Managing free pairs to prevent DSL sync. loss
Hello everyone. I am having some very bad time due to my ISP's poor last mile (in India). DSL is loosing sync. consistently and this time problem seems quite interesting so I though to ask how ISPs across world managing it. Problem is high attenuation & low SNR because of "lot of free pairs" in the cable. My connection is coming from something like 100 pair > 50 pair > 20 pair > 5 pair. Now 100 pair has less then 30 active lines but based on testing it seems like at 100 pair DP there's very low noise and everything is pretty good (usual BSNL pillars in India have 100 pair terminations). Next 20 pair has just 4 active lines (and 16 free lines causing issues for those 4 working lines) and at the end my line comes from 20 > 5 with only one (which is my) line active on one of 5 pairs. Now argument of my ISP (BSNL) is that due to excessive number of free pairs, they are causing huge noise and they likely need to reduce these DP's by putting 1-2 line wire from my home till 100 pair pillar termination (which is down in other street and so needs effort in digging and putting new wire). But I just never heard about this problem anywhere else. Do DSL providers really suffer due to free pairs? Assuming other pairs are all crossed/shorted, can they still produce significant noise in other working lines? Also, what exactly was "bonding" used by AT&T in US? I thought it was actually making use of free pairs, bonding them together and having more bandwidth for end user, isn't it? If someone can pass me some detailed whitepaper or document explaining about this noise, it will be very much helpful. Thanks. -- Anurag Bhatia Web: anuragbhatia.com Skype: anuragbhatia.com Linkedin <http://in.linkedin.com/in/anuragbhatia21> | Twitter<https://twitter.com/anurag_bhatia>| Google+ <https://plus.google.com/118280168625121532854>
Hello Anurag. I have not heard of this problem before, but I imagine that the non-terminated pairs could be acting like antennas and picking up noise. Have you considered grounding one end (or both) of the free pairs? Perhaps this would reduce the amount of noise they pick up. Regards, John John Souvestre - New Orleans LA - (504) 454-0899 -----Original Message----- From: Anurag Bhatia [mailto:me@anuragbhatia.com] Sent: Tuesday, July 17, 2012 5:19 am To: NANOG Mailing List Subject: Managing free pairs to prevent DSL sync. loss Hello everyone. I am having some very bad time due to my ISP's poor last mile (in India). DSL is loosing sync. consistently and this time problem seems quite interesting so I though to ask how ISPs across world managing it. Problem is high attenuation & low SNR because of "lot of free pairs" in the cable. My connection is coming from something like 100 pair > 50 pair > 20 pair > 5 pair. Now 100 pair has less then 30 active lines but based on testing it seems like at 100 pair DP there's very low noise and everything is pretty good (usual BSNL pillars in India have 100 pair terminations). Next 20 pair has just 4 active lines (and 16 free lines causing issues for those 4 working lines) and at the end my line comes from 20 > 5 with only one (which is my) line active on one of 5 pairs. Now argument of my ISP (BSNL) is that due to excessive number of free pairs, they are causing huge noise and they likely need to reduce these DP's by putting 1-2 line wire from my home till 100 pair pillar termination (which is down in other street and so needs effort in digging and putting new wire). But I just never heard about this problem anywhere else. Do DSL providers really suffer due to free pairs? Assuming other pairs are all crossed/shorted, can they still produce significant noise in other working lines? Also, what exactly was "bonding" used by AT&T in US? I thought it was actually making use of free pairs, bonding them together and having more bandwidth for end user, isn't it? If someone can pass me some detailed whitepaper or document explaining about this noise, it will be very much helpful. Thanks. -- Anurag Bhatia Web: anuragbhatia.com Skype: anuragbhatia.com Linkedin <http://in.linkedin.com/in/anuragbhatia21> | Twitter<https://twitter.com/anurag_bhatia>| Google+ <https://plus.google.com/118280168625121532854>
On Tue, 17 Jul 2012 09:15:59 -0500, "John Souvestre" said:
Have you considered grounding one end (or both) of the free pairs? Perhaps this would reduce the amount of noise they pick up.
Grounding both ends will probably result in "hilarity ensues". And I suspect that Anurag can't ground the free pairs, because the copper belongs to the provider.
Yeah, grounding both ends will result in some current traversing across the pairs all the time because of differences in ground potential over long-ish distances. Ken Matlock Network Analyst 303-467-4671 matlockk@exempla.org -----Original Message----- From: valdis.kletnieks@vt.edu [mailto:valdis.kletnieks@vt.edu] Sent: Tuesday, July 17, 2012 9:00 AM To: John Souvestre Cc: 'NANOG Mailing List' Subject: Re: Managing free pairs to prevent DSL sync. loss On Tue, 17 Jul 2012 09:15:59 -0500, "John Souvestre" said:
Have you considered grounding one end (or both) of the free pairs? Perhaps this would reduce the amount of noise they pick up.
Grounding both ends will probably result in "hilarity ensues". And I suspect that Anurag can't ground the free pairs, because the copper belongs to the provider. *** SCLHS Confidentiality Notice *** The information contained in this message may be privileged and confidential and protected from disclosure. If the reader of this message is not the intended recipient, or an employee or agent responsible for delivering this message to the intended recipient, you are hereby notified that any other dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify me immediately by replying to the message and deleting it from your computer. Thank you. *** SCLHS Confidentiality Notice ***
Yes, but would this result in more or less noise than an open end acting like an antenna? And would the ground loop noise be in the DSL spectrum? John John Souvestre - New Orleans LA - (504) 454-0899 -----Original Message----- From: Matlock, Kenneth L [mailto:MatlockK@exempla.org] Sent: Tuesday, July 17, 2012 10:14 am To: valdis.kletnieks@vt.edu; John Souvestre Cc: NANOG Mailing List Subject: RE: Managing free pairs to prevent DSL sync. loss Yeah, grounding both ends will result in some current traversing across the pairs all the time because of differences in ground potential over long-ish distances. Ken Matlock Network Analyst 303-467-4671 matlockk@exempla.org -----Original Message----- From: valdis.kletnieks@vt.edu [mailto:valdis.kletnieks@vt.edu] Sent: Tuesday, July 17, 2012 9:00 AM To: John Souvestre Cc: 'NANOG Mailing List' Subject: Re: Managing free pairs to prevent DSL sync. loss On Tue, 17 Jul 2012 09:15:59 -0500, "John Souvestre" said:
Have you considered grounding one end (or both) of the free pairs? Perhaps this would reduce the amount of noise they pick up.
Grounding both ends will probably result in "hilarity ensues". And I suspect that Anurag can't ground the free pairs, because the copper belongs to the provider. *** SCLHS Confidentiality Notice *** The information contained in this message may be privileged and confidential and protected from disclosure. If the reader of this message is not the intended recipient, or an employee or agent responsible for delivering this message to the intended recipient, you are hereby notified that any other dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify me immediately by replying to the message and deleting it from your computer. Thank you. *** SCLHS Confidentiality Notice ***
On Tue, 17 Jul 2012 10:16:17 -0500, "John Souvestre" said:
Yes, but would this result in more or less noise than an open end acting like an antenna? And would the ground loop noise be in the DSL spectrum?
No, it will be strictly a DC current, with the amperage easily calculated from the voltage difference between the two ends and the resistance of however many cable-feet of wire is involved. Not usually a big deal, unless your termination design didn't include the ability to sink a DC current 24/7. (Of course, actually measuring the voltage and resistance may be non-trivial :)
-----Original Message----- From: valdis.kletnieks@vt.edu [mailto:valdis.kletnieks@vt.edu]
No, it will be strictly a DC current, with the amperage easily calculated from the voltage difference between the two ends and the resistance of however many cable-feet of wire is involved. Not usually a big deal, unless your termination design didn't include the ability to sink a DC current 24/7.
(Of course, actually measuring the voltage and resistance may be non-trivial :)
That brings up an interesting question. I assumed the ground potential stays the same between 2 points, but have there been any studies to see if it's actually DC, or if there's an AC component to it? If there's an AC component in the ground at either end (or both) that may introduce EM into adjacent pairs across the cable. And are they more or less than the EM ungrounded pairs would pick up? Ken Matlock Network Analyst 303-467-4671 matlockk@exempla.org *** SCLHS Confidentiality Notice *** The information contained in this message may be privileged and confidential and protected from disclosure. If the reader of this message is not the intended recipient, or an employee or agent responsible for delivering this message to the intended recipient, you are hereby notified that any other dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify me immediately by replying to the message and deleting it from your computer. Thank you. *** SCLHS Confidentiality Notice ***
On Tue, Jul 17, 2012 at 11:16:07AM -0600, Matlock, Kenneth L wrote:
That brings up an interesting question. I assumed the ground potential stays the same between 2 points, but have there been any studies to see if it's actually DC, or if there's an AC component to it?
Thaat's not a safe assumption, since most power companies use earth grounds for their distribution systems. That means that potential between two points, and the current through the ground between those two points, may vary depending on what's happening in the electrically-near parts of the power distribution system. That's not a happy thought, but it is Real Life. It's one of the reasons we went to fiber between widely-separated buildings in our field sites. In my experience, there are AC and DC components both. They're generally -- but not always -- negligible, unless something goes wrong or one end of the line takes a lightning strike, in which case "ground" can rise to bunchty KV.
If there's an AC component in the ground at either end (or both) that may introduce EM into adjacent pairs across the cable. And are they more or less than the EM ungrounded pairs would pick up?
Whatever is picked up by ungrounded pairs should be common-mode -- the same on both wires in the pair. Even if it is induced into the "live" pairs in the bundle, it shouldn't affect signalling. In theory, that is. -- Mike Andrews, W5EGO mikea@mikea.ath.cx Tired old sysadmin
You could "ground" then via some small capacitors. This would block DC and the low frequency power line trash and even act somewhat as a fuse should there be a lightning strike. John John Souvestre - New Orleans LA - (504) 454-0899 -----Original Message----- From: Mike Andrews [mailto:mikea@mikea.ath.cx] Sent: Tuesday, July 17, 2012 12:46 pm To: NANOG Mailing List Subject: Re: Managing free pairs to prevent DSL sync. loss On Tue, Jul 17, 2012 at 11:16:07AM -0600, Matlock, Kenneth L wrote:
That brings up an interesting question. I assumed the ground potential stays the same between 2 points, but have there been any studies to see if it's actually DC, or if there's an AC component to it?
Thaat's not a safe assumption, since most power companies use earth grounds for their distribution systems. That means that potential between two points, and the current through the ground between those two points, may vary depending on what's happening in the electrically-near parts of the power distribution system. That's not a happy thought, but it is Real Life. It's one of the reasons we went to fiber between widely-separated buildings in our field sites. In my experience, there are AC and DC components both. They're generally -- but not always -- negligible, unless something goes wrong or one end of the line takes a lightning strike, in which case "ground" can rise to bunchty KV.
If there's an AC component in the ground at either end (or both) that may introduce EM into adjacent pairs across the cable. And are they more or less than the EM ungrounded pairs would pick up?
Whatever is picked up by ungrounded pairs should be common-mode -- the same on both wires in the pair. Even if it is induced into the "live" pairs in the bundle, it shouldn't affect signalling. In theory, that is. -- Mike Andrews, W5EGO mikea@mikea.ath.cx Tired old sysadmin
participants (5)
-
Anurag Bhatia -
John Souvestre -
Matlock, Kenneth L -
Mike Andrews -
valdis.kletnieks＠vt.edu