Had problems since before Xmas with Broadband connected but random no internet. Anyone help p...

[#1]

@Gandalf 

The sync speed graph on post #134 is the same before and after the router restart albeit the SNRM is different.  Does that suggest that the line is banded?  The resultant SNRM is not that of a recognisable 3dB interval target.

@joseph1 

You need to look under HELPDESK (IIRC) and therein there should be the router's log file.  Look for a record around the time of the resync - note that the router log is GMT not BST.

[#2]

---

@joseph1 wrote:

Yes, that line loss is excessive.

Info here suggests a lot lower would be expected:

---

Well that depends.  The length is the electrical length, assuming the whole circuit is copper of a given gauge.  Change the gauge or material (eg. aluminium) and the electrical length can quickly become different to the physical length.  Also note that the cable route might not be the most obvious which you measure on a map.

[#3]

@Gandalf 

Here is the log for reset first,

11:57:25, 14 Jun.	( 96.940000) PPPoA is up -​ VPI: 0, VCI: 38 Down Rate=7199Kbps, Up Rate=828Kbps; SNR Margin Down=9.7dB, Up=9.0dB
11:57:25, 14 Jun.	( 96.920000) CHAP authentication successful

then the first drop of UL

14:18:21, 14 Jun.

OUT: BLOCK [7] ICMP replay (ICMP type 3 code 1 81.174.250.84-​>52.97.212.114 on ppp0)

second drop UL

16:57:47, 14 Jun.

OUT: BLOCK [7] ICMP replay (ICMP type 3 code 1 81.174.250.84-​>52.97.146.162 on ppp0)

Route I followed is that of cabinets. There is one river bridge in direct line so assumed that is teh crossing point of cable.

Just done the PN speed test and 6M DL  0.6 UL

Thanks, Rob

[#4]

@Townman 

Regarding line loss. From my occasional records of stats page of 2704N router:

About a year ago OR were down our road and a little way back towards exchange replacing cables. He was over a day outside our house in the manhole with huge bundles of wires and chatting to his mate at the next cover along. I know the last section, was installed as aluminium when the houses were built in1976. I can only assume the aluminium was replaced with copper on this last leg.

In May 2019, one log:

26 May after hols

6 - Uptime:	13 days 18:43:42
7 - Data rate:	952/8143
8 - Maximum data rate:	1364/8360
9 - Noise margin:	5.9/6.3
10 - Line attenuation:	22.4/41.0
12 - Data sent/received:	1.6G/2.1G

A year later:

12 May 2020 after bt work completed in street

6 - Uptime:	0 days 03:10:16
7 - Data rate:	968/8179
8 - Maximum data rate:	1376/8080
9 - Noise margin:	6.1/5.6
10 - Line attenuation:	22.4/40.5
12 - Data sent/received:	20.6M/395.9M

Didn't make much difference to line attenuation!

Thanks, Rob

[#5]

@joseph1 

Sight of those stats is useful. thank you.

I have re-read the whole of this topic and to some extent feel that collectively we might have lost sight of the problem we seek to remedy.

• Earlier this year you were seeing over 9mbps at 6dB but reported buffering issues
• Since then you've had changes made
  • Line L&S
  • Replacement router
  • SNRM fixed to 9dB
• Monitoring shows some indication of REIN
• Current stats report circa 8mbps at 9dB

What is the problem which remains to be fixed?

Do you now still see buffering on the TV?

It might be the case that your line is what it is - That'll Do - the unlocated REIN drives the need for the line to be set at 9dB, with the slight reduction in speed to maintain connectivity during REIN events and reduce retransmission rates which might be contributors to buffering.

@Gandalf might we have your view here please?  Can you please give an answer to the specific question of AC balance, which might make a line more susceptible to REIN?  Was a hot VLAN eliminated as a potential contributor to buffering?  Would a L&S have delivered a change of SVLAN?

[#6]

@Townman 

Earlier this year it was constant random dropouts, usually overnight, this triggered the OR visit and work was then carried out swapping circuits over back to exchange.

More recently, since all the dropout problems were all now virtually eliminated, buffering was noticed and at times, the Internet led on 2704N flickered rapidly and this coincided with web pages failure to load straight away, (LAN and wifi) for sometimes a minute or more. This could go on several minutes, day or night.

The new router was then provided and since then have not seen buffering or noticed web page loading problems.

Apart from any speed changes made by PN since Xmas, my speed and noise margin are unchanged (slightly lower though).

To sum up, don't know if old router was cause of problem or rein or even if fault is still around?

Will monitor one last night then give it a try over a few days.

Been out today, and now, at top of 2 poles down our road some metal plates have been fitted, each with 4 rectangular slots. Plates look around 350mm High x 150mm Wide. Any ideas what they are for?

Thanks, Rob

[#7]

No idea.

Photos might help please.

[#8]

@Townman 

https://lightspeedbroadband.co.uk/

Same price as my adsl? NOT!

Rob

[#9]

@Townman @Gandalf 

Been looking up what ac balance is.

Seems it can have a huge effect on adsl susceptibility to REIN.

Considering all the faults and repairs done these past 10 years or more, it would be a good idea to find out the actual figure and maybe eliminate that as an effect assisting REIN. The switch over to a different pair by OR in March, might that not have been as good balance pair even?? Or even a redundant pair due to a previous problem?

Another unknown at the moment.

Rob

[#10]

Sorry for the delayed staff response here.

@Townman wrote:

The sync speed graph on post #134 is the same before and after the router restart albeit the SNRM is different.  Does that suggest that the line is banded?  The resultant SNRM is not that of a recognisable 3dB interval target.

No the line's not banded: 

"WBC 160K - 24M Medium delay (INP 1) 9dB Downstream, UC Medium delay (INP 2) 9dB Upstream (ADSL2+)"

@Townman wrote:

@Gandalf Can you please give an answer to the specific question of AC balance, which might make a line more susceptible to REIN?  

Our tests don't really show this specific detail as far as I'm aware.

@Townman wrote:

Was a hot VLAN eliminated as a potential contributor to buffering?  Would a L&S have delivered a change of SVLAN?

There are no signs of a hot S-VLAN, and a lift and shift shouldn't change the S-VLAN as all it'd do is move the line onto a different port on the same equipment at the exchange. 

@joseph1 wrote:

The new router was then provided and since then have not seen buffering or noticed web page loading problems.

Apart from any speed changes made by PN since Xmas, my speed and noise margin are unchanged (slightly lower though).

To sum up, don't know if old router was cause of problem or rein or even if fault is still around?

Will monitor one last night then give it a try over a few days.

Looking very promising! Let us know if there are further issues.

[#11]

---

@Gandalf wrote:

No the line's not banded: 

"WBC 160K - 24M Medium delay (INP 1) 9dB Downstream, UC Medium delay (INP 2) 9dB Upstream (ADSL2+)"

Anoush,

Sorry for "slugging away" at this one but your findings seem at odds with what is being observed.  If the line is not banded, then why do we see that through a resync...

1. The line comes back with a SNRM higher than the target?
2. The sync speed before and after the resync is identical?

Normally the observation of the SNRM being higher than the target immediately after a resync is indicative of the the DLM restricting the sync speed or the user having changed the SNRM target settings in the router, rather than allowing the router to negotiate best sync speed at the target SNRM.  I note that this user is using the PN Hub One so we can eliminate user tweaking of line metrics as some other routers facilitate.

---

@Gandalf wrote:

@Townman wrote:

@Gandalf Can you please give an answer to the specific question of AC balance, which might make a line more susceptible to REIN?  

Our tests don't really show this specific detail as far as I'm aware.

---

There are interesting observations on the BT forum - https://community.bt.com/t5/Archive-Staging/AC-line-balance/m-p/676840#M311853

"AC balance can affect Broadband. characteristics are intermittent sync which occurs randomly. AC balance is measured in dB and it is the ability that the copper pair has to resist an Induced signal and its ability to hold a steady DSL connection. if a line has had a line lest and there are no conditions on the line i.e. No battery or earth contact then the AC balance should be the deciding factor (excluding line length) a poor AC is <50dB a good AC is >60dB"

This is an interesting (long) read - AC Balance issue... (kitz.co.uk)

AC Balance is also known as longitude balance.  Do none of the ISP (or CP) tests report the Ring / Tip to earth electrical characteristics?  See Longitudinal Balance of a Copper Cable Pair and Its Effect on Bandwidth | ISEMAG

"Longitudinal balance is accomplished with good pair balance on a vacant pair. Pair balance is broken down to 2 parts: capacitive balance, and resistive balance. ... "

Do none of the tests indicate anything like these measures?

[#12]

@Townman @Gandalf 

Thanks Townman for finding those items. I had read the BT community one last night.

So, as I understand it. IF the REIN is caused by a MW transmission or any other electrical noise situation, then, put simply, the wire pair to my house can act as an aerial to pick it up if it lacks good balance. It seems above 60dB is considered ok and 70dB excellent.

What good balance will achieve is any interfering signal that is induced in both wires simultaneously, they will pretty well cancel each other out. However, if a poor balance, they do not cancel out as well, may even add and can result in a stronger interfering signal.

Good ac balance is essential to correct operation and cannot be assumed to be ok if it is not actually measured.

Thanks, Rob

[#13]

@Gandalf 

See attached. U/L has dropped several times last 2 days. Look at this latest!

Something is just not right!

Rob

[#14]

@Townman wrote:
If the line is not banded, then why do we see that through a resync...
The line comes back with a SNRM higher than the target?

Possibly some interference on the line caused the SNRM to fluctuate?

@Townman wrote:
The sync speed before and after the resync is identical?

It'd just be sync'ing at that speed, just because it sync's at the same speed doesn't mean the line's banded.

"AC balance can affect Broadband. characteristics are intermittent sync which occurs randomly. AC balance is measured in dB and it is the ability that the copper pair has to resist an Induced signal and its ability to hold a steady DSL connection. if a line has had a line lest and there are no conditions on the line i.e. No battery or earth contact then the AC balance should be the deciding factor (excluding line length) a poor AC is <50dB a good AC is >60dB"

This is an interesting (long) read - AC Balance issue... (kitz.co.uk)

AC Balance is also known as longitude balance.  Do none of the ISP (or CP) tests report the Ring / Tip to earth electrical characteristics?  See Longitudinal Balance of a Copper Cable Pair and Its Effect on Bandwidth | ISEMAG

"Longitudinal balance is accomplished with good pair balance on a vacant pair. Pair balance is broken down to 2 parts: capacitive balance, and resistive balance. ... "

Do none of the tests indicate anything like these measures?

Does this help?

Copper Line Test
Circuit ID:	[redacted]	Service ID:	[redacted]
Telephone NO.:	NA	Test Executed On:	18-06-2021 08:24:40

Status:	Pass	MFL:	OK	OR Test ID:	[redacted]
Test Outcome:	CIDT LINE TEST OK - ALL SOCKETS UNPLUGGED			DTR:	T302

Copper Test Details
	A to E	B to E
Capacitance:	135 NanoFarad	137 NanoFarad
DP Line Length Estimate:	2011 Metres	DN Line Length Estimate:	2029 Metres
Celerity:	32.3 dB	Line Loss:	33.84 dB
Line Stability:	Stable
Fault Report Advised:	N
Service Level:	2
BRAGOutcome:	Bad
Faceplate:	Not Detected

@joseph1 wrote:

@Gandalf 

See attached. U/L has dropped several times last 2 days. Look at this latest!

Something is just not right!

Rob

Your broadband's remained stable though and not dropped in 3 days. You've just had a drop now as I've been running an intrusive line test to see whether we can see the AC Balance^

Are you still experiencing real world performance problems?

If so, I think our options is to book another engineer or look at upgrading you to fibre on a good deal. The problems may carry across to fibre but generally, it's easier to fix a problem on fibre as there are more diagnostics available.

[#15]

@Gandalf @Townman 

Thanks Gandalf. You will see the change (what did it?) on attached RS shot.

Your test shows a difference in line length estimate. Is that a good result?

Just looked for BRAG info and found this:

https://www.ofcom.org.uk/__data/assets/pdf_file/0037/186688/openreach-broadband-experience.pdf

Page 1  "The lines classified as “Very Bad”and “Bad”have a high likelihood of seeing a speed improvement following an engineering visit, whereas the “Very Good”and “Excellent”lines are unlikely to see a speed improvement following an engineering visit."

Would you agree about this as a basis for another visit? In the meantime, I will try and get in touch with the original guy who did all the work on my line and get his opinion if I can.

If at around 0825, time of your test, which is well outside the times of the suspected MW REIN issue and a time which shows less disturbances on RS, the BRAG test shows Bad, could that mean Very Bad in the night? I appreciate you may not know the answer and I don't expect another test at say 0200.

My wife did not use TV last night but did the night before and had 2 or 3 buffering issues.

Regarding dropouts, as there have been none I don't know if this current router would reconnect every time automatically. The old 2704N would not and that was the big problem.

Following your test today, the UL & DL speed is now increased.  Current speeds are UL 828kBps and DL 7300kBps.  Following this disconnection, is my router likely to try and adjust to a different speed either way?

Regarding my changing to fibre, I have not seen any deals of interest cost wise and with the unknown result regarding the noise issue I could have similar issues but be paying a lot more for the privilege. I doubt there is a free trial available. When my adsl is working ok I have no need for fibre speed as it has always done all that I want.

I appreciate your time and effort in helping.

Many thanks,

Rob