Re: A wire is a wire is a wire

Bicephale

join:2005-09-24

reply to TakeTheFifth
Re: A wire is a wire is a wire

Hi Phil,

I'd like to use a more descriptive expression than
just the term "60 Hz noise", which is quite vague,
but my GNet only provides a simple diagnostic tool
through 'DMT' and my SpeedTouch even less.  In any
case, i did suspect there was a wider problem this
winter and i still do so the next step would be to
put the MoDem on battery since that's all i can do
about such external nuisance;  when it comes to AM
radio stations and storms i'm short of imagination
but it wouldn't matter with these numbers, though:

             Source: GNet BB0060B                Date: 2007-Sep-8                Time: 11:27:52        Bits-per-Bin: 13       UpStream Bins:   6 to 31 inclusively     DownStream Bins: 33 to 254 (idem)      Tx Power Attm.: -1 dB         Coding Gain:   7 dB             Up Time: 12:12:37 (43957 sec)           Local CRC:   142/day (72/43957 sec)          Remote CRC:    16/day (8/43957 sec)           Local HEC:    90/day (46/43957 sec)          Remote HEC:     4/day (2/43957 sec)      Local Tx Power: 10.95 dB     Remote Tx Power: 19.40 dB Local Line Atten.: 31.5 dB Remote Line Atten.: 29.0 dB    Local SNR Margin: 18.5 dB Remote SNR Margin:   8.0 dB

Even a twelve hours record has little meaning here
and i don't believe this is the end of peak rushes
but that's nowhere near a thousand errors per day,
(around four thousand until recently to be exact)!

As i wrote, the wires got to go back into the wall
so there's no other way to manage with the limited
space than by shielding Bell's main while the rest
of the phone lines are moved by a few inches only.

Thanks for your link, i already feel apprehensive!

to forum · permalink ·

· 2007-09-08 12:42:38 ·

TakeTheFifth

join:2004-04-20
Anjou, QC

1 edit

reply to Bicephale
I would be more concerned about interference riding the power lines than the 60Hz signal itself: hi-frequency junk (from electric motors, arc welders, amateur radio, etc). If have seldom seen nice and clean sine wave house power; If you live near any kind of garage, industry, machine shop (or are on the same power grid), and you suspect your problems are related to power line interference, they would be the culprit. At any rate, keeping phone lines away from power lines is a good idea (and if you can avoid having them run parallel, even better).

Some other sources of RFI: »cable-dsl.home.att.net/#RFI

Phil

to forum · permalink ·

· 2007-09-08 09:50:40 ·

Bicephale

join:2005-09-24

·TekSavvy Solutions..

reply to ArthurS

60 Hz Noise - The AC coupling is actually revealed!

Hi ArthurS,
Hi Angelo168,

My numbers ain't as accurate over a few hours than
they can be over a week so i better hold judgement
based on them until enough records collect but the
graphic shown above should speak for itself!  This
looks like some threefold improvement to me and it
also suggests i only need to shield Bell's main as
it turns out there was only a single phone instead
of the whole wiring when the 18 h peak occured:  i
put back the original circuit for the 19 h peak so
that's a strong indication that the stacked double
filter setup does work at 60 Hz, after all...  The
single phone provided some loading and i put it at
a good distance from the electrical panel in order
to control noise pickup;  there should have been a
difference in the peak's amplitudes if the filters
couldn't keep the 60 Hz noise away compared to the
reference load.  Great, i was quite concerned that
having Bell's main besides of the home wiring in a
close shielded space might actually couple them to
a point that noise from the house wiring would end
up being injected on the main line anyway - i just
hope the impedance won't be affected too much once
i'm done inserting the wires into the 1" sq. steel
tube i've found!  Well, at least there's no splice
to deal with and the shielding tube just needs the
length to be cut then i must get screw-pads added.

Piece of cake!

to forum · permalink ·

· 2007-09-08 07:56:59 ·

ArthurS
Watch Those Blinking Lights
Premium
join:2000-10-28
Hamilton, ON

reply to Bicephale

said by Bicephale

:

I tought of using galvanized
steel tubing because it would shield my DSL signal
against electric and magnetic fields, i guess, but
i prefer to evaluate simpler remedies for a while:
maybe displacing the demarcation box would suffice
to fix this issue permanently (i once had a 6 Mbps
profile without much trouble, after all!) but then
i'd probably need to displace the MoDem as well so
i'm considering my alternatives.  In the meantime,
the possibility that a pair of stacked filters may
not have enough punch at low frequencies to reject
60 Hz noise is a big concern, i might have to move
the other phone lines for the duration of the next
few tests to the very least!  I wish i could do my
little experiments using the SpeedTouch so i asked
about Linux Live CDs (in the TekSavvy forum) but i
doubt this will be sucesssful as it wasn't before.

After six months of wondering, my nightmare may be
over, finally.  I can now turn my frustration into
a renewed motivation to improve the situation.  It
seems i got no choice but to grasp the opportunity
to reduce the negative effects of cross-talk which
are bound to grow while DSL gains more popularity!

Even galvanized steel pipe is limited in its shielding from low frequency (60 Hz) electromagnetic fields. For best results, use rigid galvanized conduit (yes it's the most expensive kind of conduit and a pain to work with), and maintain a separation of at least 6 inches between high voltage and low voltage (which is required by code). In my specifications for high end commercial applications, I insist on one foot minimum, with both AC power and low voltage wiring being in separate conduits. In your situation, maximizing separation distance between the two is the cheapest and most cost effective solution for home use.

There is a lot to be gained from a neat installation. Starting over and doing it right sometimes is the best option.

to forum · permalink ·

· 2007-09-07 14:01:20 ·

Bicephale

join:2005-09-24
·TekSavvy Solutions.. reply to ArthurS
Hi,

I increased the distance between the AC wiring and
Bell's main (including its ground wire) but i left
the internal phone lines in place. This is only a
temporary configuration because it must go back in
the wall somehow, eventually... The sampling rate
being fifteen minutes these tests are very slow so
i need to be very patient before i can evaluate my
error rate reliably. I tought of using galvanized
steel tubing because it would shield my DSL signal
against electric and magnetic fields, i guess, but
i prefer to evaluate simpler remedies for a while:
maybe displacing the demarcation box would suffice
to fix this issue permanently (i once had a 6 Mbps
profile without much trouble, after all!) but then
i'd probably need to displace the MoDem as well so
i'm considering my alternatives. In the meantime,
the possibility that a pair of stacked filters may
not have enough punch at low frequencies to reject
60 Hz noise is a big concern, i might have to move
the other phone lines for the duration of the next
few tests to the very least! I wish i could do my
little experiments using the SpeedTouch so i asked
about Linux Live CDs (in the TekSavvy forum) but i
doubt this will be sucesssful as it wasn't before.

After six months of wondering, my nightmare may be
over, finally. I can now turn my frustration into
a renewed motivation to improve the situation. It
seems i got no choice but to grasp the opportunity
to reduce the negative effects of cross-talk which
are bound to grow while DSL gains more popularity!

to forum · permalink ·

· 2007-09-07 10:05:34 ·

ArthurS
Watch Those Blinking Lights
Premium
join:2000-10-28
Hamilton, ON

1 edit

reply to Bicephale
Yikes! What a mess! I hate to see the rest of the wiring!

A few points:

1. This is clearly an electrical code violation. Low voltage wiring (CATV, telephone, data) must be kept separate from high voltage (120 VAC) wiring!

2. Distance is your friend. For every doubling of distance, you noise goes down an order of magnitude (inverse square law here). I would try 6 inches separation at the very least, especially close to any significant current carrying conductors.

3. Be careful how you bundle/route your wires, particularly your data cables. Interference rejection is dependent on the consistency of the twist in the wire pairs. Using zip ties wrapped tightly around cable, routing cable around sharp corners, snagging wire into knots disturbs this twist pattern, making another point of entry for noise.

Let us know how you do!

Edit: I advise caution when dealing with a rat's nest of wire so close to AC power wiring! I would turn off AC power at the breaker panel before messing around there as a safety precaution! Under the right conditions, 120 VAC can kill!

to forum · permalink ·

· 2007-09-07 00:53:46 ·


Saturday, 28-Nov 17:52:16	Terms of Use \| Privacy Policy \| Hosting by www.nac.net - DSL,Hosting & Co-lo \| feedback \| contact over 10 years online! © 1999-2009 dslreports.com.republican-creole page compression OFF