New ADSL Install - Sim Provide - Line Changes

Hi All,

Yesterday had a simultaneous provide from MPF LLU which could've been a ton better; landline down for 10 hours, broadband could've been active all day when I was trying to work but Plusnet don't activate username on RADIUS until BT Wholesale give the ok.

Since this provide the line characteristics seem to have changed.

Obviously we're under DLM still however there is a clear issue. When the land line phone is in use the SNR margin drops from 12dB to 4dB upstream, and from 6dB to 5dB downstream.

This is a long line however this did not happen before the move.

There is a couple of dB drop in attenuation which seems to be due to the BT Wholesale MSAN using ADSL 2 rather than ADSL2+, downstream sync has dropped from 2,020kb/s to 1,767kbps and upstream from 605kbps to 443kbps.

Land line usage bleeding into the ADSL concerns me. It's not an issue on my side, the same Openreach filtered NTE is in place, which leads me to think a problem was caused at the exchange when the line was rejumpered.

Of course just to complicate things it's also possible that the fault was induced by work in the street cabinet as new lines are being provisioned but I struggle to image a fault that would cause this and am far more inclined to blame a dodgy SMPF splitter port in the exchange.

This seem a reasonable thought process?

Here are line stats. The low up time is from a loss of sync when the land line was used.

My next step is to dig out a router with the best ADSL/2/2+ chipset I have ever seen, the Argon AR432, and hope it likes the Infineon chipset in the BT Wholesale MSAN.

Along of course with getting the line itself fixed.

Uptime: 0 days, 1:23:28
DSL Type: ITU-T G.992.3
Bandwidth (Up/Down) [kbps/kbps]: 443 / 1.767
Data Transferred (Sent/Received) [MB/MB]: 34,45 / 33,98
Output Power (Up/Down) [dBm]: 12,8 / 0,0
Line Attenuation (Up/Down) [dB]: 39,3 / 68,5
SN Margin (Up/Down) [dB]: 12,0 / 6,1
System Vendor ID (Local/Remote): TMMB / ----
Chipset Vendor ID (Local/Remote): BDCM / IFTN
Loss of Framing (Local/Remote): 0 / 0
Loss of Signal (Local/Remote): 0 / 0
Loss of Power (Local/Remote): 0 / 0
Loss of Link (Remote): -
Error Seconds (Local/Remote): 181 / 0
FEC Errors (Up/Down): 0 / 0
CRC Errors (Up/Down): 15 / 318
HEC Errors (Up/Down): 6 / 393

FEC Errors (Up/Down): 0 / 0 - is it on Fastpath ?

Does sound a bit like a flaky joint on the jumpering, unless a downpour has infiltrated overnight.

In reply to a post by yarwell:

FEC Errors (Up/Down): 0 / 0 - is it on Fastpath ?

Does sound a bit like a flaky joint on the jumpering, unless a downpour has infiltrated overnight.

Yes Phil, I presume DLM hasn't seen fit to switch to an interleaved / INP profile.

This was happening yesterday as well before the rain came to visit, I was just busy with other things.

opting into interleave on such a long line is common advice.

I am sure that DLM will be interleaving the line in time...


Uptime: 0 days, 2:47:34

Line Attenuation (Up/Down) [dB]: 39,3 / 68,5
SN Margin (Up/Down) [dB]: 12,1 / 6,0

Error Seconds (Local/Remote): 301 / 0

CRC Errors (Up/Down): 32 / 506
HEC Errors (Up/Down): 12 / 620

A picture is worth a thousand words.

Uptick in Bitswaps too.

As a sideline Plusnet's order support was horrendous. The last person spoken to was a robot who kept talking over the account holder and interrupting her to recite the same information over and over again.

She stubbornly refused to pass her to a manager which was just bizarre. Are the Plusnet CS team on some kind of incentive / disincentive programme to avoid escalating issues?

Said manager once she finally relented and got him on the line was able to do exactly what we asked for in 5 minutes flat.

Really incredibly unimpressed with the whole deal so far.

Edited by deleted (Tue 23-Jul-13 14:30:09)

I would opt-in to interleaving in the hope of keeping the SNR margin at 6.

DLM is already there:

Channel Mode: Interleaved

Zero FEC still. Correcting nothing.

In reply to a post by Ignitionnet:

As a sideline Plusnet's order support was horrendous. The last person spoken to was a robot who kept talking over the account holder and interrupting her to recite the same information over and over again.

You're not alone. That was an experience I had recently with the support team - on three separate occasions.

The problem today is there is too much talking and not enough listening!

A result of mobile phones?

No idea.

Recommendation on the Plusnet forum is to plug directly into the test socket, presumably on the off chance that the passive face plate went bad at exactly the same time as a guy in the exchange was rejumpering the line.

I think it's fair to say it's not the face plate.

is there not a small chance the HR fault is actually in the faceplate, hence the suggestion to rule it out?

Sounds like you had a relatively smooth transition compared to the "Right Pig's Ear" we experienced!

We migrated from a superb TalkTalk LLU ADSL2+ line to Plus net, for its FTTC service. There were ongoing problems from the outset, mostly related to the ordering process.

So far, the transfer has taken five weeks, and still no sign of fibre broadband. Still waiting for an engineer's appointment to connect us at the fibre cabinet. Apparently no appointments are available for another three weeks, at the earliest.

Our cherished phone number of ten years has been lost in the transfer process, although it was originally promised by Plus Net that we would keep it.

We've been billed £50 for a "new line" when we were told the transfer would be free. Plus Net made the flaky promise that the £50 would be refunded after we had transferred. A promise which no one seems to have recorded.

We specifically asked for an afternoon booking from an Openreach engineer ( to "trace the pair"), only for it to be booked for the morning. That had to be cancelled and a new afternoon appointment made. Nevertheless, the engineer arrived in the morning! It was a Quinns contractor who'd travelled 50 miles and wanted to get his jobs finished by lunchtime.

Instead of migrating us directly to Plus Net FTTC, we were told we would be migrated first to Plus Net ADSL "so that we can get you on our system". We were promised a seamless transfer of broadband provision. But in the event, broadband was down for eight days. Thank God for 3G.

So far Plus Net ADSL has been very poor, and the much vaunted support has been substandard, too. We were getting 17Mbps with TalkTalk. That's now been slashed to just 7Mbps with Plus net. No one from plus net seem to know how to resolve that and provide us with a similar bandwidth that we got with TalkTalk. For reasons unknown, upstream is routinely capped by Plus net, too.

After two hours waiting in phone queues to report the broadband fault ("we are currently experiencing high call volumes") we were fobbed off with the claim that "7Mbps is all that your line can support!"

We are being billed for Caller Display, but that wasn't provisioned properly by Plus net either ("it's recorded in our front office system, but for some reason the order didn't pass to back office")

And when it finally came to upgrading to Plus Net FTTC from ADSL, the half price deal for 6 months had finished!

We requested Plus Net to honour that half price deal (since they delayed the provision by migrating us first to ADSL). But they would agreed only if we took out a 24 month contract rather than the standard 18 months.

Checking the support tickets, it seems the next problem is that Plus net has requested a 40/10 service, rather than 80/20.

To be honest, not impressed. With the benefit of hindsight, we would definitely have stayed with TalkTalk.

Edited by deleted (Thu 25-Jul-13 13:57:35)

On the ADSL speeds, what are the stats from the router, i.e. actual connection speed, attenuation and noise margin. They are the real indicator of what a line can do rather than the guess-estimate system

In reply to a post by MrSaffron:

is there not a small chance the HR fault is actually in the faceplate, hence the suggestion to rule it out?

Very small given the face plate worked perfectly for 8 months and if it is the issue simultaneously went faulty at the same exact time as the line was moved from Sky MPF to BT Wholesale.

The voice service was down for over 9 hours, it seems reasonable to think that there were some issues with the provisioning.

Probability heads in that direction but have learnt over the years coincidence is a real devil

I prefer Occam's Razor and experience has shown it to be far more reliable. It would seem a bizarre way of troubleshooting indeed to eliminate the virtually impossible while leaving by far the most likely course of events untested.

can I ask why the change to plusnet?

I was going to say typical HR fault, but given what you said we have to look at whats changed which is going to be the line card.

Nothing to do with my face plate or anything else local to me, was an issue in the exchange.

In their infinite wisdom BT Wholesale left a 5.5km line on 3dB fastpath. Packet loss and drops ahoy.

Not that it matters, 6dB interleave has left the packet loss there and failed to stop 70% of the errors from being uncorrectable.

Now it's jumping through Plusnet's hoops to get an SFI to come to the property. Risky given that Plusnet technical support don't even seem to know what the actual issue is, and annoying given the line was loss free until work was done in the exchange, specifically the move to Plusnet, but it strengthens the case to get the year's line rental in advance back.

BT Wholesale are a nightmare, an absolute nightmare that seems to exist purely to ensure faults take longer to get fixed, Plusnet's support is script driven and robotic and they don't pay attention to the fault being reported.

27 days into the Plusnet experience, not a single day of decent service so far. This is the worst experience I have had with an ISP having had broadband at home for over 10 years and will be broadcast far and wide. The measure of any service is how they deal with faults and in the case of Plusnet their support would fail the Turing test if it weren't for the fact that a computer would actually pay attention to what is sent to it, hence I know I'm dealing with humans.

It's tragic when moving from a bargain basement, mass market ISP like Sky who have being an ISP as a value-add and retention tool to their TV product to a Which? recommended provider who apparently pride themselves on offering low prices without compromising on service and for whom their primary business is being an ISP is an unmitigated disaster.

Had they just done the bleeping lift and shift I asked for weeks ago this would be fixed but Plusnet run through their scripts in order to run through the BT Wholesale scripts and the customer suffers.

Name: portal.plus.net
Address: 212.159.9.2

Ping statistics for 212.159.9.2:
Packets: Sent = 100, Received = 96, Lost = 4 (4% loss),
Approximate round trip times in milli-seconds:
Minimum = 57ms, Maximum = 418ms, Average = 118ms

Uptime: 0 days, 13:54:59
DSL Type: ITU-T G.992.3
Bandwidth (Up/Down) [kbps/kbps]: 663 / 2.103
Line Attenuation (Up/Down) [dB]: 39,0 / 66,5
SN Margin (Up/Down) [dB]: 5,8 / 6,0
Error Seconds (Local/Remote): 18.408 / 3
FEC Errors (Up/Down): 3.507 / 32.610
CRC Errors (Up/Down): 202 / 87.910
HEC Errors (Up/Down): 639 / 21.189

In reply to a post by Chrysalis:

can I ask why the change to plusnet?

I was going to say typical HR fault, but given what you said we have to look at whats changed which is going to be the line card.

Everything changed bar a tie cable from the MDF onwards. The move was from LLU to BT Wholesale.

This'll be why VoIP's knackered then.

Upstream Loss: 0.0% (=0/500)
Upstream Order: 100.0% (=249500/249500)
Upstream Jitter: ave=0.59ms, max=36.868ms
Downstream Loss: 11.2% (=56/500)
Downstream Order: 94.18677354709419% (=234996/249500)
Downstream Jitter: ave=78.82ms, max=154.32882882882882ms

Sounds like a bad experience so far, can you PM me your username and I'll get someone to take a look into what's going on and what the next steps are.

3dB and interleaved has been helpful to me in the past - at 9km. Does G992.1 do better ?(unlikely, but always worth a try).

An SFI engineer came today, Chris, did his thing and as expected found nothing wrong with the line.

He was then spun a load of complete nonsense by BT Wholesale.

Here's the post relating to this on your own forums.

This is in no way personal to you Mr Taylor.

Your supplier's incompetence is breathtaking.

As expected the Openreach engineer informed that the line itself is in great shape right back to the MDF, as expected given that this household has been featured on national TV over the quality of broadband provision, has been the centre of a ward-wide superfast broadband campaign, and has had 2 REIN engineers base themselves in it to diagnose a fault.

However he called BT Wholesale and they informed him that the difference between the stability on Sky and Plusnet is apparently because when we were on Sky LLU, and note I said LLU, we were on a 20CN 'up to' 8Mb service, whereas we're now on a 21CN 'up to' 24Mb service.

However it's all fine now. BT Wholesale have fixed the SNR margin at 6dB which means it can't drop any lower and all the issues should apparently now be fixed.

Of course there are a few issues with this.

1) Sky is ADSL2+ / G.992.5
2) We're on ADSL2 / G.992.3, which you'd expect to be more stable.
3) There is absolutely no way that BT Wholesale can fix downstream SNR margin.

So this wonderful, fixed SNR margin, super reliable service has in the time I was writing this forum post already dropped and packet loss has returned.

Ping statistics for 212.159.9.2:
Packets: Sent = 100, Received = 96, Lost = 4 (4% loss),
Approximate round trip times in milli-seconds:
Minimum = 37ms, Maximum = 483ms, Average = 144ms

Strangely enough the SNR margin is also still unstable. There I was thinking that BT Wholesale could defy the laws of physics in one direction and control the signal quality my modem reported in the other.

Any other ideas or ready to cut losses, refund the line rental paid in advance, and provide a MAC so that we can take the lesson as learned and return to our obsessive avoidance of BT Wholesale?

Thank you.

Incidentally in the time it's taken me to write this post the line actually dropped twice, as I'm sure will be reflected by your RADIUS logs.

3) There is absolutely no way that BT Wholesale can fix downstream SNR margin.

They can fix the target downstream SNR margin rather than let DLM play with it, that's what the shorthand meant.

A line should be fine running at 0 dB SNR margin as that gives an acceptable error rate, so going to 6 dB gives more headspace. Does your kit have an ATM layer BER test function ?

how did you get 2 rein engineers to stay overnight and get previous issues on tv? O_o

what you seeing now is the sort of rubbish BTw have been doing for years, I used to moan alot in the past about my own adsl line years ago before I moved to ukonline.

The question is what specific task do you think is required to get the area of the fault looked at?

lift and shift
cease and reprovide?
something else.

I would think a cease and reprovide gets the line physically discconected and recconected again so perhaps thats what plusnet should be ordering. If BT still do it a managed cease and reprovide is done all in one day.

Sorry you couldnt say why the line owner jumped to plusnet, to save money? the old rule dont rock the boat.

I'm aware they can fix the target margin. If you could tell me just why they'd do this on a line that's showing errors and proclaim it fixed that'd be great.

I don't need an ATM layer BER test function.

FEC Errors (Up/Down): 188 / 9.692
CRC Errors (Up/Down): 25 / 5.292
HEC Errors (Up/Down): 80 / 1.754

It is perhaps worth my mentioning I've worked for a datastream OLO and an LLU operator to save the teaching grandma to suck eggs thing.

In reply to a post by Chrysalis:

how did you get 2 rein engineers to stay overnight and get previous issues on tv? O_o

They didn't stay overnight, they based themselves in the property to diagnose the wider issue but obviously ensured that the connectivity from house to exchange was tip top.

In reply to a post by Chrysalis:

The question is what specific task do you think is required to get the area of the fault looked at?

lift and shift
cease and reprovide?
something else.

I asked about a lift and shift however it isn't my problem to diagnose this and I'm fed up of this [censored].

In reply to a post by Chrysalis:

Sorry you couldnt say why the line owner jumped to plusnet, to save money? the old rule dont rock the boat.

It wasn't about money, the line was previously with Sky. Stupidly it was more about getting rid of Sky TV hence wanting to be 'free' of the company entirely and deal with a 'specialist' ISP for whom it's their main business.

That'll teach us.

well good luck in getting it resolved, I would push the cease and reprovide route or at least a lift and shift.

I agree with you that its a joke to suggest fixing it at 6db would fix it when already a DLM'd 6db interleaving has not fixed it.

One last thought.

We both know setting a dsl connection isnt just snrm and interleaving settings, there is many tunables including also INP. So its possible sky did have a different configuration that was stabilising the line which of course got lost on the switch and BT's configuration is simply inadequate. Then there is also the modem to dslam compatability been a possible factor.

Edited by Chrysalis (Fri 16-Aug-13 14:42:18)

In reply to a post by Ignitionnet:

I don't need an ATM layer BER test function.

always nice to put the kit into test mode and see what's happening at the most basic level.

Did the other two types of router you'll have tried behave identically?

In reply to a post by yarwell:

In reply to a post by Ignitionnet:

I don't need an ATM layer BER test function.

always nice to put the kit into test mode and see what's happening at the most basic level.

Did the other two types of router you'll have tried behave identically?

The SFI engineer's JDSU behaved identically. Next question.

EDIT: As, incidentally, did a Billion BiPac 7402. Those have an ADSL chipset (Globespan / Conexant Argon 432) that cost a small fortune and is absolutely excellent.

The standard router uses Broadcom, the Billion Conexant, connecting to an Infineon chipset MSAN. I've covered the bases and done my diligence.

I'm not meaning to do the big 'I am' or be rude but every post you've written on this thread has told me things I already know and have already done, hence my comment about grandma and sucking eggs. I'm trying to save your time and mine..

Edited by deleted (Sat 17-Aug-13 10:38:12)

As a follow up to this SFI engineer's JDSU connected to test socket showed the same symptoms.

In reply to a post by Ignitionnet:

I'm not meaning to do the big 'I am' or be rude

but you do achieve this, so carry on playing to the gallery.

I'm completely lost as to where I did that but no matter. I wouldn't see fit to speak authoritatively to you on chemical engineering and fuel technology and accuse you of playing to the gallery when you say it's not necessary.

FYI - Both Openreach and BTWholesale returned it as no fault found to the provider, who with the information I provided returned it to BT Wholesale.

It appears the issue was resolved over the weekend. Had I not 'played to the gallery' or done the big 'I am' so much it wouldn't have been fixed as quickly or potentially at all.

Anyway all fixed for now, may this thread die quietly.

Or not - it behaved for a couple of days and since a resync on Monday has been hosed again.

Anyway due to a DLM reset on Friday Plusnet will not revisit the fault until the 27th. In the interim I'm left with this.

Any of you lovely people think of anything else that can be done? DLM has been 'tweaked', SFI visit found nothing wrong with the line, wondering what's next if/when we get to 27th and the line is still sluggish and lossy.

aside from switching back to sky, then get a router where you can take matters into your own hands and pump the target snrm up to about 15db.

if you have the router you mentioned earlier then you can already do that.

Your connection certainly looks better than it was but still there are certain issues. I'll speak with our faults team to see if we can get this picked up, It may be worth us setting the SNR target on the DLM higher.

In reply to a post by chrispurvey:

Your connection certainly looks better than it was but still there are certain issues. I'll speak with our faults team to see if we can get this picked up, It may be worth us setting the SNR target on the DLM higher.

Alternately rather than reducing the speed on an already pretty slow line to cover up the issue we could try figuring out why it worked just fine on Sky at 6dB but is a bit of a nightmare on BT Wholesale at 6dB.

Perhaps deeper INP and interleave would be an option?