Plusnet to BT complete BUT . . .

Well it seems my migration from Plusnet to BT Infinity 2 (76Mbps) happened at 7am this morning with the BT comfirmation email sent shortly afterwards. The only thing is, my sync speed for well over a year has been around 68Mbps but is now a full 10 meg lower at 58Mbps.

I'm using the same Billion 8800NL modem and nothing else has changed. I noticed that the SNR is now 5.6 when it's usually around 6.1.

Is the sync likely to increase?

What's the max attainable rate showing as?

I'm in work now with no access to modem (one of the disadvantages of no static IP with BT) but it was only slightly higher (by a few Kbps if I remember) than the sync rate.

There's not reason why the sync speed shouldn't be the same given it's the same wire, terminating at the same cabinet, run by the same outfit (OR) and, possibly, even on the same line port. (I don't know if the ISPs have any ability to bias the DLM setting priorities towards stability or speed).

I wonder if it's just an artefact of Interleaving being turned on or some other issue which will settle out in time.

It would be best to post the full line stats when you have the chance.

If you could post before and after line stats that would help a lot mate.

Cheers

Interleaving and g.inp have been on for probably 6-8months and were still on when I checked this morning. I did power the modem off for 20mins this morning before work and then back on again to see if this made a difference bu t no change at all.

I'll get the full stats when on my lunch break and post them.

Yep will do. I'll get the latest when on my lunch break. The previous ones (before switch over today) are available on mydslwebstats - I'm Tenbyboy2. I'll post these later along with the new ones though for people that haven't got access to them.

In reply to a post by WelshWArrior:

Yep will do. I'll get the latest when on my lunch break. The previous ones (before switch over today) are available on mydslwebstats - I'm Tenbyboy2. I'll post these later along with the new ones though for people that haven't got access to them.

Do you recall any loss of sync during the migration?

Migration happened at 7am this morning and from looking at my logs, there was no loss of sync that I can see. I did only have a quick look though.

In reply to a post by WelshWArrior:

Migration happened at 7am this morning and from looking at my logs, there was no loss of sync that I can see. I did only have a quick look though.

I'm wondering how the sync rate decreased without a loss of sync.

Was this an IP profile decrease?

There may have been a loss of sync but I didn't see it upon the quick look through the logs before work. I'll take another look at lunch.

Bit off topic I know but I was thinking of changing from BT to PlusNet for a cheaper deal.
Just wondered why you changed?

Various different factors came into it really. I used to have about £15 in referral value with Plusnet but this has dwindled to about £2 making the amount I paid a lot more. Also, my mobile contract with EE was up and as I already had a new Galaxy S7 edge, I just wanted a sim only package which BT have very good deals on. Added to this, as I was also changing Fibre to BT, they give an extra £5 BT mobile.

I'm glad I changed and it was very smooth. Just hope I get my extra 9-10Mbps back

In reply to a post by smsmasters:

In reply to a post by WelshWArrior:

Migration happened at 7am this morning and from looking at my logs, there was no loss of sync that I can see. I did only have a quick look though.

I'm wondering how the sync rate decreased without a loss of sync.

Was this an IP profile decrease?

Of course it wilp have came with an IP profile decrease as the IP profile is set as a direct result of the sync value.

Thanks for taking time to reply.

Good luck with getting the extra back!

Thanks! And no probs at all about getting back to you. It really has been the most seamless migration I've ever had. Don't need a MAC code anymore and I didn't even have to speak to Plusnet about leaving.

Tip: go through a cashback site like topcashback. I used this one for the first time and got £192 for signing up to BT!!

In reply to a post by lee111s:

In reply to a post by smsmasters:

In reply to a post by WelshWArrior:

Migration happened at 7am this morning and from looking at my logs, there was no loss of sync that I can see. I did only have a quick look though.

I'm wondering how the sync rate decreased without a loss of sync.

Was this an IP profile decrease?

Of course it wilp have came with an IP profile decrease as the IP profile is set as a direct result of the sync value.

What I was trying to ask was whether the loss of speed was a result of actual sync rate decrease or IP profile decrease. He mentioned there was no loss of sync.

So perhaps FTTC migrations entail no loss of sync but software IP profile changes?

In reply to a post by smsmasters:

So perhaps FTTC migrations entail no loss of sync but software IP profile changes?

No.

Is that no to no loss of sync, no to software IP profile changes or no to both??????

In reply to a post by WelshWArrior:

Migration happened at 7am this morning and from looking at my logs, there was no loss of sync that I can see. I did only have a quick look though.

In reply to a post by RobertoS:

In reply to a post by smsmasters:

So perhaps FTTC migrations entail no loss of sync but software IP profile changes?

No.

So how did this occur then?

Edited by deleted (Mon 25-Apr-16 11:46:34)

I believe Openreach's VDSL service does use lSeamless Rate Adaption, so your sync rate can change without a retrain.

In reply to a post by lee111s:

I believe Openreach's VDSL service does use lSeamless Rate Adaption, so your sync rate can change without a retrain.

SRA is not enabled on OR's VDSL circuits.

http://forums.thinkbroadband.com/fibre/t/4239755-re-...

Edited by deleted (Mon 25-Apr-16 12:35:02)

As I mentioned previously today, I'll check the log in more depth at lunch time. I didn't say there wasn't a resync, I said I didn't see it in the logs.

That to me reads they do use it although don't simply call is SRA as they can use the algorythm to modify the interleaving depth too, without a retrain?

"The SRA is limited to tweaking the minimum and maximum data rates (banding), but the Openreach DLM algorithm can also modify the interleaving delays and impulse noise protection levels (and theoretically the TSNRM too)"

The use of the word "also" being key.


Additionally, that post was made in 2013, some 3 years ago. I note that Openreach did carry out some SRA trials along with vectoring so it's possible SRA in one form or another is in use.

I guess we'll wait and see what the OP comes back with.

Edited by deleted (Mon 25-Apr-16 13:00:12)

In reply to a post by lee111s:

That to me reads they do use it although don't simply call is SRA as they can use the algorythm to modify the interleaving depth too, without a retrain?

"The SRA is limited to tweaking the minimum and maximum data rates (banding), but the Openreach DLM algorithm can also modify the interleaving delays and impulse noise protection levels (and theoretically the TSNRM too)"

The use of the word "also" being key.

OR's DLM requires a retrain for changes to take effect. OR does not use SRA on FTTC circuits.

SRA dynamically adjusts sync rate on the fly which I've never seen occuring.

"No" to software IP Profile changes. (Other than, of course, any change to IP Profile caused by a change of sync).

FTTC IP Profile is 0.9678 x sync (no G.INP) or 0.9669 x sync (with G.INP), though some recent results from ECI kit seem to be clouding that simplicity.

I think I see what has caused this. G.inp has been disabled on the line. Obviously DLM must also have been reset if G.inp has been turned off. Would I be right in my assumption?

Sounds that way. I think we saw one a week or so ago where this happened but it wasn't BTW >> BTW.

Fingers crossed then when G.inp kicks in (and hopefully soon as there's lots of errors on line now) the speed will increase. I seem to remember when G.inp kicked in last year for the first time, I gained 7-8Mbps!

Have a look at your FEC graph ... It is interesting that from either 18:30/19:00/19:30 to 22:30 every night they drop down to near zero.

That suggests there is a noise source somewhere that is causing a problem and the source is turned off during the evening. I would certain not expect such a specific notch every day.

I've noticed that in the past but not really been able to put my finger on what could be causing it. I might have to get my thinking cap back on and try to find out. Not that I've much chance of being able to do anything about it mind.

I've known for ages that my line is underperforming. It is only 275m long according to OR Engineer's hand held tool but attenuation is always between 19 and 19.2.

IF and it is a big IF, you could sort out why the FECs drop to near zero in the evening you would probably find that G.INP is not invoked and you would see the higher speeds continually.


As for distance, yes, 275m and 19dB attenuation is a little odd. At 400-450m I see 16dB and would expect you to be down at 10-12dB. It could possibly be that the copper feeding you is smaller diameter that standard - BT has used different sizes in the past.

In reply to a post by WelshWArrior:

.... It is only 275m long according to OR Engineer's hand held tool but attenuation is always between 19 and 19.2.

My line is 272m long according to BT engineer and attenuation as measured by my Billion 8800NL is about 17.4dB for some sort of comparison with your figure. According to the OR engineer, some of my line is aluminium (where it goes underground for about 10m). Do you happen to know if some or all of your line is also aluminium? That may account for extra 1.6 to 1.8dB attenuation?

Update on this:

Ran BT speedtest and was prompted to complete TAP3 which failed miserably. Online chat with BT who initially said nothing wrong. I then asked him to check my TAB3 test result which then prompted him to do more tests, confirming 'network issue'. He then raised the fault.

10 mins later, sync drops and immediately reconnects. Attainable is now 70Mbps but actual sync rate is still only 57Mbps i.e hasn't moved. I now notice though that G.INP is now back on.

Is my throughput likely to return to pre-migration levels? Considering I'm now sync'd at 70Mbps (albeit with 220602 FECs a minute!!) I'm hoping so. Guess I'll just have to wait and see what OR engineer finds out.

Edited by WelshWArrior (Tue 26-Apr-16 15:54:49)

Current stats showing the hideous amount of errors latency is also rubbish, averaging 60ms at moment:

adsl info --stats
adsl: ADSL driver and PHY status
Status: Showtime
Last Retrain Reason: 0
Last initialization procedure status: 0
Max: Upstream rate = 17443 Kbps, Downstream rate = 68209 Kbps
Bearer: 0, Upstream rate = 17443 Kbps, Downstream rate = 57676 Kbps

Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 6.2 6.3
Attn(dB): 19.2 0.0
Pwr(dBm): 12.7 7.5

VDSL2 framing
Bearer 0
MSGc: 18 20
B: 51 237
M: 1 1
T: 64 64
R: 12 16
S: 0.0287 0.4342
L: 17848 4680
D: 1129 1
I: 64 127
N: 64 254

Counters
Bearer 0
OHF: 10745906 436581
OHFErr: 73 11
RS: 2750814320 2095667
RSCorr: 35931786 35
RSUnCorr: 112 0

Bearer 0
HEC: 20 0
OCD: 0 0
LCD: 0 0
Total Cells: 2196793356 0
Data Cells: 46465947 0
Drop Cells: 0
Bit Errors: 0 0

ES: 16 11
SES: 0 0
UAS: 32 32
AS: 19809

Bearer 0
INP: 3.00 0.00
INPRein: 0.00 0.00
delay: 8 0
PER: 1.84 6.97
OR: 104.17 29.82
AgR: 57780.29 17472.53

Bitswap: 4782/4782 25/25

Total time = 5 hours 30 min 41 sec
FEC: 35931786 35
CRC: 73 11
ES: 16 11
SES: 0 0
UAS: 32 32
LOS: 0 0
LOF: 0 0
LOM: 0 0
Latest 15 minutes time = 41 sec
FEC: 6779 0
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 15 minutes time = 15 min 0 sec
FEC: 199435 0
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Latest 1 day time = 5 hours 30 min 41 sec
FEC: 35931786 35
CRC: 73 11
ES: 16 11
SES: 0 0
UAS: 32 32
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 1 day time = 0 sec
FEC: 0 0
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Since Link time = 5 hours 30 min 9 sec
FEC: 35931786 35
CRC: 73 11
ES: 16 11
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
>

It will be better when G.INP is enabled

G.INP is enabled. It got enabled about 10 mins after I rang them today. The sync dropped and reconnected along with G.INP being enabled. This is confirmed if you look at mydslwebstats. I am tenbyboy2

Have a look. It says " G.INP is inactive on this line (has been enabled but is currently turned off)"

I meant ON

I was going to say - an interleaving depth of 1129, and so many FEC errors doesn't look at all like G.INP being active.

Does whatever router he is using show Bearer 1 when active?

Yeah, the Billion 8800 NL does indeed.

That's weird - it was active earlier on after my phone call to BT this afternoon and it was showing in dslstats as active. Only thing I can think of is that BT OR have reset the DLM again!

Same again last night ... FECs down from 200,000 to nearly zero from around 19:00 to 22:30 - is there anything that you have turned OFF at that time? Or maybe something that goes from idle to ON ... a PSU in an unloaded state could generate a lot of noise and as soon as it goes under load the noise disappears or changes? Or it could be filtering out noise from another device.

Hi! I didn't get chance to have a think about it last night but it's definitely weird isn't it?

Only thing off the top of my head that is on at that time and not others is the TV and Sky box. They are both about 10ft away from the modem. Perhaps I'll try powering them off at the wall later and check the graphs.

Could be worth trying ...

Power them OFF completely and even unplug them - leave for 30 minutes. Plug in and leave on standby for 30 minutes, then power up for 30 minutes. Start it on the hour or half hour so there is an easy reference.

You say 10feet - that is still close enough for RFI to cause problems and it could be radiated or conducted.

OK thanks for that. I'll give it a go.

Just s shot in the dark but but I have just remembered something. The router is in the corder of the lounge on the floor right against the skirting board. On top of the skirting board are some radiator pipes and on top of that are 2 old telephone extension cables from before we bought the house last year. They aren't connected at either end but the cores are bare and could be touching the copper pipes. I wonder if there is somehow some static in the pipes that is transferring to the cable ?? Is that possible?

Reason I ask is that the heating could be on at that time if the wife has put it on for the kids OR they could be having a bath too so the boiler would definitely be in use. The boiler itself was brand new in Jan so surely it would't be interference from that directly?

Edited by WelshWArrior (Wed 27-Apr-16 11:03:17)

It looks too consistent for a bolier or heating related issue ... same time every night and not in the mornings. Although, if you have continual hot-water and only have the heating one for those few hours in the evening, then it might the valve controls ...

If your Sky & TV tests don't show anything, try putting the heating ON and see if that makes changes - or this eveing, have periods of CH & HW off, CH on, HW on and both off.

Yea I agree - way to consistent. Just looked back a few months. It's the same time each night. Even when we've been away too.

I'll do the SKY TV tests this eveing and report back tomorrow. Thanks for the help so far . . .

BT have an BT Engineer booked for me next Tues AM because of the difference in attainable (68583) and actual sync rate (57935) but it would be nice to get to the bottom of it on my own

It might even be worth turning off everything in the house including fridge, freezer, alarm and then unplug everything from the one ring which the hub/modem uses. Then power up that one ring and turn on the modem and a PC and see what that shows. Do it at a time when there are high error rates and see if it disappears - if it does, the problem is yours, if it does not then you need to consider neighbours or external.

Sounds like a good plan. Might have to wit until I can ship the kids/wife out for a few hrs on weekend though

I'll certainly still try the SKY/TV option later though.

I'm hoping it is my problem as at least if it is, I can do something about it. Obviously if it's external then I'm pretty much stuck!

Your lines been reset during the switch... meaning as your where G.INP before you've been dropped down to Interleaving with an 8ms delay on DS... G.INP should reappear on downstream soon.

I've found that any tiny changes now that are made seem to trigger this reset.... BT recently messed with my account whilst applying a disocunt this mean an order was gernarated fro my line to go from 80/20 to 80/20...

You'd think no changes would occure but my line reset as a result..... lost G.INP on my upstreas as a result **

Hopefully it wil be back next week though I'm currently having an IP Profile issue too, so unless that budges I think I'll be getting back on to wretched BT care.

That's not good! Fingers crossed it'll be back on for both of us soon!

I'm still determined to try to get to the bottom of the strange noise issue causing very high FEC on the line that seems to disappear every night at 7pm on the dot until 10.30pm when it returns!! See a few lines down on this thread for a discussion between myself and MHC.

As I recall, it was your connection that showed G.INP dealing very well with large amounts of errors. I wonder if the errors you're seeing now are caused by the same thing.

Yes G.INP was doing really well. It only got turned off as DLM got reset by my move to BT from Plusnet this week. Now I'm sync'd at 57000ish but attainable never drops below 68000Kbps. I was seeing upwards of 400,000 fec/hour yesterday. Slightly lower at 200k today!

Something funny is going on as if it was only noise, surely the attainable wouldn't be showing so high would it?

My line was pretty much the same, not much that can be done about it other than allowing G.INP to do its job... as intended.

My lines has stabilised very well now, still mising speed but that just the way it is it seems.

Once thing though, I noticed you resynced a good few times and two of the times where within 15 minutes of each other... this won't help DLM in deciding what to do... if anything may make it stress and put you on higher depth of interleaving.

I feel that if you leave it be from now onwards it will have G.INP by Monday morning... I found it takes around 48-72 hours after reset to reappear on downstream.

In reply to a post by WelshWArrior:

Something funny is going on as if it was only noise, surely the attainable wouldn't be showing so high would it?

The attainable seems to increase when interleaving is increased, for some reason.

OK thank - good to know. I'll leave it be and just check the electrical stuff in/around the house now to see if that's what's causing the very (as in exactly same time each night!) consistent drop/subsequent rise in noise early evening.

What happened at around 17:00?

Errors seems to have dropped?

I'd love to say I know but I haven't a clue! I was home but wasn't doing anything that I wouldn't normally be doing. The only that is different to usual is that the washing machine was on. BUT I put that on at about 4.30pm.

I am thinking about making a note of each time I turn a new electrical device on. At least if the noise drops again, I can compare times.

And tonight 18:55 to 22:25 - check anything you have on a timer and is the timer out by 5 minutes.


One thought - do you have evening only street lights?

Yeah it's like clockwork isn't it?

I checked around the house at lunchtime and there's nothing on a timer. I also left the TV and SKY off and unplugged all afternoon but the issue was still there. The streetlights are evening only but are still on at the moment (they go off at midnight I think) but the noise is back so can't be them.

I have tracked the start of the issue down. It was Fri 20th November at 23:00hrs. The date isn't significant at the moment but I'm still racking my brain!

It might not even be something in your home, it could be a neighbour, or anywhere on the route of your line back to the cabinet.

In reply to a post by lee111s:

It might not even be something in your home, it could be a neighbour, or anywhere on the route of your line back to the cabinet.

He is fully aware of that. The idea is to try and eliminate everything, part by part, starting with his own property and electronics. They to start looking for outside influences.

Yeah I realise that. The line goes stright down the road into the cab about 270m away. It doesn't pass any factories/sub stations or anything like that. I suppose it could be anything. How close does the line have to pass to be affected by the source?

In reply to a post by WelshWArrior:

BT have an BT Engineer booked for me next Tues AM because of the difference in attainable (68583) and actual sync rate (57935) but it would be nice to get to the bottom of it on my own

As BatBoy has said, in less detail, that huge difference has always been normal on FTTC when interleaving is applied. Pre-G.INP when interleaving was later removed by DLM then the sync rate would rise and the attainable fall so they became similar somewhere in between.

An engineer is unlikely to find a fault to do with that, but you may be lucky and he finds what is causing all the errors. However, in either case, I'd be worried about a charge, if the error cause is one of those recently suggested. Unless BT CS have guaranteed there won't be one.

Any more thoughts?

I suppose it could be anything. How close does the line have to pass to be affected by the source?

It depends, how strong is the source, does it have an easy path onto the network. i.e. a duff sky box connected to telephone line,
I've seen one of these blitzing a 300m radius no problems.

The cheapest and easiest way to hunt for culprits is to use a battery operated MW radio. Tune to around 612 ..... but the frequency may be different as it's affecting VDSL. Remember also, lots of stuff makes 'noises' you'll hear on the radio .... not always the culprit.

Hunting these things is dark ju-ju indeed

In reply to a post by lee111s:

That to me reads they do use it although don't simply call is SRA as they can use the algorythm to modify the interleaving depth too, without a retrain?

"The SRA is limited to tweaking the minimum and maximum data rates (banding), but the Openreach DLM algorithm can also modify the interleaving delays and impulse noise protection levels (and theoretically the TSNRM too)"

The use of the word "also" being key.

And also with SRA enabled the sync rate changes both up and down according to the amount of errors/noise, and can work on a fast path profile, infact a former ISP who i was with would enable sra on some users lines, if their modem was SRA compatible as an alternative to packet loss due to errors or having interleave applied, that ISP was UkOnline BE also tried to utilise it but the results of their trial weren't great, possibly had something to do with how they configured it


Additionally, that post was made in 2013, some 3 years ago. I note that Openreach did carry out some SRA trials along with vectoring so it's possible SRA in one form or another is in use.

I guess we'll wait and see what the OP comes back with.

Yep it's certainly a challenge

Post deleted by WelshWArrior

If it was something as simple as a Sky Box though, surely it wouldn't be so regular. The noise drops on the dot at 7pm and appears again at 10.30pm - exactly same time each day.

It's been happening since 20th November by looks of it. 20th November has no significance as far as I can remember though.

Can't find anything in my house on a timer, I've tried turning off all electrical device apart from router but nothing.

Met office weather for 20th November:

The 20th was a day of sunshine and scattered showers, with the south-east remaining mainly dry, but it turned much colder in a north-west wind.

The start of last winter was very mild but looks like it was possibly this date when things turned colder... That points possibly to a central heating system somewhere. It could be that they switched central heating back on and perhaps there is a fault in a central heating system that means when the heating is "off" it is causing load interference that doesn't occur when it is running (which could explain the counter intuitive timing)?

Another excerpt from Met Office:

November began dry and exceptionally mild but often cloudy with widespread fog. The rest of the month was generally mild, dull and changeable with south-westerly winds. It was very windy at times with persistent rain across the Pennines and Cumbrian fells. There was a brief cold snap on the 20th to 23rd.

Edited by ian72 (Fri 29-Apr-16 08:38:54)

Do you have anything that will show like a 'bits per bin' type graph for your line ? If you do you may be able to find the part of the signal that's going when the source gets turned on, and that way listen more accurately with a radio,

The problem with errors though is continuous apart from the 3,5 hours every evening. That points away from heating - who leaves heating and hot water on 24 hours/day?

The two questions are: what is causing the errors? And what is suppressing them in the evening? which will the lead to: is it the device being turned off? or going into active mode? or another close-by device having a filter on the power lines which is absorbing e interference?

That is unlikely to show it as there are large numbers of FEC errors. So the bins are still being filled however the data is being corrupted which requires the correction.

I am vaguely wondering if there is a fault with the heating that is having the opposite effect to what you would normally expect. The timings of lower errors occurs at the sort of time I would expect heating might be on - perhaps an electric fire somewhere that when turned on full blast causes no noise but when essentially off is causing an overload through a resistor somewhere.

I know it seems a long shot but when things are running full power they can sometimes be fine but when "throttled down" can cause issues. Light bulbs can cause no interference at full power but when dimmed down to a certain level can cause interference. If the device is not fully turned off at the mains then the method of it being "off" could itself cause interference.

EDIT : In fact a timer clock itself could cause this? The timer itself is turned on at the mains. But, whilst the timer is off it is doing something to stop current being passed on - if it is a dodgy timer this could cause interference that may disappear when that block is released.

Edited by ian72 (Fri 29-Apr-16 09:47:55)

Go back to some of my previous comments and they are in the same vein.


For example, a power supply when unloaded can be very "noisy", put it on load and the noise is quashed.. Or noise from one device is being suppressed by something else turned on for those hours - which is almost certainly on a timer.

The OP is going to turn off everything and see if there is any effect ... if that fails it points to the potential for it being outside and as Zarjaz says they are even harder to locate - he has yet to be taught the full Ju-Ju dance though!


The OP has checked for timers and has yet to find anything as far as I can remember.

Yes I've checked for timers and turned off most stuff in the house to no avail. I'm planning on turning everything apart from router off on Sunday morning (at the fuse box) when i can get the wife and kids out of the house for a few hrs.

Good point.

If the OP can do it, he might the ask the engineer to do a DLM reset, that'll lose the error correction and then maybe it'll be visible ?

Oh, and it'll be Sod's Law the engineer attends when the source is off.

In reply to a post by Zarjaz:

Good point.

If the OP can do it, he might the ask the engineer to do a DLM reset, that'll lose the error correction and then maybe it'll be visible ?

Oh, and it'll be Sod's Law the engineer attends when the source is off.

not much chance ... the off period, or when it is suppressed is 18:55 to 22:25

Have a very detailed look at Weds and Thurs evenings ... at 22:00 there is a little blip too. It may be coincidence, however it will be interesting to see if it occurs this evening too. And it may have been there on Sunday too.

Oh yeah ......

I wonder if WW wants to rent me radio and 444 meter off me ? Competitive rates

I will do, cheers

The DLM was rest on Monday when I migrated from Plusnet to BT. I'm seriously considering cancelling the OR visit now due to RobertoS comment about a possible charge.

I must add though that I didn't ask for an engineer visit, the guy on the end of the phone did a test, and said that there was a fault and that he was going to send one. Does that mean I could still get charged - even though they say there is a fault?

I must add though that I didn't ask for an engineer visit, the guy on the end of the phone did a test, and said that there was a fault and that he was going to send one. Does that mean I could still get charged - even though they say there is a fault?

It depends on what fault they think they saw, what type of visit they have booked the engineer for.

I'd say don't cancel.

At very least you'll be able to see a PQ test done on your line .... earwig over the engineers shoulder, ask them to explain each of the results as they come in (if they don't know, well then, I'm sorry) Pay especial heed to a nice balanced pair of results (no more than 5 nanofarad's difference) between the A and B legs, and a good high A/C balance, preferably over 60db, this being your pairs ability to repel external signals (amongst other things).

They should put the tester in manual sync mode (the scripted one they are obliged to run ain't worth a [censored]) look at the errors screen, hopefully they will be piling in, and they'll have something to chase.

I wouldn't worry about a charge, if raised, you could always argue that the CP raised the task.

As far as I've heard, you don't get a charge for no fault found. Only if a fault is found to be within your home set up.

Depends what actual fault the CP is raising a visit for .

I know on the BT forums, the moderators always advise customers that no charge will be raised if no fault founds. They don't specify what type of engineer visit, so I can only assume all.

It may well be retail absorbing the cost for some visits

I know on the BT forums, the moderators always advise customers that no charge will be raised if no fault founds. They don't specify what type of engineer visit, so I can only assume all.

It may well be retail absorbing the cost for some visits

Aww come on lee, I am 95% sure you are on the tools ......

Experience should have made you at least doubt the sense in your statement.

I'm not on the tools and never have been. I've not worked for retail for a number about 4 years now but you're right, I should have known better haha.

Still curious as to why the moderators over there always advise there's no charge if no fault found...

In reply to a post by MHC:

The problem with errors though is continuous apart from the 3,5 hours every evening. That points away from heating - who leaves heating and hot water on 24 hours/day?

The two questions are: what is causing the errors? And what is suppressing them in the evening? which will the lead to: is it the device being turned off? or going into active mode? or another close-by device having a filter on the power lines which is absorbing e interference?

What happens to the noise margin when the error rates rise and fall?

Stays the same.

If errors are caused by noise I would expect the SNR(M actually) to fall while they are occurring, and rise when they stop.

Which suggests to me something else is going on in the router causing it to screw up in some way . Is there any sort of timed backup takes place? Or remote scan of a computer on the LAN?

I don't think so. It's a Billion 8800nl and I factory reset it on Monday as I wanted a fresh start with BT. I will take another look now though.

Checked router and can't see anything. I have uploaded images of my graphs for FEC, SNR andBitSwaps for the last 24hrs so you can see what's happening. They are here.

And looking on MyDSLWebStats at your SNRM/band - they are consistent at 6dB with the expected short burst of noise or interference. There is nothing matching the evening FEC rate change.

I've managed to ship the wife off for a few hrs with the kids now so I'm in the process of turning everything off to see what happens!!

Looks as though you were unsuccessful in your hunt! And that the 19:00 change has also occurred again!

Cancelled engineer visit yesterday but OR turned up today anyway! I happened to pop home from work to get a fibre patch cable and van was outside. The guy said it wasn't chargeable job but that he hadn't found anything wrong. Back to the drawing board I guess.

One thing I did notice though is that the noise disappeared early yesterday - at 16:34 instead of 19:00. As yesterday was a very warm day, this says to me that it may not be somebody's heating coming on. Doesn't help much but least it's one thing probably ruled out.

I saw that ... although the noise did come back for a short period fro around 18:40 to the usual 19:00.

I presume all your tests at home have show nothing ...

Unfortunately not. Turned off everything but no joy at all.

Also, should G.INP have kicked in by now? It's been a week since my switch-over to BT from Plusnet.

Well I decided to power off the Billion 8800nl for 30 mins at about 19:52 last night. Turned it back on about about 30 mins as wife was moaning she needed internet for work.

I noticed that the noise has been negligible since I turned router off!

I think I might have a dodgy router - at least I hope that's what it is!! I bought an unlocked HG612 from eBay last night so time will tell.

I have just given the router a reboot to see if I can now regain my speed.

Edited by WelshWArrior (Tue 10-May-16 08:19:54)

I bought a HG612 off eBay and guess what? Errors are no longer! Its only been plugged in since about 9pm but its now past the usualy 22.30 noise return and still nothing.

Only thing is though, I've lost exactly 10meg off my sync speed. Before I unplugged the 8800nl, I was sync'd at 59Mbps and ow I' sync'd at 49MBps with the max. attainable also dropping from 10Mbps from 69Mbps to 59Mbps.

Any ideas why that is? Can I expect sync to go back up to what it should be now the noise is (hopefully!!) gone for good??

I did noticed that Interleaving when viewed online through MyDslWebStats is 969. Sounds high - is that normal??

Yes, interleaving depth on FTTC can often be into 4 figures, and interleaving also typically drops the sync by about 10Mbps.

Two things are strange though.

One, it is also normal for attainable to increase considerably when interleaving is introduced. In other words the sync and attainable move in opposite directions, and converge to very close together figures when it is removed. Both having fallen is unusual.

Two, I wonder why your weren't on interleaving using the Billion. Do you happen to know which make of FTTC cabinet you are on? Has your latency risen by 8ms?

Edit - I wonder if you were on G.INP and the modem swap has removed that? In which case it should return in time.

Edited by RobertoS (Wed 11-May-16 23:26:32)

Looking at my history on dslstats, it seems that interleaving has actually dropped from 1,123 before 9pm to the current 969. G.INP hasn't been on since my switch from Plusnet to BT on 25th April. That's when the DLM was reset, I dropped from 69Mbps to 59Mbps in actual sync and the issue began.

I am on a Huawei cab. Latency had already risen during the migration to 8ms. It is still 8ms.

If you ave access to MyDslWebStats, I'mTenbyBoy2