Military Appreciation: Exclusive savings for military, veterans, and their families!
Need help connecting your router?
msecher2's profile

New Member

 • 

9 Messages

Mon, Jan 25, 2021 7:04 PM

The infamous 100/20 VDSL2 Lines balancing act; 3 tech visits different sync rates each time(posted below):

Hi All,

I've been trying to get my lines working right for a few weeks now. I've reached a point where local tech's are losing control/power on the issue as it seems to be an under lying issue within the network from the corporate office as I've been told. I've sent an informal complaint to the FCC so I've been in touch with the office of the president in efforts to fix the issue.  Especially as I'm hoping to get access to fiber eventually once made available for FTTP and I wouldn't want this outside line issue that is effecting all my neighbors on 100/20 as well to also be an issue in higher tier speeds. Which the person I am speaking with from OOP is very nice I'm afraid their actual knowledge of difference between speed qualities are lacking, so this issue may not be resolved unless I can get some proper guidance.

Little backstory: installed on a move service mid october no real issues just inconsistent speed tests. Originally my connection started to disconnect so we had a tech come out to service and these were the numbers and this was from first week of January:

 

And here's how the numbers were altered where most errors went away, lines were balanced a little closer after the 2nd tech came(1st tech just called from the box saying everything was ok and it wasn't) This tech explained a problem of the two sources of line coming from outside because these speeds shouldn't be under 40-70mbs on a single line test when the FTTN box is only 200 feet away:


Finally here is after the last tech came just last night, numbers changed slightly errors seem to have gone down further but the balance was further off. Knowing that this is some what "to spec" been told to have SELT tests ran but no tech has actually done so yet. My main issue I'm trying to resolve here is to achieve atleast 20% of the 100/mbs which it's not as I said at times it's under 40 and that is hard wired in on cat6:

Main question(s): Will FTTP connections be effected by this outside line issue? And is it even possible to achieve within 20% range as promised in my contract or am I just wasting my time with this outdated technology? As tmobile 4g LGE towers are within full strength within my location, and they offer home internet, and I can get 110 single line test from them down as it seems like the speed option with the most integrity in the area currently. So I'm strongly considering this route as it is the most competitive option available. That is if ATT's 100/20 is not achievable on a single line test within the range of 20%. Please advise and let me know if you need anymore screenshots; speed tests, pings, traceroutes, or etc. I truly hope to be able resolve this and keep ATT as my main home internet provider. Thanks in advance

Best,
-m   

Accepted Solution

Official Solution

my thoughts

Employee

 • 

19K Messages

Il y a 1 a

Your bonded pair profile (user rate) is 110/22 ... perfectly balanced would be 55/11 per line...

On January 11 lines are 57.2/11 and 52.8/11 within less than +-5% of perfect

On next visit lines are 58.6/11.3 and 51.4/10.7 which is about +-6.5% of perfect.

Both are fully passable at+-10% and within all acceptable standards. 

Neither line will offer 100, the max rates will not support thus the bonded of sending half signal down each line and being combined at the gateway.

Do you have the Smart Home Manager app? What do you see when running the internal speed test on the app? This is the gateway connection speed to the ATT server staying on network...

If using your own device(s) testing hardwired to ATT server within 100 miles using speedtest.net with no other devices connected or in use?

Is your device capable of these speeds, there are many devices out there that are not, limited to 75 or less. 

From a tech viewpoint, the previous techs may have replaced the jumpers at the VRAD, performed a port bounce or even ports swap...

Would not know unless stated in their closing reports.

Most techs would run a SELT test as part of troubleshooting, you would not know if they did or did not unless saw the test... however the company places more emphasis on the JDSU TRUE TEST that is required for passing results in order to close the job. If the TRUE TEST is passing there is no line issues between the VRAD and the test point which could be terminal, NID or gateway jack. 

Single line test? referring to speed test... as bonded pair would not have single line testing only multi line (2 lines for bonded) that combined would produce the profile and speed tier. Will not see anywhere near 100 on single line as speed would not exceed profile which is less than 60/11 depending on the line. 

The only question I would have is on the protocol being 17a instead of 8d... considering upload max rate is greater than 20 on each line I would say this is 17a. 

Accepted Solution

my thoughts

Employee

 • 

19K Messages

Il y a 1 a

@Constructive internet 75/20 and internet 100/20 are only provided on BONDED pairs using K or N cards with 17a protocol. 

Internet 100/20 requires copper distance to be under 1000 feet

Internet 75/20 requires copper distance to be between 1800 to 1000 feet.

In the original bonded pair days (2011-2013) these were iNIDs (2WIRE 3812) over 3000 feet to provide internet 18/1.5 and UVERSE IPTV service where previous best speed would have been single pair 12/1 using 3801 out to about 4000 feet. 

With release of the 589, the company started offering bonded pair installs at shorter distance, initial offers were for internet 45/6 (POWER) on F cards using protocol 8d. As newer cards made it to the field speeds increased to current 50/10, 75/20, and 100/20 using 17a. If the company ever went to protocol 30 and/or vectoring could see 200 download at shorter distances. However with FTTP releases since 2016 I doubt the FTTN VRADs will see upgrade to vectoring. 

Constructive

Employee

 • 

27.3K Messages

Il y a 1 a

fiber is a totally different monster it has nothing to do with your copper lines. but im wondering why you even have a bonded pair at less than 200 feet from the box. we were never allowed to install bonded pairs for anything closer than 5 times the distance you are.

Constructive

Employee

 • 

27.3K Messages

Il y a 1 a

yeah i installed from 2008-2011

New Member

 • 

9 Messages

Il y a 1 a

@Constructive I hear you about fiber having nothing to do directly with my copper lines however fiber is involved in my setup as the system is FTTN and then copper to the property. So in theory if there is something wrong with the fiber lines to the node I imagine there could be something affected by issues in the fiber lines, no?

@my thoughts Are you implying that when you say that "if the company ever went to protocol 30 and/or vectoring could see 200 download at shorter distances." that it currently would not be able to achieve 100/mbs on a single line test, due to non upgraded vrad for vectoring, while it would be able to sometimes hit 100 on multiline test as it currently stands?

(edited)

JefferMC

ACE - Expert

 • 

28.4K Messages

Il y a 1 a

Those error counts you see are totally artifacts of the VDSL2 protocol over copper, only measured between your Gateway and the VRAD.  Any errors that may occur on the Fiber (FTTN) would not show up there at all.  When you have FTTP, you either have an ONT or a module that takes care of the signaling and hides whatever error information may exist on the fiber behind the interface to the Gateway.

New Member

 • 

9 Messages

Il y a 1 a

@JefferMC 10-4 makes logical sense. Which makes it more obvious that fiber deployment is one of the best tools to quiet the bandwidth complainers like myself. My main concern within the numbers is the balancing of sync rates and why they can't gettem to match after 3 tries, second to that were the errors being produced. Thanks!

@my thoughts Thank you for further clarifying what is capable with the VDSL2 lines as far as single threaded power. Where two lines combined in theory should equal one, as that is the definition of bonded, but this is likely just a limitation of the product not working with vectoring, which is an interesting concept in it’s self. Obviously i'm no engineer just a wee ol' tech tester on my trusty geek boxes remoting into my office for work 5 days a week. And I think I knew it from the start instinctually, about the limitations that were to come on my current plan, but combo of 100mb/s blast being needed for my work and when I'm paying for a speed advertised and it's not being met makes me feel like theres a mouse eating the cheese somewhere if I can't get a constant 80 multitest in peak hours even. But it makes me think sometimes that if my lines were perfect balanced at 55 each maybe I would achieve a more steady flow closer to what I need to function more properly since it is a marginal difference that is needed to achieve to have a perfect connection for what I do. So to compensate I try to trouble shoot to make it better as much to my power so I can learn and understand it better especially if it needs to be serviced frequently as it seems. I also have this hunch because I've researched my area for ATT BUSINESS Internet service and what is offered at my location is only the 75/20, not 100/20 this may be an incorrect theory but I think they don't offer the 100/20 business in my area is because they can't guarantee it as close as they can the 75 in my exact location with how the FTTN is setup. Business also claims they will offer fiber to my premise but I’m reluctant to go that route because I’ve had some bad luck with what ATT cs speed promises they claim that can achieve if it’s not fiber and it’s not 100% certain in writing come the day of the install. And then I’m left yelling at CS like what the heck did I get strung along for meanwhile the tech is running copper when I had fiber promised.

Based on your last point, and correct me if I'm wrong, do you have a feeling that it could be the uncertain protocols causing the errors on the upload speeds?

And great question and glad you asked about the smart home manager. I use it religiously. Useful tool. for qos and that home speed test which measures a multithread test on how much is coming into the house, it almost always stated 108 to the house and hardwired on my most updated device I would get 100 at times and even 103 on multithreaded but definitely fluctuation present. ATT corp server is 2 miles away I test it with local networks always and sometimes a bit away to see how it reacts to bottlenecks. I've tested multiple devices with it each from a different year to see the fluctuations in speeds at different times of the day. since these last two tech visits(not the last third that got me to -6.5. I haven't been able to get my router to work with the smart home manager test specifically, when I did the swap it transfered everything then said call CS at the end I called and they didn't have more visibility on the issue than me. Reason being is I tried to swap out the modem recently which I have the old one it didn't make a difference really except now my smart home manager speed test doesn't work. If you have any info on how to make it go back I'd appreciate it, can you let me know? If there's nothing else I could unless have a tech come configure I may just swap the router I exchanged back to see if it clears it so I can use it correctly again, I may give it a shot. But TBH I’m not that stoked on the smart manager test for myself as it’s probably a better gage for cs and legal… I trust speedtest.net and use a single threaded test because it's the most thruthful. Again, multispeed tests are kinda misleading especially when not all servers will utilize connections the same. Last night I was getting 85-90 on a single test. And 100 on multi. And right now i'm actually getting 75-91. Plan to test later in the afternoon again when it bottlenecks and record data. I can usually only achieve these numbers in the late or early hours and only from a hardline with top NIC card but I don't think new technology of devices should be relied on to balance out dated technology, this in it's self is a flaw. So combination of some throttling and lines that don’t match and fluctuating speeds throughout the day averaging below I’m wishing I still had a little more to consider my rate truly consistently competitive unfortunately.

An actual dream of mine is that I could actually have a tech that wouldn’t be so shady to share things like a SELT test. I know that my last techs likely had passed the TRUE test because they all signed off and closed out their tickets real quick lol. It’s the local tech’s in my area involved in some kind of politics that don’t seem to want to make waves and call out issues on the lines to that go out pretty frequently and cause a proverbial backfire from the main office on themselves. If you were on my next-door app and heard the amount of customers complaining (and not just for ATT)during outages I’ve never witnesses so many internet outages in any other part of the city I’ve lived in.. But likely the local techs are staying quiet to keep their jobs secure especially during these troubling times. Which I understand fully. But somethings gotta give.

All the other stuff aside I still would like to stick it out with ATT, as the tech’s here actually have an empowering community who enjoy what they do and I appreciate that. I have a lot of un answered questions from my account still of being directly connected to my account manager, and I’m left to feel like Dorthy searching for the Wizard of Oz. If I could talk to them I could then explain how bringing in fiber to the block would turn all of us complainers about losing 20+40 mbs that are forcing the company to spend money on loaner out of town repair techs. Why keep having to waste tech's time to rebalance some lines when you can just shut the custies up and greenlight neighborhoods like mine with FTTP where my neighbors are forced to have multiple internet providers to deal with local bottlenecks. Because it’s basically here, across the street less that 400 feet away. I’ve had CS tell me they would have a fiber line ran from that address only to have someone else shut it down etc. ATT Business even said they could make it happen, but the lack of sales integrity w speed promised for my current situation doesn’t make me want to go that route currently. Meant to say this earlier but didn’t want to overload with info from the start my original issue stemmed from a move order of my original fiber that went through and then on the day of the order denied and switched to copper. And to read all the news articles about how att stopped installing DSL in october and that was when my copper was ran when an order to move fiber was processed it’s frustrating overall. I’m sure the tech’s feel it in one way or another as well. Cheers

-m

(edited)

my thoughts

Employee

 • 

19K Messages

Il y a 1 a

 And to read all the news articles about how att stopped installing DSL in october and that was when my copper was ran when an order to move fiber was processed it’s frustrating overall...

That was for LEGACY DSL, best speed of 6 with 150G data cap....

UVERSE based ADSL2+ and VDSL FTTN deployed from 2006 to end 2015 is still being installed providing internet access to over 46 million addresses, more than 3x the number of addresses that can currently receive FTTP direct fiber installations or 75% of all hardwired internet addresses are UVERSE based ADSL2+ / VDSl with less than 25% being FTTP.

I expect the FTTN types of installs to be continued to offer this entire decade. The FTTN had 33 million addresses at end of 2015 which was about 44% of all addresses in then footprint... took 10 years to reach less than 44% of addresses. I do no believe you will see FTTP reach the same numbers within ten years which would be 2016 to end 2025, maybe 28+ million (?) or about 35% of then estimated addresses within 21 state footprint. Conclusion is end of 2025, 65% of addresses will still have either fiber/copper based internet or be using fiber to the cell tower as 5G with a small percentage (rural, semi rural) using 5G mmwave from electrical poles known as Project Airgig.

Just my opinion... but the future looks more wireless fiber service than it does hardwired fiber. If desire hardwired fiber need to consider moving to an address where it is currently offered... 14+ million with 25% of those (3.5 million) being apartment / condo complexes.

(edited)

New Member

 • 

9 Messages

Il y a 1 a

but the future looks more wireless fiber service than it does hardwired fiber. If desire hardwired fiber need to consider moving to an address where it is currently offered... 14+ million with 25% of those (3.5 million) being apartment / condo complexes.


I can feel it coming for

sure.. I’m still traditional in a sense so

id be most likely going w a physically wired service just because atleast to get through the first few generations of air fiber..

I’ve been wondering about living right next to a multilevel apartment building that has the same 100/20 if they get fttp Can I piggy back somehow since we likely share the same vrad? There’s like one apartment complex on each side from east to west and they’re greenlit and i know Im light green had techs saying it’d be done about 1st quarter not holdin my breath tho. I’m at an apartment that has two units would they ever consider going that route like homes that were converted to multi unit homes as well?

update: after testing again in the higher use times my speed test this afternoon to eod was 100-90 multi test and it was actually 80-90 single speed for once 🤷🏼‍♂️

(edited)

New Member

 • 

9 Messages

Il y a 1 a

..And another update: Been testing only with local att server on speedtest.net all week and I noticed throughout the week as Wednesday afternoon arrived my speed started to decrease on the multi test from which is now around 80 while upload is always 20 single or multi and on single download it's giving a solid 64. I do notice when I use the local frontier server I'm getting about 18 more mb on Multi and 23 more on a single threaded test. Keep in mind I'm using a gaming computer hardwired for these tests which has an up to date nic card. Makes me wonder when connected for a speed test why Frontier's server has a stronger flow than ATT's especially at peak hours during peak days of the weak. Speed tests to my gateway have been 105 since the last tech no longer 108 while it's seeming to perform a little better. I'm on the hunt for a tech to see if they could actually get the balance perfect 55/11. Thanks!

JefferMC

ACE - Expert

 • 

28.4K Messages

Il y a 1 a

Trying to keep a perfect 55 is likely a fools errand.  Look at this graph of my VDSL2 Max:

It varies all the time.  Doesn't even need a reboot, just retrains and comes up with a new rate.  No, the bump ups (like the very nice one this Tuesday) don't reflect a service call and someone fixing things up... I didn't even notice the retrain.

New Member

 • 

9 Messages

Il y a 1 a

@JefferMC Interesting. how are you pulling this graph? Not to pass judgement but If mine steady fell well below 80 like it did for you in may and april CS would hear it from me for sure..

I've just noticed on my vpn that I'm using to connect my work computer that's on a network that has about 100 other users on it currently. At times there are bottle necks with when important deliveries are going on with it so every little bit counts for me. Since balancing has been working better my lag is seeming less through the rdc connect. Sure speed tests will always vary but if there's bottlenecks in the stream it will be seen in a delay on my VPN network immediately by allowing less mb's to go through so I am unable to quality control my renders.

(edited)

Constructive

Employee

 • 

27.3K Messages

Il y a 1 a

I’ve found my speeds are 1/3 of regular speeds when connected to my vpn

New Member

 • 

9 Messages

Il y a 1 a

@Constructive do you know what kind of connection the vpn you have that you’re connecting to? Mine is a 1gig/sec connect. Users are allowed about 10mb/sec each roughly and with my vdsl connection I get in about 8.5m/sec max on a good hour

(edited)

Constructive

Employee

 • 

27.3K Messages

Il y a 1 a

I get 200 mb off of vpn when I connect to my At&T Cisco vpn I get somewhere around 20 

Need help?
New to the AT&T Community? Start by visiting the Community How-To.
New to the AT&T Community?
Visit the Community How-To.