Electromagnetic Interference

File: 20161211_122751.jpg (1MB, 1753x986px) Image search: [Google]

1MB, 1753x986px

>>1097792
Pic 2/2, this is with the insulation on the phone line.

Why not cut the line and resplice it away from the power...

>>1097796
>Without splicing wires

Trying to avoid that. Mostly because I'd have to run the extension cord about 45 feet across the attic to get my soldering iron over there, but it's looking like either that or unhook it from the wall in that room, untangle it, then run it back down.

It's CAT3 line, but that really doesn't make much of a difference at the moment.

You don't need to splice phone wires, you use jelly crimps because it's a million times quicker.
If you think that the power cable is creating rein check it first, turn it off and on while watching packet loss.
Also plugging the router into an extension is all kinds of wrong, ideally it should be at the lead in, first point of intervention with a filtered socket taking the dsl off before it fucks off around your house and goes wrong.

>>1097797

Get a butane soldering iron

>>1097798
This

Every piece of this is right

>>1097792
the EASIEST would be to wrap some aluminium foil around the telephone wire
will it help? possibly, if you can ground it to something even better i guess, any sockets up there?

the BEST solution is to separate the cables.

i have helped a few friends who had shitty internet and thought that something like this was the problem. turns out it almost never is and fixing it doesn't help shit.

of the potentially hundreds of feet of copper wire buried underground or strung between poles in the sky and all the associated connections and junction boxes between your house and the exchange the likelihood of the error coming from a few inches being 'near a wire' is doubtful.

let us know how you get on however!
if its cat3 and theres a junction box either end anyway you might as well put in something shielded to put your mind at ease. leave the original wire there in case the phone company throw a fit or something i guess.

>>1097832
Well the phone company has checked the lines about two dozen times. (Literally over 20 times. I'm sick of it).
So the issue is inside the house. And even when I disconnected all other lines in the house leaving just the one from outside to this jack, it still had the issue. So I'm positive this few inches is the problem.

>i have helped a few friends who had shitty internet and thought that something like this was the problem. turns out it almost never is and fixing it doesn't help shit.
Elaborate, please?

>>1097832

Shielding won't do anything, also if you are hearing anything on the line, that's a physical fault, depending on the noise. If it's a rustle, probably a high resistance dis in one of the pair, if it's a
Constant hum then it's probably picking up an earth from somewhere on its route, there's a fuckton of possibilities, and it can be a few inches, it can be a 100th of a mm that causes it, the trick is in finding it, which you will struggle to do without the proper kit, just cut that wire, buy some decent twisted pair wire and fix your house out. Then when it's still shit tell your provider that there's noise on the line and that your internet goes down when the phone rings.

>>1097835
In my own house, adsl had shit snr, phone cable came in attic and ran along with mains wiring. Ripped it out right back to demarcation, no mains wires anywhere. Snr did not improve.
Friend of mine found chewed cable feeding telephone socket, complaining service was intermittent, cable replaced no improvement. Checked all wiring back no other damage.
Another friend complained to be about intermittent WiFi, thought it might be you do with the messy cables at the router, turns out WiFi dropped out when using a certain phone, no filter. Added a filter worked great.

Phone company can run a resistance check to make sure you aren't stealing offhook current from them or ping your router to see if there are any packets dropping for maybe 5 seconds. If a tree branch is rubbing against your line or a mouse had made a nest in a street cabinet or I don't know harmonics from a pedestrian crossing beeper are traveling up the power supply cable next to your phone line under the street and the frequency interferes with your broadband whenever someone crosses the street, maybe its just a capacitive line when its damp, the few seconds that some headset in India have available to waste on you before they need to hit the average call time target aren't going to let them do anything significant.
Even if you got a guy out to run a proper frequency response of your line what are they going to do, tell you it's fucked? Then what? If they can't fix it then you don't want out, and they don't get paid. Either that or rip up the street to replace the cable. Do you know how much that shit costs? Just the permit to close the road is not worth it. For your intermittent issue.
But by all means I wish you every success in your endeavour and sincerely hope to hear that it was this cable causing your problems.

>>1097878

Most things can be fixed, most, there are major issues that can't, but then you have to find a workaround. As for interference, once you find it then it can be cured to am extend.
Poor snr can almost always be improved, most of the time is going to be high resistance somewhere. I don't know what it's like over in americuckistan but over here in britbongland there are multiple tests done on the copper remotely that can give early indication of an issue. Measures resistance, capacitance, AC balance, resistive balance & voltage.

>>1097882
All of those tests have been performed multiple times (20+) and the issue all but one time were determined to be in the house.

Aside from the odd white noise-like sound on the phone line, the internet connects at its max speed all the time and only has issues during storms (That's expected) and when the treadmill is in use.

The EMI from the treadmill is enough to kick the modem off of the internet every single time like clockwork, and it will always reconnect afterwards, though sometimes at lower speeds.

>>1097885
>when the treadmill is in use.
maybe you should have explained you had it narrowed down.
maybe stop using the treadmill? get a treadmill without a motor in it? go outside to run? hand wind a choke at the emi?
>though sometimes at lower speed
thats how the hardware is supposed to work, if it can't connect it will work its way down in speed until it finds a speed that it can work at (lower speeds aren't as succeptible to absolute fuckery from interference)
when you stop the treadmill the exchange will slowly try to renegotiate at a higher speed to see if it was just a temporary interference and work its way back up to the full speed the line can handle.
this is why they tell you not to turn off your router regularly, it takes time to negotiate to the max line speed because the exchange is actually retarded and thinks the line is damaged rather than you just turned the router off.

>>1097886
>when you stop the treadmill the exchange will slowly try to renegotiate at a higher speed
No...? I have never, ever, ever seen that happen. Once it connects, it's connected. It doesn't renegotiate anything until service is interrupted again, either by me resetting the modem, or through EMI, or what have you.

>>1097887

Its called dynamic line management, it can take up to 72 hours for it to go back to where it should.

But stop using the fucking treadmill.

>>1097898
>stop using the fucking treadmill
No.

Anyone else have experience in handling EMI?

You're probably on the right track in thinking the problem is inside the house, but it may not be the phone line. Crossing the power line at an angle is not going to induce a problematic amount of interference.

However, any phone tech that would do shit like that probably made other mistakes as well. Start looking at jacks and splices. They're bound to be sloppy. Sloppy jacks and splices mean poor electrical connections. I would check near that room.

Your problem may also be power or the electrical devices you are using. Poor splices and terminations could put out more noise and if you're running a tesla coil or some shit in your room then yeah, it's gonna cause issues.

Check anything near a phone, phone outlet, phone wire, phone anything in that room that can be moved.

If you can't figure it out, you could have the lines in the house checked.

>>1097792
AT&T tech here, that power line isn't doing shit. I've seen a drop wire completely wrapped around the mains power line from pole to house and come up dead clean. Does your phone wire daisy chain from room to room, or does each jack have it's own homerun? Have you tested the phone line at the demarc and established that the static is present on the drop before it enters the house? Do you have an external splitter or those stupid plugin filters?

>>1097887
You wouldn't see it happen unless you checked the router connection status page continuously for the next 4 or 5 days.

How are you measuring the emi to know its significant or are you just basing it on the internet getting fucked up?
If you could scope or ideally get a spectrum analyser on the power line and then find out the frequency of the interference and then get some capacitors on the motor or ferrite beads as chokes on the power cable for the treadmill might help. If you want to investigate the shit in your power lines there are a few projects where you get a very low voltage mains transformer and a voltage divider for maybe 1v output then plug it into your soundcard line in and analyse the frequency and harmonics and shit from all your smps. Might help you tune in to see what helps removing noise.

OP i used to work for AT&T, and quit about 7 months ago.

If you have any questions about scotch locking the cat 3 let me know. I'll be ITT for a while.

>inb4 no splicing


its very, very, easy.

I'm confused as to why you'd 'splice' the network wire when the power wire would literally be a box with some wire nuts in it.

>>1097954
Are you mentally retarded?

>>1097954
Because power splices should be accessible.

Ops treadmill is killing his line quality. It might be worse because it's running over that wire. You would have to lift it up substantially to reduce the effect too. The cat3 isn't doing enough.

I would second plugging modem directly into nid and would then add you should use the treadmill to see if it drops at the nid. If it doesn't then you know you can improve your wiring.

Easiest thing is to cut it and bring it above the power and even suspend it a foot or something.

Next easiest is to run a new cat5 from the nid directly to modem and just do it outside he house as easy as you can.

Hard mode is replacing the existing wire in the attic all.the way back to the nid.

>>1097997
I have seen equipment bad enough it knocks out whole neighborhoods when used. We had to buy a guy a new TV because his was throwing out so much emi it would kill all the dsl in the area.

Buy some snap on ferrites. I had a LED light with a bad power supply that was completely fixed by putting ferrites on the lamp cord.

>>1097792
Also check any older compact flourescent "curly" bulbs you might have around the house. I struggled with DSL issues like yours for months only to accidentally discover they happened when my porch light was on. Went back to incandescent and problem never came back.

>>1097998
Yep. Plasma TVs are particularly bad in this regard.

>>1097997

this

I would move the DSL modem to directly next to where the phone line comes in and disconnect any circuits that branch off that aren't in use.

I loathe daisy-chained phone wires, every time I renovate a room I run 2-4 cat5 wires down to the basement. get shielded cables and you're all set

>>1099900

Get a quality LED and save the power