but he didn't fix the problem and refused to admit that there was a real problem. It sounds a lot like you. There are other problems than just bad connectors. And the way to help reduce them is to follow good design rules in all cases.

Bill Gill

[avid_hiker]

This is highly correct.....running parallel to electric ( AC ) lines will cause interference. AC lines create a magnetic field around the AC wire.......when running Cable line parallel to it; the magnetic field basically raps or surrounds the Cable line and therefore causing interference. Note-The magnetic field will not be felt if cable line is run at a 90 degree angle across the AC line nor if cable line is far enough away from the AC so that the "M" field does not encircle it. But- you got it and found it.....good job!!

Dean

[gfretwell]

I don't suppose "common mode" means anything to you guys.