dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
1393

andrew_b
@teksavvy.com

andrew_b

Anon

Replacing 25 pair phone cable. Any telco people here?

I want to improve my 2x VDSL lines by replacing the 25 pair cable from the closet to my modems.
This length of this segment is 50% of the total VDSL loop length to the DSLAM.





Goal is to improve the lines by 20%, is it realistic? Share your thoughts.
Can someone recommend a contractor for this in Toronto?


23awg Cat6 has considerably better attenuation/noise properties at ~12Mhz (12a profile) than the 24 awg cat3 which i assume is the same as 25 pair currently in the conduit.
All these cables are rated for 100ohm +- 15 impedance, not sure if impedance mismatch is a problem if cat3 is 85ohm and cat6 is 115ohm.





HiVolt
Premium Member
join:2000-12-28
Toronto, ON

HiVolt

Premium Member

Is this in an apartment building, or do you have access to that area? if so, why not pull just the CAT6 cables and bypass whatever is there?

andrew_b
@teksavvy.com

andrew_b

Anon

condo/apartment
it's a locked common closet one floor below
bigger pic »s13.postimg.org/gizltn2u ··· t_pg.jpg
InvalidError
join:2008-02-03

InvalidError to andrew_b

Member

to andrew_b
Unless there is a wiring fault, you aren't going to get much of a measurable benefit - much less a "considerable" one - from replacing wiring if your loop is indeed as short as you seem to think it is.

If you are unconvinced, just move your modem directly to the punch-down, punch in a jack with your unit's wiring disconnected and see how much of an improvement you get from completely eliminating the wiring you want to replace.

My bet is: almost no improvement.

In my apartment, I had 27.5dB SNRM using individual POTS filters for each phone and still have 27.5dB SNRM after TSI's tech visit to install the usual Corning splitter and re-wire DSL to use the second pair. No measurable gain from eliminating 60' of phone wiring, three filters and three phones from the DSL signal path.

Teddy Boom
k kudos Received
Premium Member
join:2007-01-29
Toronto, ON

Teddy Boom to andrew_b

Premium Member

to andrew_b
said by andrew_b :

All these cables are rated for 100ohm +- 15 impedance, not sure if impedance mismatch is a problem if cat3 is 85ohm and cat6 is 115ohm.

I'm just looking at a site that lists cat3, cat5, and cat5e as 100ohm +/- 15, but cat6 at 250ohm +/-15. My guess, impedance of 85 to 115 is a minor concern, but impedance of 100 to 250 is a significant if not major concern:
»www.atl-technology.net/p ··· at6.html

I think with home ethernet applications, impedance matching doesn't matter, because the switch ports buffer things. If you were running cat5 in the wall to cat6 patch cable, with no active port buffering the connection, the impedance mismatch is probably an issue. That is, even though places like this say it is not:
»www.broadbandutopia.com/ ··· aco.html

For a DSL, I think impedance is probably even more important, and there definitely isn't any active buffer. I'd stick with 100ohm cabling. If you are sure the cat6 you are thinking of is 100ohm, fine, but be careful

No doubt somebody who *knows* instead of *guesses* will be along shortly..
Teddy Boom

Teddy Boom to InvalidError

Premium Member

to InvalidError
said by InvalidError:

If you are unconvinced, just move your modem directly to the punch-down, punch in a jack with your unit's wiring disconnected and see how much of an improvement you get from completely eliminating the wiring you want to replace.

I suspect he has already done this. Specifically:
said by andrew_b :

This length of this segment is 50% of the total VDSL loop length to the DSLAM.

Doesn't that imply that he has tested, and there is some room for improvement?
InvalidError
join:2008-02-03

InvalidError

Member

said by Teddy Boom:

Doesn't that imply that he has tested, and there is some room for improvement?

You would need to check SNRM for that.

If his SNRM is already well beyond the "worry zone", he would be wasting his time and money.

There is also the possibility that the DSLAM will eat whatever attenuation he manages to shave off his wiring by reducing TX power levels at both ends so he'd end up with pretty much the same net SNRM anyway.

andrew_b
@teksavvy.com

andrew_b to Teddy Boom

Anon

to Teddy Boom
cat6 is definitely ~100ohm, at least under 100mhz

haven't tested, it's a locked common closet. need a contractor with a vdsl signal meter.

i have 53/12 down/up "max attainable data" rates.
20% is 60/14 which would allow for a solid 50/10 connection, and future proof my home.

a waste? maybe, but it's my money.
i like the idea of having my cat6 as close to the fiber as possible.

- modem to closet is about 150ft, most people have 1x line, since i have 2x, there is some cross talk all the way to the modems
- the other 150ft is alongside other peoples lines to DSLAM with major crosstalk, but i cannot change that.
InvalidError
join:2008-02-03

InvalidError

Member

said by andrew_b :

- modem to closet is about 150ft, most people have 1x line, since i have 2x, there is some cross talk all the way to the modems
- the other 150ft is alongside other peoples lines to DSLAM with major crosstalk, but i cannot change that.

The problem with signal integrity is that once it has been compromised somewhere you cannot fix, nothing you can do can undo the damage. If your building's wiring is in good condition and whatever is causing your modem to have relatively low sync rate for your distance is in Bell's outside plant, bypassing the building's wiring will have negligible to no effect.

Perfect wiring prevents signal from getting worse but does not magically undo what has already been done prior to signal reaching it.

xsbell
join:2008-12-22
Canada

xsbell to andrew_b

Member

to andrew_b
It looks like the riser is a 99-10, meaning Bell doesn't own it and it's up to building's owner to maintain it, that's why you see POTS splitters in the IT room.

As for the cable, you want lower resistance, not higher, so look to get something that's 250ohm or higher.

A tech has to check the metallics first, if there's power influence, resistance, or it's unbalanced.. attainable speeds will suffer.
AsherN
Premium Member
join:2010-08-23
Thornhill, ON

AsherN to andrew_b

Premium Member

to andrew_b
Where on earth did you get 20% improvment? Replacing 150' of wiring is not going to get you measurable improvments. .

LazMan
Premium Member
join:2003-03-26
Beverly Hills, CA

LazMan to andrew_b

Premium Member

to andrew_b
150' of cable isn't going to make a big difference, assuming what's there is in decent shape.

More importantly; I VERY much doubt the LL/property manager/riser management company would allow you to place (or have a contractor place) new cable for your suite...

jmck
formerly 'shaded'
join:2010-10-02
Ottawa, ON

jmck to AsherN

Member

to AsherN
said by AsherN:

Where on earth did you get 20% improvment? Replacing 150' of wiring is not going to get you measurable improvments. .

you know that VDSL2 means people can be 200-300 feet away (modem to remote). so replacing 30-50% of the cable with a higher quality one and better shielding would make a difference.
InvalidError
join:2008-02-03

InvalidError to xsbell

Member

to xsbell
said by xsbell:

As for the cable, you want lower resistance, not higher, so look to get something that's 250ohm or higher.

Huh? No.

You want uniform cable impedance from end to end because any impedance discontinuity will introduce phase errors, signal reflections and differential noise across the differential pairs; all of which much nastier than whatever he would gain from a minute attenuation reduction.

BTW, typical differential pairs are in the 100-110 ohms impedance range.
InvalidError

InvalidError to jmck

Member

to jmck
said by jmck:

you know that VDSL2 means people can be 200-300 feet away (modem to remote). so replacing 30-50% of the cable with a higher quality one and better shielding would make a difference.

That would be METERS, not feet. If Bell rolled remotes at 100m (300') from clients, they would be in range to start rolling VDSL2 30a for 100+Mbps.

On a good unshielded pair, you can do VDSL2 at 50+Mbps over 400+ meters. Changing 150' of it (~50m) would make almost no measurable difference assuming the existing wiring he's replacing is still in reasonably good shape since he is much too close to the remote to start worrying about attenuation at only ~1/4th the maximum practical distance.
AsherN
Premium Member
join:2010-08-23
Thornhill, ON

AsherN to jmck

Premium Member

to jmck
said by jmck:

said by AsherN:

Where on earth did you get 20% improvment? Replacing 150' of wiring is not going to get you measurable improvments. .

you know that VDSL2 means people can be 200-300 feet away (modem to remote). so replacing 30-50% of the cable with a higher quality one and better shielding would make a difference.

Ah, no.

I'm about 300 METERS away from my remote, on 40 year old buried copper for my Fibe connection. with all 3 TVs streaming HD, I still get full 25/10. Replacing 150' of 25 pair indoor cable is not going to do anything unless that cable is completely shot.
SLAMtech
join:2009-12-03

SLAMtech to andrew_b

Member

to andrew_b
Most apartment risers are fine and you will not see a significance of running your own cat5/6. Although if enough people are on that riser with vdsl it may help out with crosstalk by seperating yourself from the riser. On older building bell will replace defective risers if there a problem or not up to standard, can be tedious and requires access to every apartment on that riser.

The one way to tell if there is a significant amount of crosstalk on your pair is to have a meter to test the wideband frequencies on that pair and psd noise.

The limiting factor in your case is that you are on a stinger and not a 7330.