cancel
Showing results for 
Search instead for 
Did you mean: 

Lumpy QLN graph = crosstalk ?

FIXED
Anotherone
Community Veteran
Posts: 19,107
Thanks: 455
Fixes: 21
Registered: ‎31-08-2007

Re: Lumpy QLN graph = crosstalk ?

That lower speed result could be congestion which can even be at the tester. Try the TBB one and speedtest.net for comparison.

The precise tones REIN can hit could in theory change if the source's fundamental frequency is not Xtal controlled - most REIN sources aren't but they do often (luckily) come up on the same tones +/- one or two.

hitchhiker43
Aspiring Pro
Posts: 200
Thanks: 29
Fixes: 4
Registered: ‎06-07-2016

Re: Lumpy QLN graph = crosstalk ?

Nope all speed tests roughly the same, anyway I think it's stuck because I cannot reset my PPP without busting the DSL again, just asked Pnet if they can independently reset PPP but they say they cannot either. I am getting away with it atm by telling the wife it's the sites are slow! hopefully I can get the REIN investigated/fixed before she gets to mad Smiley

Nope it's not a fixed frequency at all, it sweeps and changes in amplitude much like a very very large motor drive might or a b****y wind turbine!! It's now knocked out all my tones above 167.

Tp-link W8960n on longgggggggg line!
Anotherone
Community Veteran
Posts: 19,107
Thanks: 455
Fixes: 21
Registered: ‎31-08-2007

Re: Lumpy QLN graph = crosstalk ?

Strikes me you need to get busy with the radio set, also keep your eyes peeled and have a brew ready Wink

hitchhiker43
Aspiring Pro
Posts: 200
Thanks: 29
Fixes: 4
Registered: ‎06-07-2016

Re: Lumpy QLN graph = crosstalk ?

BT visited Thurs 25th Aug as I reported phone line fault (noise), cannot say what there conclusions are as Pnet have not updated fault ticket so I also don't know if there gonna charge me to add to my misery!!

6 drops of 6dB or more in last 48 hours 4 of which caused re-sync, BT wholesale speedtest now down to 2.6Mbps

One big problem I have here is MODERATORS!! This is a relatively obscurely named thread so how am I supposed to draw attention to the unanswered fault ticket or the ongoing snr drops, I am beginning to think this site doesn't serve Pnet customers very well if we all tread in fear of raising any more issues!!

Tp-link W8960n on longgggggggg line!
Estragon
Rising Star
Posts: 811
Thanks: 10
Registered: ‎07-02-2012

Re: Lumpy QLN graph = crosstalk ?


@Anotherone

I think it's worth pointing out, as neither of you have commented on it, that in posts 10; 14; 36 the line is on ADSL2+, but in posts 17 (presumed from the attenuation) 22 and 23 it is on ADSL2.

As you will know but the OP may not, ADSL2 often (and in his case does) give a 3dB drop in attenuation but cuts out the top end frequencies/bins. Around this range the net effect on sync speed is likely to be small, and in either direction. But the effect on the error rates due to the presence or absence of the upper frequencies could be marked. As could the knocking out of high end bins when the noise occurs.

I haven't gone through trying to correlate the error rates v ADSL2/2+ or what bins get knocked out when the noise comes along. 

 Edit - and I don't like the fact there are 4 wires coming out of the gable junction box. If they are all connected inside it then there are bridge taps, whether or not there is anything at their other ends.

 Plus the OP's picture showing the back of the NTE5 faceplate seems to show the ring wire connected. Two of?

 All that needs sorting. (Barring the one going from gable box to the NTE5 of course).

hitchhiker43
Aspiring Pro
Posts: 200
Thanks: 29
Fixes: 4
Registered: ‎06-07-2016

Re: Lumpy QLN graph = crosstalk ?

Yes I am very sorry for any confusion caused I can confirm the modem is set to ADSL2 only! And as you rightly point out this limits the frequencies to ~1100Khz and frankly on my length of line there is nothing worth having above that anyway. Typical bin knockout is everything above 170 or sometimes 150. The net effect on sync speed is usually a drop from ~4300Kbps to ~2600Kbps (this dramatic drop is caused by the resynch occurring during the peak period of the interference event).

The connections in the gable junction box are just an inline connection between two pieces of old figure 8 drop cable, there is no bridge tap there. These joints were replaced during the recent BT visit though sadly the drop cable remains figure 8.

The ring wires used to be irrelevant because the filter was directly into the master socket but is no longer, thankyou for reminding me Smiley

 

Tp-link W8960n on longgggggggg line!
hitchhiker43
Aspiring Pro
Posts: 200
Thanks: 29
Fixes: 4
Registered: ‎06-07-2016

Re: Lumpy QLN graph = crosstalk ?

A small update, I removed the close coupling between the drop cable and power cables and since then no events on my DSL. I am having to come clean as startlingly Pnet want to send another BT engineer to investigate REIN.  I very much hope he might replace the cr*p drop cable altogether or at least decide it needs re-routing. In any case my opinion of Pnet is rising, they escalated my ticket internally did a decent investigation and decided it needs more work, so here is hoping, it's also possible they read this forum or someone had a quiet word many thanks Smiley

Ohh and nobody at Pnet this time told me I was getting more than 1Mbps so should be happy, perhaps the penny is dropping that this really p*sses of customers.

Tp-link W8960n on longgggggggg line!
hitchhiker43
Aspiring Pro
Posts: 200
Thanks: 29
Fixes: 4
Registered: ‎06-07-2016

Re: Lumpy QLN graph = crosstalk ?

Fix

A final note and closure BT investigated and agreed there was severe REIN on the power cables and "officialised" my re-routing to avoid them. This has had the effect of reducing the REIN event amplitude to 3db and that is no longer sufficient to knock the modem/dslam out of sync. The line is now stable as Pnet have also locked the DLM to 6db and set it to ignore errors. A very peculiar problem caused by REIN that both OFCOM and the BBC have refused to investigate and will probably remain. Many thanks to everybody here and the support staff at Plusnet for getting this situation under control and getting my service stable Smiley

Tp-link W8960n on longgggggggg line!
Anotherone
Community Veteran
Posts: 19,107
Thanks: 455
Fixes: 21
Registered: ‎31-08-2007

Re: Lumpy QLN graph = crosstalk ?

Sorry for lack of attention here - am away and have limited access atm. Will reply more shortly. 

Anotherone
Community Veteran
Posts: 19,107
Thanks: 455
Fixes: 21
Registered: ‎31-08-2007

Re: Lumpy QLN graph = crosstalk ?

The following comments are as much for anyone else subsequently reading this saga as they are for hitchhiker43. 

I don't know why the engineer that attended on the 28th did not replace the drop-wire. Surely he was made aware (by yourself if not in his notes) that the fault report was for Noise on the line - specifically Hum,  which would be the result of Untwisted drop-wire and it's proximity to power cables. If need be I would have over-egged the impact it periodically had on phone calls.  I might not have mentioned the "secondary" impact on the broadband depending on how understanding the engineer was a the time - remember this is a phone line fault that is being dealt with. Should he not have deemed it necessary after it being explained I would have promptly replied to the Ticket complaining that he didn't replace it despite it not being fit for purpose - don't wait for a Plusnet ticket update in such circumstances. 

Your last post implies the drop-wire was replaced and re-routed on the subsequent visit. It should be emphasised that anyone with old Untwisted drop-wire and having drops in BB whether due to Noise/REIN/SHINE etc. should be insisting it be replaced - by hook or by crook. 

OpenReach can pursue the REIN investigation with their own REIN engineers - this should be happening even though the impact is now reduced  - it will still cause a significant increase in errors when it appears. You should pursue this via the ticket. 

Just one other small point, it's not the Moderator's function to flag posts etc.  for attention any more than anyone else although often they will do so if they spot one requiring attention,  as do other posters such as myself.

@Estragon whilst your observations are perfectly valid,  as far as the Faceplate/bellwires are concerned, the OP did mention in one post that the extension wiring was no longer used,  the problems experienced were present when plugged into the test-socket. Also in REIN cases such as this where the impact is significant at lower tones and the availability of higher tones above 255 is virtually non-existent or has dropped  off significantly  due to line length, I've never found it worth the effort of worrying about things like ADSL2/2+ until the primary cause - Untwisted drop-wire - is first  resolved. It's frequently lots of time consuming effort for very little,  if any,  gain on such occasions. 

hitchhiker43
Aspiring Pro
Posts: 200
Thanks: 29
Fixes: 4
Registered: ‎06-07-2016

Re: Lumpy QLN graph = crosstalk ?

Well I think I can explain and defend the 1st BTeng, he was investigating a PHONE fault and did as much as he could within that remit and TBH the drop wire could not be implicated though he did re-crimp the gable connection for insurance.  It was only after that failed that Plusnet finally did what they should have done in the first place and raised a BB fault with BT that resulted in the drop wire replacement.

The drop wire is not a panacea, firstly it only has a telephone standard twist rate of once every few inches so nothing like the standard of CAT5, yes it's better than fig8 but not that wonderful. The major effect was the re-routing, I did try for the opposite side of the house to be completely away from the power cables but that was asking to much!

As for continuing to pursue the REIN I would prefer to return to my life before all this began as frankly it's a large amount of effort for ever diminishing returns and as it's a village wide problem, let some others have a go Smiley

P.S. just got a hit but its only 2.5db not enough to knock me out, I can cope with that Smiley

Tp-link W8960n on longgggggggg line!
Anotherone
Community Veteran
Posts: 19,107
Thanks: 455
Fixes: 21
Registered: ‎31-08-2007

Re: Lumpy QLN graph = crosstalk ?

Obviously it's entirely your choice as to how much, if any, time you choose to spend on pursuing the REIN issue - if and by how much it impacts on the performance of your connection. I did not recall you mentioning it's effects were more widespread - "village-wide".  I will simply wish you well there and hope the effects are minimal and that whatever the source of it is - it breaks itself completely and so disappears as a problem altogether. 

However, when it comes to the drop-wire,  your comments could leave the impression that the old figure 8 cable is quite acceptable which it most certainly is not. There's no defence for the 1st engineer not replacing it with the correct standard CW1411 cable - the PHONE line was picking up audible Hum, clearly exacerbated by proximity to power cables, and movement away from the power cables would in any event involve replacement! As far as the BB goes, it was only because of the major impact of proven REIN - from what you have said,  primarily being distributed and radiated by the power cables, that resulted in its replacement further from the power cables. Unfortunately I've seen too many cases of BB faults with dropping connections where such old cable is involved and it was not replaced, so your implied  assumption that a BB fault will automatically cause replacement is not valid. ANY (and all) opportunities to get old cable replaced must/should be used.

Whilst CAT5 cable is a higher standard,  BT don't use it. Normal CW1411 (external) and CW1308 (internal) has an acceptable twist rate even for FTTC installations, so although it might not be a panacea in the worst of situations, it is for the great majority. Your situation is not the worst I've encountered Wink

hitchhiker43
Aspiring Pro
Posts: 200
Thanks: 29
Fixes: 4
Registered: ‎06-07-2016

Re: Lumpy QLN graph = crosstalk ?

You may be amused to know I am now pursuing the power company for poor quality of power supplied! no response so far but that's cool because after 6 weeks I can raise it to ofgem lols

 

The point is someone is putting sh*t on the powerlines bigtime and should be held to account for messing up everybodies broadband don't you think ?

Tp-link W8960n on longgggggggg line!
Anotherone
Community Veteran
Posts: 19,107
Thanks: 455
Fixes: 21
Registered: ‎31-08-2007

Re: Lumpy QLN graph = crosstalk ?

Excellent idea. We know it messes up MW-AM radio (obviously) it may also effect the HF end of LW or is it more widespread on LW? How about FM frequencies?  You could add all that to the mix with your moan at the Power people.

I hope you are moaning at the people that maintain the power-lines etc for your area (not the same as those you pay your bill to) the former are the ones you'd contact if/when you have a power-cut.
If you've moaned only to the people you pay the bill to, they should pass it on, but I'd expect you could possibly get a BS response so I'd contact the other lot directly.

hitchhiker43
Aspiring Pro
Posts: 200
Thanks: 29
Fixes: 4
Registered: ‎06-07-2016

Re: Lumpy QLN graph = crosstalk ?

Its always 425Metres! UK power networks not my power provider, so far they emailed me a response then a few days later nice young lady tried to fob me off on the phone then admitted she wasn't technical and promised an engineer would call me.

Here are my recent REIN records, still in the middle of the latest one so last entry incomplete!

As for moaning, I think deservedly so and I am sure many would if there internet was unusable sometimes for 10's of minutes a day, almost certainly someone is operating an illegal piece of equipment causing this!

 

P.S. Event just finished as I was exploring the radio dial and finding some lower harmonics, just getting another set out marked in Khz when it stopped!! better luck next time lols

1st plot (14:29) is during event 2nd is after, finally corresponding snr plot for start of event

Tp-link W8960n on longgggggggg line!