Kitz Forum
Broadband Related => FTTC and FTTP Issues => Topic started by: jt999 on October 04, 2017, 11:59:19 AM
-
Hi all,
I've been trying to read up more and more on DSL but to be honest, it's just very confusing to me :'(
I used to get speeds of 50-60Mbps but went through a period earlier in the year where my line was dropping a lot and it got slower and slower. BT Openreach got involved and an engineer visited the cabinet to run a test, he found no faults but since then my line became stable (he insisted nothing had changed). It did recover a bit to around 20-25Mb, I then changed ISP to Zen, and whilst I'm happy with the service my line is still fairly slow (23Mb-ish). Can anyone take look at my line on MDWS (jt999) and tell me if I'm banded or any other reason I shouldn't return to a decentish speed?
I had a few power outages in the area a while back (August 17th, 22nd, and Sept 7th) which I assume caused the BT cabinet to reboot also, they are noticeable on the sync rate graphs as the short increases in speed after a power cut (I then rebooted the router each time as MDWS was saying my sync rate was higher than my attainable).
Many thanks
jt999
-
Doesn’t look to be banded.
23648kbps against an attainable of 26976.
What is the speed estimate for your line shown on BT Wholesale’s DSL checker?
Your attenuation of almosg 29dB would be about right for your current speeds.
-
Doesn’t look to be banded.
23648kbps against an attainable of 26976.
What is the speed estimate for your line shown on BT Wholesale’s DSL checker?
Your attenuation of almosg 29dB would be about right for your current speeds.
I've attached a screenshot of the checker as I ruined the formatting trying to paste! :)
Even the low for impacted seems higher than what I actually get, not sure what this really means though
-
Seems you need another engineer visit to find out what’s wrong.
Seems you should be getting quite considerably higher speeds.
-
Seems you need another engineer visit to find out what’s wrong.
Seems you should be getting quite considerably higher speeds.
Thanks, at least it's not just me going crazy then. I might give it another few weeks to see if any speed increases and then contact Zen to see if they can help.
-
Nothing will change without physical intervention.
-
Nothing will change without physical intervention.
Ahh Ok, thanks, will raise it with Zen in that case :)
-
Nothing will change without physical intervention.
Zen ran a test and can see a copper cable fault or something, unfortunately they want me to check for a dial tone and I never actually bothered to buy a phone. I'll have to get one of those and test, then see what happens from there!
-
Have you checked the sync speed when connected to the test socket of the NTE master socket?
This helps rule out any internal wiring issues.
You used to get 50-60Mb? I've never seen an attenuation as high as yours even get 40Mb without G.INP and vectoring. Your connected to an ECI cabinet so that seems even more unlikely.
Something physical likely changed at some point that's impacted your speed and your attenuation.
Your estimates also suggest this, as you shouldn't have those estimates with such a high attenuation.
-
Have you checked the sync speed when connected to the test socket of the NTE master socket?
This helps rule out any internal wiring issues.
You used to get 50-60Mb? I've never seen an attenuation as high as yours even get 40Mb without G.INP and vectoring. Your connected to an ECI cabinet so that seems even more unlikely.
Something physical likely changed at some point that's impacted your speed and your attenuation.
Your estimates also suggest this, as you shouldn't have those estimates with such a high attenuation.
I haven't tried the test socket lately (only for the original line fault) but I can perhaps try it, I'd be surprised if it helped though as I don't have anything else connected to the phone line, and it's only the master socket, so no extensions. The modem is on a 5m cable so I can perhaps try changing that for a short one temporarily.
Would I be likely to see an attenuation drop fairly quickly if it is a wiring fault? Just it'd be quite difficult to leave it cabled to the master socket on the short cable for more than a day or two.
Up until around May (I thought it was earlier in the year, but checked and I have emails of it being in May) it was much higher than it is now but unfortunately have no logs aside from some old Speedtest app results on my phone (I have one for February 18th which is 49.2Mb down and 12.26Mb up, and again on February 20th of 48.83Mb down and 10.11Mb up - both connected to my Wireless AP).
-
I just changed to the shorter cable, unfortunately I couldn't find my spare filter to try using the master socket (I have the NTE faceplate thing installed normally). Speeds did go up a little, but not a noticeable amount so perhaps just due to rebooting everything. I'll try to find my spare filter or order some in.
-
You need to try the quiet line test.
-
You need to try the quiet line test.
I'll pop out later and buy a phone if I get time and try that.
Thanks for all your help so far! I'm sorry I'm so useless with this! :(
-
Quiet line test seemed to be OK (it took a few attempts to get a dial tone, but I bought the cheapest corded phone I could so perhaps it's to be expected with the quality), BUT when I plugged the phone in my DSL resynced and jumped massively in speed, and the attenuation dropped... I don't really understand why?
-
Also another update for those interested, I've had 0 errored seconds since then. It's most odd but I'm obviously very happy with it!
-
So it's an internal wiring issue. A drop in attenuation from 29dB to 19.6dB is huge and shows a problem.
Have you tried connecting directly to the test socket behind the front cover of the master socket?
-
So it's an internal wiring issue. A drop in attenuation from 29dB to 19.6dB is huge and shows a problem.
Have you tried connecting directly to the test socket behind the front cover of the master socket?
I'm not sure how it can be an internal wiring problem, the vDSL modem is connected the same way today as it was yesterday and has not been touched. The only change is that a phone is now also plugged in (it's one of the Openreach faceplates with 2 sockets, so the DSL cable wasn't even touched). MDWS gave a reason of "Reason: 1 Remote Defect Indicator/DLM" when the resync happened today, so it looks like it was updated from the DLM. Perhaps I'm barking up the wrong tree but I've ordered some higher quality DSL cables to try out anyway.
I'm trying to find a vDSL filter to try using the master socket, but no luck so far, may have to order online but no idea of quality!
-
It certainly sounds like something is wired incorrectly. DLM doesn't massively reduce attenuation. That only comes from a physical change in the line.
Perhaps others may have seen something similar, or know what would cause an improvement in attenuation + causes a line to resync when a phone is connected.
My guess would be the master socket and/or the MK3 faceplate is wired incorrectly or perhaps faulty. A picture of the wiring would help.
-
An xDSL modem/router will not loop the pair and draw current from the exchange.
A telephone was added to the circuit and, when the handset was lifted, there was not an immediate "proceed to initiate a call" tone (a.k.a the "dial tone"). Eventually the expected tone was heard and a decrease in attenuation was noted.
As soon as I had processed the above information, I got a "tingle in my whiskers". Why? Because I strongly suspect that there is a defective joint somewhere in the circuit . . . a joint that is either exhibiting semi-conductive or HR (high resistance) tendencies.
Where would that joint be found? Anywhere between the NTE5 and the serving PCP . . . somewhere where two dissimilar metals are in contact and also damp.
-
It certainly sounds like something is wired incorrectly. DLM doesn't massively reduce attenuation. That only comes from a physical change in the line.
Perhaps others may have seen something similar, or know what would cause an improvement in attenuation + causes a line to resync when a phone is connected.
My guess would be the master socket and/or the MK3 faceplate is wired incorrectly or perhaps faulty. A picture of the wiring would help.
I can provide a picture, but not sure what wires you want a picture of - do you mean if I take the entire master socket off the wall? The NTE itself doesn't seem have any wires, it just seems to plug in. It is a Mk3 as you suggested.
Too add to your thoughts, where the line comes into the property there is a small box that the OpenReach engineer did replace when I first had vDSL installed, to be honest it looks like a junction box or something, just joining the internal wiring to external - I guess it's possible he didn't wire it right?
-
An xDSL modem/router will not loop the pair and draw current from the exchange.
A telephone was added to the circuit and, when the handset was lifted, there was not an immediate "proceed to initiate a call" tone (a.k.a the "dial tone"). Eventually the expected tone was heard and a decrease in attenuation was noted.
As soon as I had processed the above information, I got a "tingle in my whiskers". Why? Because I strongly suspect that there is a defective joint somewhere in the circuit . . . a joint that is either exhibiting semi-conductive or HR (high resistance) tendencies.
Where would that joint be found? Anywhere between the NTE5 and the serving PCP . . . somewhere where two dissimilar metals are in contact and also damp.
That sounds like it'd be a nightmare to get investigated :(
I'm happy with the current speeds, just hoping it remains that way. Would you suggest I leave a phone attached to the line? It'll be annoying but possible (the master socket is in my 9 year old sons room, I didn't really want a phone in there! :))
-
That sounds like it'd be a nightmare to get investigated :(
Intermittent faults are indeed the hardest to troubleshoot.
A change in attenuation is something that should never happen ... but when it does, it also seems to be an indicator of one of these tricky cases.
The easiest thing for you to do yourself is to take the faceplate off, and run the modem through a rat-tail filter, and just leave it like that to see if you see the same behaviour (now, or at some point in the future).
I'm happy with the current speeds, just hoping it remains that way. Would you suggest I leave a phone attached to the line? It'll be annoying but possible (the master socket is in my 9 year old sons room, I didn't really want a phone in there! :))
The thing that might have "fixed" things temporarily is the current that was drawn by the phone (after a few goes); this current can have "burnt through" a layer of oxidation, to make a proper circuit again ... and if so, the current is called a "whetting current".
It is never more than a temporary fix, as only a small amount of the oxidation will have been removed, and it will soon build up again, so cause problems again.
When it does go wrong, then you might need to use the phone again. But you don't need to leave it in place 24x7 ... only when you need it. Even if you choose to go off-hook every day, hoping to act as a preventative.
-
Intermittent faults are indeed the hardest to troubleshoot.
A change in attenuation is something that should never happen ... but when it does, it also seems to be an indicator of one of these tricky cases.
The easiest thing for you to do yourself is to take the faceplate off, and run the modem through a rat-tail filter, and just leave it like that to see if you see the same behaviour (now, or at some point in the future).
The thing that might have "fixed" things temporarily is the current that was drawn by the phone (after a few goes); this current can have "burnt through" a layer of oxidation, to make a proper circuit again ... and if so, the current is called a "whetting current".
It is never more than a temporary fix, as only a small amount of the oxidation will have been removed, and it will soon build up again, so cause problems again.
When it does go wrong, then you might need to use the phone again. But you don't need to leave it in place 24x7 ... only when you need it. Even if you choose to go off-hook every day, hoping to act as a preventative.
That's very interesting! Thank you all so much for taking the time to reply to me.
I assume a rat tail filter is just a normal filter that hangs out of the socket? If so do you have any recommendations on a decent quality one? I did get one with my Zen router but it appears to have gone missing.
-
Well today has been annoying, the attenuation is fluctuating. Today a quiet line test is also quite crackly. I'll raise it with Zen :)
-
Attenuation shouldn't fluctuate. Sounds like a line fault
-
Attenuation shouldn't fluctuate. Sounds like a line fault
They couldn't find a fault, but have agreed to get a BT engineer to come out, the line is quite crackly so there must be a problem. I've attached my last 48 hours attenuation, both times when the lowers it's when I've been on the phone to do a quiet line test.
-
Remember if an engineer is sent and it's a fault on your side of the master socket then it's a £130 charge.
You really need to connect to the test socket. That's what it's for.
Most modems come with 2 of the "rat tail" filters. Do you have a friend or a neighbour that you could borrow 1 from?
Does the line resync every time you use the phone? That suggests the MK3 is faulty or something is incorrectly wired.
-
Remember if an engineer is sent and it's a fault on your side of the master socket then it's a £130 charge.
You really need to connect to the test socket. That's what it's for.
Most modems come with 2 of the "rat tail" filters. Do you have a friend or a neighbour that you could borrow 1 from?
Does the line resync every time you use the phone? That suggests the MK3 is faulty or something is incorrectly wired.
It looks like a line fault as I plugged a phone into the test socket, and the modem was unplugged entirely for the test (I had no rat tail filter to use so couldn't actually plug it in if I wanted to). The phone line was very crackly at the dial tone or in a call.
I don't lose connection every time I used the phone when the Mk3 faceplate is on, but sometimes I do and it comes back with lower attenuation as seen on the graphs, I don't have to hang the phone up for the DSL to come back.
-
Reading and looking at the stats I can only think that the problem lies within the lower NTE5 or upper filtered faceplate on this circuit, reason being any insertion in that location is causing a loss of sync or different attenuation the A & B pairs to the back of the NTE5 needs re terminated something is loose
-
Reading and looking at the stats I can only think that the problem lies within the lower NTE5 or upper filtered faceplate on this circuit, reason being any insertion in that location is causing a loss of sync or different attenuation the A & B pairs to the back of the NTE5 needs re terminated something is loose
But I have crackling on the phone even when the faceplate (upper and lower) is removed and phone is plugged straight into the test socket - surely a faceplate issue would disappear when I unplug it? Unless I'm missing something (which is likely! ;D)
-
The more I look at the way your DS & US SNRM falls off over time after a resync says that line needs a Oprenreach appointment it's almost like a battery or capacitor slowly draining away it's voltage, if your up to it you could replace your Modem and the NTE5 & SSFP and see if that helps if not call your ISP.
-
The more I look at the way your DS & US SNRM falls off over time after a resync says that line needs a Oprenreach appointment it's almost like a battery or capacitor slowly draining away it's voltage, if your up to it you could replace your Modem and the NTE5 & SSFP and see if that helps if not call your ISP.
Openreach are coming Tuesday, so I'm hoping they can find/resolve it, if not it sounds like it'll be expensive! :(
-
Audible crackles should not get charged ... but they might only get investigated by a voice engineer rather than a broadband-skilled one.
-
Audible crackles should not get charged ... but they might only get investigated by a voice engineer rather than a broadband-skilled one.
I think it's a voice engineer coming, I'm not sure though (Zen did suggest we get the crackling fixed before worrying about the DSL, so I agreed to that). I'm sure they won't be able to deny a fault if they try to use the phone.
The more I look at the way your DS & US SNRM falls off over time after a resync says that line needs a Oprenreach appointment it's almost like a battery or capacitor slowly draining away it's voltage, if your up to it you could replace your Modem and the NTE5 & SSFP and see if that helps if not call your ISP.
I had a resync a short while ago (this time without using the phone) as the SNRM dropped to basically 0. Currently I have a rat tail filter (£10 in Maplin, what a rip off!) to the test socket and a new/overpriced DSL cable just to ensure none of the old stuff I had was the issue.
I see what you mean on the graph where it's falling over time - I've attached a screenshot for anyone interested.
I can perhaps change the modem to the one Zen sent originally if I can work out how to get it into a bridge type mode. I don't much fancy trying to replace the SSFP though as that'll likely end in tears! :)
-
That is bizarre.
-
That is bizarre.
Isn't it just! :D
Will report back tomorrow once BT have been out, hopefully it'll be straightforward enough that it can be repaired.
-
BT engineer came, he thinks there is a bad joint around 200m away, so is going to get someone else out to look at that as he can't do that work, progress at least! :)
-
All sorted hopefully, they had to move me to some new pairs for part of the line, and redo a joint a bit further down the line.
-
BT engineer came, he thinks there is a bad joint around 200m away, so is going to get someone else out to look at that as he can't do that work, progress at least! :)
Ah, so he wasn't trained for UG (under ground) work.
-
All sorted hopefully, they had to move me to some new pairs for part of the line, and redo a joint a bit further down the line.
Hopefully all will now be good with your broadband Internet access, carried on that physical circuit.
-
Ah, so he wasn't trained for UG (under ground) work.
No, he said he can only do above ground, but he got the under ground engineer within 30 minutes of establishing that there was a fault, so I was very happy - especially given the reviews I've heard of OpenReach elsewhere!
Line looks much better since the work was completed, I'm just hoping it stays that way :fingers:
-
No, he said he can only do above ground, but he got the under ground engineer within 30 minutes of establishing that there was a fault, so I was very happy - especially given the reviews I've heard of OpenReach elsewhere!
In view of that, would you be willing to go to the Openreach "Tell Us Something" page (https://www.openreach.co.uk/orpg/home/contactus/tellussomething/tellus.do), scroll down to the "Let us know how we did" heading and then take the first link in the following paragraph to send a "thank you" message?
-
In view of that, would you be willing to go to the Openreach "Tell Us Something" page (https://www.openreach.co.uk/orpg/home/contactus/tellussomething/tellus.do), scroll down to the "Let us know how we did" heading and then take the first link in the following paragraph to send a "thank you" message?
Thanks, I've done that now ;D
-
Thanks, I've done that now ;D
Purrfect. :)
-
DLM seems to be having some fun with my connection today, had 4 resyncs all "Remote Defect Indicator/DLM" but not sure why :no:
-
Hmm . . . perhaps ignore it for a day or two, then look closely.
-
It doesn't look promising. The line appears to still be showing the plummeting snrm prior to each resync.
That's certainly not normal behaviour.
The line also jumped in attenuation by 0.9dB after the last resync. All the waiting in the world won't fix that. The attenuation should not change like that.
edit: Looking further each time the snrm crashes there is 10,000's of CRCs. Fortunately it's only a small amount of ES. However such high error rates and subsequent resyncs may see the line banded in time.
-
DLM seems to be having some fun with my connection today, had 4 resyncs all "Remote Defect Indicator/DLM" but not sure why :no:
It seems to me that the majority of people see this reason, and immediately discard the "remote defect indicator" as being a valid cause.
A "Remote Defect Indicator" is a bit in the OH frames sent between modem and DSLAM; if the DSLAM triggers a resync because of an RDI it is because it has seen the RDI bit set to 0 in the incoming OH bitstream from the end-user modem. In turn, the end-user modem chooses to set RDI to 0 when it has detected a flaw in the synchronisation of the incoming bitstream from the DSLAM (labelled officially as a "sever errored frame").
So an RDI becomes a valid possibility when the end-user modem could be receiving a particularly bad bitstream.
It always struck me that, on a line with bad faults or noise, even if only intermittent, an RDI could be a more likely explanation than DLM.
DLM should only be suspected if there is also a significant change to other framing parameters. In older times, that would be just the INP and delay parameters, but now we need to also watch for target SNRM adjustments. It hasn't happened here, so I would discount DLM.
edit: Looking further each time the snrm crashes there is 10,000's of CRCs. Fortunately it's only a small amount of ES.
Something that causes tens of thousands of CRCs in a few seconds would be just the kind of thing that will bring out a "sever errored frame" and a consequent "remote defect indicator".
It doesn't look promising. The line appears to still be showing the plummeting snrm prior to each resync.
That's certainly not normal behaviour.
The line also jumped in attenuation by 0.9dB after the last resync. All the waiting in the world won't fix that. The attenuation should not change like that.
Definitely still something wrong...
-
It always struck me that, on a line with bad faults or noise, even if only intermittent, an RDI could be a more likely explanation than DLM.
The actual response AFAIAA is just 'Remote Defect Indicator' - I added the '/DLM' part back in 2014 when this all first ran on the assumption that the RDI triggered the DLM system to act.
I'll remove it preferred if that isn't what is happening.
-
There certainly looks to be times when it *does* indicate a DLM intervention, so it is probably OK to leave that indication in there. The only thing that has gone wrong is where people assume it only means DLM.
I've often wondered how DLM triggers the resync; forcing an RDI would indeed be one way, but I haven't seen how in a MIB.
-
You're all far too clever for me :(
Is there anything I can try to do to help it? Or are we talking another BT visit?
-
I tend to think that if the attenuation is varying, and we've already ruled out internal stuff (wiring, filters, routers) then it is something that can only bend to the call of the copper whisperers.
-
. . . then it is something that can only bend to the call of the copper whisperers.
b*cat approves of that turn of phrase. :)
-
To be safe I've switched to my ECI modem from the Huawei one to see if it helps, although it does mean I lose my monitoring so I'll have to hope I catch the outages manually. If I still see them I'll get back onto Zen.
Thanks again for your help everyone! :)
-
I switched back to my Huawei modem as had a drop on the ECI, but touch wood things are back to normal.
Is it possible DLM was reset and was trying to work out how my line was?
-
Is it possible DLM was reset and was trying to work out how my line was?
That would have been done when the technician concluded his work on the circuit. I very much doubt that you were "blessed" with another "circuit recalc" (to use the Openreach turn of phrase).
-
Definitely not.
The attenuation should NEVER change between resyncs like it did. This indicates your problem is not completely fixed. The plummeting snrm confirms this.
Keep the HG612 connected with DslStats uploading to MyDSLWebStats so you can show the ISP of it continues to happen.
-
OK will do, thanks again.
I seem to have bursts of ES, oddly about the same time as yesterday. Will keep logging for a few weeks to see if there is a pattern :)
-
Zen can't find a line fault, and have asked for me to collect stats for longer (which is understandable) to see if this is a pattern - I can probably live with it for a few weeks (I'm away for a week in a week, so that'll help). Will let you know how it goes I guess :)
They also mentioned this site by name, so that's something cool ;D
-
Just tried the quiet line test again, no noises, but it did again cause my DSL to drop and come back fast. I'm just going to move house :lol:
This time the reason for the drop was:
4 MIS: CReverb1Misdetection
-
FEC has gone crazy today, MDWS is showing over 2m a minute (if I'm reading that right!). I've tried to find anything that happened around that time, it's possible I put the washing machine on, so will have to wait for that to finish to see if it's somehow related :no:
-
It started at 1744 hours and reached a maximum of over 3 million at 2028 hours . . . it is still on-going. That is a long wash cycle! :-\
I would be tempted to power down the modem (or modem/router, if you use a one box set-up) and leave it off until tomorrow morning. I see that you are one of the unfortunates who is connected via an ECI Hi-FOCuS M41 DSLAM. :(
-
It's now finished (cycle takes about 5 hours 45 minutes!) and it hasn't yet dropped, so doesn't look related - I was just racking my mind for what changed around that time that I was aware of for such a drastic difference.
I'll try turning it off overnight and will report back tomorrow, thanks :)
-
It started at 1744 hours and reached a maximum of over 3 million at 2028 hours . . . it is still on-going. That is a long wash cycle! :-\
I would be tempted to power down the modem (or modem/router, if you use a one box set-up) and leave it off until tomorrow morning. I see that you are one of the unfortunates who is connected via an ECI Hi-FOCuS M41 DSLAM. :(
Turning it off for the night seems to have stopped the millions of errors now, will keep an eye on it :fingers:
-
My line seems quite stable at the moment (2 days 19 hours and counting). I find that if the SNRM drops a bit I can use the phone to sort it out - not a permanent solution of course but if it can last until my holiday so I can then re-engage Zen I'll be happy. ;D
-
Had to say it didn't I!
-
Things seem to be improving, not sure why. Very little ES, line isn't fluctuating as much, and interleaving is being lowered. So far so good, but wish I knew why it seems to be healing itself :no:
-
I notice a resynchronisation / retrain event at 1354 hours that has an RDI status.
-
I notice a resynchronisation / retrain event at 1354 hours that has an RDI status.
I saw that, but it's not really clear what caused it, I figured DLM was turning off interleaving - since when it came back the interleaving was gone
-
Ah, that was probably the reason.
-
Been away a few days, my line has been going mental. Will raise it with Zen! I had considered getting an old 56k modem and trying to dial numbers once a day, since a phone call a day seems to sort the issue out :D
-
Just to update you all, my line has been rock solid for past 14 days, and that's with only the odd phone call on the landline. I'm sure the issue will return at some point, but until it does I don't suppose anyone will be able to do anything about it.
Thanks for all the help you have all given! :)