Kitz Forum
Broadband Related => FTTC and FTTP Issues => Topic started by: jizzerman on January 29, 2015, 02:56:19 PM
-
Sorry, for butting in, but I have also been driven nuts by interference from my mains supply.
I have tried Tacima mains filters and RFI cleaners for the router/modem power supply.
Removed powerline adapters.
Put ferrites everywhere.
Shielded cable from my new MK3 faceplate to the modem, but still the jittery SNRm especially during the mornings and evening, when we are all active at home.
I have a REIN case open luckily and the engineer has been trying to help, but no luck yet.
Does anyone know if BT will consider changing the cable feeding into my house to a shielded one?
As I recon my master socket is right on my downstairs ring main cabling, which is why it's so sensitive.
My HG612 stats are online should anyone wish to compare.
Cheers.
[Split from "DLM Banded Profiles" topic - Admin]
-
I don't think there's any chance that Openreach would fit a shielded drop cable.
What you need to do is establish what kind of interference it is, and then try to find the source. 24-hour monitoring of the connection can be very helpful in diagnosing these sorts of issues. If you have a monitoring program running, or can install one, the first port of call would be the SNR margin graph.
-
I'm running the stats program and upload them to mydslwebstats, it's great for checking what's happening!
There are regular swathes of interference every morning and most of the evening.
I know my boiler kicks in at 6.30am, 11.30am and 16.30pm for the hot water provision, which coincides with about a 1dB drop like clockwork, but I am not sure how else I can protect my line from it.
I have attached a 24hr sample, some days my SNRm will drop entirely for about 30mins, causes re-sync at a much lower rate, and a reboot a bit later is needed to get the re-sync back to normal.
The BT cable comes from underground right next to my front door, goes through the wall about 18 inch thick straight into my MK3, through some CAT5 put in by the house builders I recon. (brand new house)
I am thinking about putting a metal plate between my wall and the grey plastic cover they have screwed to my wall and changing that tiny cable for a shielded equivalent.
What you recon?
-
A few questions Jizzerman
1. is the FTTC Modem situated beside the Master Socket ?
2. is the Master Socket very close to any mains sockets or mains cables ?
-
Hi,
1. the FTTC modem is about 1m from the mains socket
2. the master socket is very close to the mains power, about 6 inches away from a double socket.
For all I know the ring mains cable might run right next to the twisted pair coming in from outside in the wall behind the socket.
Thanks.
EDIT: Added pic of how close to mains :/
-
I have tried all sorts and the REIN/crosstalk continues.
Ran a new shielded cable through the wall, and even put that in a steel sleeve, to be totally sure.
Every 65 minutes it drops by 2dB for a couple of minutes, then once in a while its catastrophic and drops me to 25MBit.
Have some interesting graphs I'm hoping someone would be kind enough to interpret for me?
Up until last night I was getting 3 million FEC/min for about 2 weeks, then it went back to normal for a reason that must be external to me.
REIN engineer determined it was OK, but I'm not so sure. What's even left to be done? £20k for BT FTTP?
Cheers.
-
Both graphs show that there is something distinctly wrong with (the DS of) your circuit. The QLN graph is not something that I have ever seen before. :o
I am at a complete loss as to what to suggest. :(
-
Both graphs show that there is something distinctly wrong with (the DS of) your circuit. The QLN graph is not something that I have ever seen before. :o
I am at a complete loss as to what to suggest. :(
Noise on the Downstream tones but not on the Upstream tones that's interesting any twisted pairs should run 90° from a mains cable like a criss cross never have the pairs in parallel with mains wiring.
-
Perhaps a fault in the cab where the DS comes from. The US from the modem is fine?
-
When I had a HR fault it only seemed to affect Downstream at first, and when it rained the upstream was also affected.
Just a hunch in your pic you have that transformer plugged in and switched on, but nothing connected. Just wondering could that be the cause of your issues as its very near the NTE?
-
I wonder if it could be a fault with the port on the DSLAM line-card?
Perhaps you could try giving the port "time to relax"? (N*Star will know what I mean. ;) ) Late one evening, when all Internet usage is over --
- Power off the router, thus dropping the PPPoE session with your CP/ISP.
- Power off the modem (so that it sends its "dying gasp").
- Disconnect the cable linking the modem to the SSFP.
- Next morning, power up the modem and allow it to initialise.
- Reconnect the cable linking the modem to the SSFP.
- Once synchronisation has occurred between the modem and the DSLAM, power up the router (establishing a new PPPoE session).
-
Just a hunch in your pic you have that transformer plugged in and switched on, but nothing connected. Just wondering could that be the cause of your issues as its very near the NTE?
This is the Tacima mains cleaner device I bought to try and suck up any loose RFI floating about in the mains, all the gear is plugged into a Tacima filtered plug bar too.
Spending a lot, getting no where. ???
The mega drop that took place when the modem last re-synced, and provided the entirely busted QLN DS graph was brief and came back a bit later yesterday a couple of times close to each other. My hunch is someone is aiming a REIN machine at the bundle of pairs into my end of the estate!
I will try the "cool down period" sounds like a plan. I have a long outstanding call with PlusNet, I am hoping they will just lift and shift now, it's the last thing left.
Cheers for taking the time to check it out.
-
Turned it off overnight. It re-connected back at normal 40/10. Then it REINed about 9.50 and dropped it again to 25/10.
Gave the same crazy QLN during that re-sync.
But the REIN that comes every 65 minutes seems to have now stopped.
Any ideas what pumps out an EMP every 65 minutes all day and all night? And wrecks the DS only.
-
Any ideas what pumps out an EMP every 65 minutes all day and all night? And wrecks the DS only.
A fridge/freezer could do this as the motor switches on, when my fridge was new it would be every hour but as it's a year old it's more like every 35 minutes bar any attemps to gain again access via the doors that is.
-
It's a good shout. But seems too precise, and we haven't changed any white goods and the issue came for 3 weeks, and stopped this morning. (hopefully)
And my beer is still cold..
Parts of the estate I live on are still being built, I'm beginning to wonder if their heavy equipment is causing it? Might explain why the big drops are only in the daytime, usually mid morning.
I also wonder why fibre wasn't blown to all the new houses :)
Perhaps a fault in the cab where the DS comes from. The US from the modem is fine?
I have not had an update from PlusNet since it was re-opened for the 3rd time this week.
Hoping OR lift and shift.
-
What would make the DS D3 be 0 for long periods and jump around?
-
At a guess, I will say a physical defect in the circuit. Perhaps a defective joint, a joint that can open and close when subject to vibrational disturbances?
Nothing appears to "jump out"; there is no figurative "smoking gun" present. :(
-
When OR came to activate my line originally, new build house, they said the line back to the cabinet wasn't installed long enough, and had to be extended.
Whatever it was only took a day to sort out, then all was fine. As did my neighbours.
I don't think my FTTC has ever been right, but it's only recently I have been checking.
Would that have any bearing on it?
-
The QLN graph is not something that I have ever seen before. :o
Jizzerman's QLN graph does not display this kind of characteristic all the time - the one visible on MDWS right now is pretty normal.
However, we have seen a graph a bit like it before, though at a slightly reduced amplitude (mine oscillated by 5dB, while OP's was 10-20dB).
When my line resynced strangely just over a week ago (discussed on this thread (http://forum.kitz.co.uk/index.php?topic=15094.0)), I got a fairly similar QLN graph too.
That can be seen on MDWS too - if you choose "Quiet Line", then "Historic", then "1st Mar". It was back to normal on the "2nd Mar" after I re-synced manually.
We pondered whether this happened at a time the DSLAM was re-booted (perhaps G.INP going live), so the QLN display looked odd while many modems were re-syncing.
Turned it off overnight. It re-connected back at normal 40/10. Then it REINed about 9.50 and dropped it again to 25/10.
Gave the same crazy QLN during that re-sync.
If it was that severe, perhaps it affected lots of other modems, and they all re-synced at a similar time. If the QLN is a side-effect of that kind of thing, that would imply it isn't just your line!
What would make the DS D3 be 0 for long periods and jump around?
Noise in the frequencies of D3 - enough to make the modem choose to not use any D3 tones during the sync.
I took a look at your SNRM/band graph on MDWS, which showed an event like this happening on the morning of the 7th March.
Then I compared your SNR/tone graph between "Mar 7th 02:00" and "Mar 7th 12:00" ...
... and Wow! Both D2 and D3 look horrible, with dips to 0dB every 12 tones for long stretches. The dips are 1-2 tones wide in D2, but 4-5 tones wide in D3. The part between tones 360 and 460 look affected too.
The matching bits/tone graph show that the modem has indeed chosen to not use a lot of tones (esp D3).
Yeuch. (Professional assessment, that ;) )
At a guess, I will say a physical defect in the circuit. Perhaps a defective joint, a joint that can open and close when subject to vibrational disturbances?
Nothing appears to "jump out"; there is no figurative "smoking gun" present. :(
I think those dips every 12 tones is a distinct smoking gun - it seems very repetitive, and suspiciously close to 50Hz - and I'd show that to a REIN guy if he ever turned up again. (Grab copies, in case we lose MDWS!)
But I haven't a clue what is causing this. It makes me think of resonance, and fast fourier transforms, and many other things long-ago forgotten.
-
When my line resynced strangely just over a week ago (discussed on this thread), I got a fairly similar QLN graph too.
That can be seen on MDWS too - if you choose "Quiet Line", then "Historic", then "1st Mar". It was back to normal on the "2nd Mar" after I re-synced manually.
Exactly the same thing happened to me a few days ago. The connection dropped around 00:54 for no apparent reason, and stayed down for 6 minutes. When it came back up the QLN was all fuzzy. I guessed that the 6 minute downtime was the result of engineering work in the exchange.
Look at my historic QLN on MDWS for 6 March 00:00.
-
The connection dropped around 00:54 for no apparent reason, and stayed down for 6 minutes.
I just checked my modem_stats.log file: the entries either side of the resync are for 27th Feb, 00:18 and 00:25. It seems like a similar 6 minute outage.
-
It does begin to look like a pattern. Pure speculation, but could it be that the DSLAM configuration/software was being updated for G.INP support?
-
I'm pretty sure this has been a problem since my line went live with FTTC about June time last year.
Although I was not monitoring until I got a DSL-AC68U, and initially blamed that device's inherent instability. Rolled in the HG612 to find it doing the same.
I had OR here on the weekend, but unfortunately the problem was not apparent while he was here, 20 minute bursts randomly through the weekdays is hard to be here for.
I had plenty of history and data but there was no interest in seeing it as no REIN detected while he was here. I was hoping for a pair swap, or port swap just to rule them out.
The GEA test he did before he arrived showed REIN, but not by his kit when he was here, fair enough. I have attached the test PlusNet did Monday last week that got the engineer booked.
I think moving house to where the cabinet is on my door step is my next move :( Drastic, I know.
It went down today at 11.15, same time as it did Monday last week... but no guarantee it will re-occur this coming Monday.
Is it possible for me to pay for a lift and shift, even just a pair swap?
I think those dips every 12 tones is a distinct smoking gun - it seems very repetitive, and suspiciously close to 50Hz - and I'd show that to a REIN guy if he ever turned up again. (Grab copies, in case we lose MDWS!)
I'm going to try shutting my heating down tomorrow, as it's the only timed device that I have.
Then I'll try full dark later in the week.
Appreciate your analysis guys, it's positively driving me nuts, I'm on brink of just scheduling daily reboots and moving on :baby:
-
Jizzerman look at my errored seconds 1500+ per day and comimg up to a month this tells me your ISP has a low level of MTBE set on your line :-\
But your US errored seconds needs looking into as they are much higher than the DS
-
I had a DLM reset on Saturday, DS error seconds went up massively.
DLM put low level interleaving back on early Sunday morning.
DS ES recovered, no interleaving on US, it's always been happy with level of error seconds upstream, must be just in their amber level perhaps.
Left my heating off, still had a couple of heavy SNRM drops this morning, looks like sync just held out.
Pretty sure interference is external. Is there a way one of the monitoring tools could email me when the DS SNRM drops below 5dB?
I have an analogue radio to hand to go ghost hunting :)
-
Is there a way one of the monitoring tools could email me when the DS SNRM drops below 5dB?
Yes, DSLstats can do this. Look at Configuration --> Alerts.
-
I'll take a look tonight at getting that up and running, as it just dropped sync again.
If I can catch it in the act and record it maybe OR will have some useful evidence to get some traction with.
-
UPDATE:
Plusnet escalated my ticket last week and arranged an engineer for today.
Then G.INP was turned on early Sunday morning. Great improvement on error rate and reduced my latency to bbc.co.uk from 17 to 10ms, wikid.
I hoped it would clear my REINy days, but no, Monday and Tuesday re-syncs both days with to a lower sync speed and a dirty QLN graph.
Rebooted both days during clean time to get the speed back.
OR engineer came today, my 5th visit, he was a great guy. Took the time to listed to what's been happening so far and what had been tried.
Decided to run a new pair from the cabinet to my door!
Stats showed a vast improvement to the interference on my QLN immediately.
He also said the AC balance had gone from 61 to 75, which apparently should improve it's resistance to REIN.
I asked for a DLM reset as a final thought, thinking it would put my line on a nice clean slate, but it actually removed G.INP and applied a deeper level of oldsk00l interleaving :-\
Hopefully the low error rate will settle itself back down in a week or so?
-
I asked for a DLM reset as a final thought, thinking it would put my line on a nice clean slate, but it actually removed G.INP and applied a deeper level of oldsk00l interleaving :-\
Hopefully the low error rate will settle itself back down in a week or so?
That's interesting a line reset takes you off G.INP i wonder how long it will take your line to go back to G.INP :-\
-
That's interesting a line reset takes you off G.INP i wonder how long it will take your line to go back to G.INP :-\
makes sense since a DLM reset is the cure to outdated ECI modems not liking G.INP,
-
makes sense since a DLM reset is the cure to outdated ECI modems not liking G.INP,
Jizzerman is using a Huawei cabinet so your theory is not holding up in this case.
-
makes sense since a DLM reset is the cure to outdated ECI modems not liking G.INP,
Jizzerman is using a Huawei cabinet so your theory is not holding up in this case.
The cabinet is irrelevant. Outdated ECI modems don't support G.INP, so the OR instructions are to reset the DLM on the line, which will allow a connection to be established, which will then allow the modem firmware to be updated to support G.INP. That supports the suggestion that resetting the DLM (temporarily) removes G.INP if enabled.
-
I see what your saying ->
The rate of deployment is now increasing. It is therefore more likely you will encounter the below scenario if you are repairing a fibre circuit and have to replace the Openreach modem.
What’s changing?
VDSL retransmission is now being deployed at a rate of around 45,000 circuits per day
It may not be obvious that you’re carrying out a repair on a fibre ReTX enabled circuit with a retransmission profile applied. And it only becomes important if you have to replace a faulty Openreach modem. If the replacement modem is ECI (indicated on the box or base of the modem) it’s likely the firmware is incompatible with ReTX and a workaround needs to be applied.
An ECI modem with old firmware will appear to be in sync (the DSL light will be on) but the customer will not be able to access the internet and this will also be reflected in your GEA service test result:
For example “VDSL modem is not in sync”.
What you need to do…
If you complete a DLM reset it will remove the ReTX profile and allow the modem to sync up and then the system will download the correct firmware before ReTX is reapplied. You do not need to wait for the download, it will happen overnight. As you will know, a DLM reset is only permissible when you have found and fixed a hard fault on the circuit - and a faulty modem is a hard fault.
It is important that you only use the DLM reset procedure, found here, when you clear a hard fault or experience a no sync situation after replacing a modem on a fibre repair task. The use of DLM resets will continue to be monitored.
Imagine all the ECI modem users asking for a line reset by a Oprenreach Engineer just to get G.INP it would cause chaos ;)
-
Thanks for the updates, glad I have no ECI in my life :)
Seems like it should sort itself back out pretty quickly.
Is G.INP applied even if your line is stable enough for FastPath?
-
Is G.INP applied even if your line is stable enough for FastPath?
Thats a good question and your stats overtime will answer that.
-
Is G.INP applied even if your line is stable enough for FastPath?
G.INP was enabled on my short and relatively error-free line on Monday (was fastpath for months). I suspect anyone's line who falls into the 'amber' category will get it enabled - but there's every chance BTw are enabling it for everyone, because I can't see that there's a downside. Speed and latency are the same, and SNR margin has increased slightly to 12dB (although I suspect this is a facet of new firmware applied at the DSLAM rather than anything G.INP is doing)
-
G.INP was enabled on my short and relatively error-free line on Monday (was fastpath for months).
Looks like it'll be G.INPing all round then. There was no change to mine last night.
Hoping it get's picked up soon.
My SNRM is strangely stable at 6dB, normally it would be bit a higher and swing by a dB or two during the evenings.
Would the forced interleaving be maintaining this, straight like an arrow profile?
-
I suspect anyone's line who falls into the 'amber' category will get it enabled - but there's every chance BTw are enabling it for everyone
What is amber category? Thanks
-
There are two thresholds used when DLM looks at the ES counter over a 24 hour period.
If you are above the "red" threshold, this will trigger DLM intervention overnight.
If you are below the "green" threshold, DLM will kick the de-intervention mechanism, although it takes a few days of being below the green threshold for de-intervention to actually happen.
Inbetween these thresholds, you are in amber territory - where DLM does not get triggered in either direction.
That's all true for old-style DLM. However, DLM is being changed for G.INP - and we don't know how far the changes cascade.
-
G.INP was turned back on at 8.56am this morning.
The SNRM went from the solid 6 back to the normal 8.5.
Plusnet ran a new GEA test today, which returns with no crosstalk or REIN detected, these have been positive in my previous tests.
I'm really hoping the new pair has finally put the issue to rest. Will keep an eye over the next week or two.
Either way, with that spare SNRM it might be worth going back up to 80/20 package, as i might get 50 down and 15 up with any luck.
But I won't rush it :)