Kitz Forum
Broadband Related => FTTC and FTTP Issues => Topic started by: matt12321 on June 27, 2016, 10:29:20 PM
-
Just installed (and unlocked) a HG612 to match the Huawei cabinet - not made any changes to the settings however, just plugged it in to the HH5.
My DSLChecker results are:
VDSL Range A (Clean) 29.7 22.4 4.9 3.2 -- Available -- --
VDSL Range B (Impacted) 22.3 10.4 3.8 1.1 -- Available -- --
I've previously only ever got circa 17.5 down and 2.3 up on Speedtest.net, which is quite disappointing.
Now I get 21.6 down but only 1.52 up when testing.
The HG612 stats are:
Mode VDSL2
Traffic type PTM
DSL synchronization status Up
DSL up time 0
Line Status
Help
Downstream Upstream
Attainable rate (kbit/s) 23056 1750
SNR margin (dB) 27.2 0
Line attenuation (dB) 11.2 3.3
Output power (dBmV) 11.2 3.3
Statistics
Help
Path 0
Path 1
Downstream Upstream Downstream Upstream
Line rate (kbit/s) 23213 1750 0 0
CRC errors 254 52 18 0
FEC errors 0 0 0 0
HEC errors 90 4294967290 32 0
Any suggestions?
Thanks!!
-
Some of stats from the web interface are incorrect. You need to telnet into the modem to get the correct and more detailed stats.
-
Welcome Visiting Huawei Home Gateway
Copyright by Huawei Technologies Co., Ltd.
Login:admin
Password:
ATP>sh
BusyBox v1.9.1 (2014-01-21 16:44:38 CST) built-in shell (ash)
Enter 'help' for a list of built-in commands.
# xdslcmd info --pbParams
xdslcmd: ADSL driver and PHY status
Status: Showtime
Retrain Reason: 0
Last initialization procedure status: 0
Max: Upstream rate = 1750 Kbps, Downstream rate = 22988 Kbps
Bearer: 0, Upstream rate = 1750 Kbps, Downstream rate = 23213 Kbps
Bearer: 1, Upstream rate = 0 Kbps, Downstream rate = 0 Kbps
Discovery Phase (Initial) Band Plan
US: (7,32) (871,1205) (1972,2782)
DS: (33,859) (1216,1961) (2793,3959)
Medley Phase (Final) Band Plan
US: (7,32) (871,946)
DS: (33,859)
VDSL Port Details Upstream Downstream
Attainable Net Data Rate: 1750 kbps 22988 kbps
Actual Aggregate Tx Power: 3.3 dBm 11.2 dBm
====================================================================================
VDSL Band Status U0 U1 U2 U3 U4 D1 D2 D3
Line Attenuation(dB): 12.4 63.0 N/A N/A N/A 24.4 80.4 N/A
Signal Attenuation(dB): 12.4 60.2 N/A N/A N/A 27.9 N/A N/A
SNR Margin(dB): 5.8 5.7 N/A N/A N/A 6.3 N/A N/A
TX Power(dBm): 0.4 0.2 N/A N/A N/A 11.2 N/A N/A
#
-
Were those the right stats?
-
It is rather difficult to formulate a diagnosis from snapshots at a moment in time. :)
The two sets of snapshot statistics that would be most appropriate are the output from --
xdslcmd info --stats
xdslcmd info --pbParams
-
Why not upload your stats to mydslstats for more indepth monitoring..
http://www.kitz.co.uk/routers/hg612stats.htm
-
Just signed up, same username - hopefully you can see something.
I've enabled 2 hour logging to start with
-
The software will update every minute if its running etc. You need a few days of continuous data so you can see if a pattern is forming or not. Best to review in a few days time.
Cant see you on the list of users yet though...
-
Yes I can't see me there either but can log in and have put the details into the Stats GUI
-
What program are you using to upload the data?
-
HG612 GUI
-
Just remembered to turn on Telnet on my server... Will hopefully work now after a reboot!
-
Seems to work now, hopefully will appear on the website
-
Looks like one upload at 23:18
-
Your line looks extremely poor. Is your modem connected to the master socket?
-
It is, but the wiring is a bit odd.
The main BT wire runs into the garage loft (where a master socket is tapped into the line) and then it carries on to a very old BT junction box on the outside of the house but it's terminated there.
I'd like to get BT out to check but they always threaten charges etc.
It was like this when I moved in, I just replaced the faceplate with the latest version.
-
So there's a bridge tap. That's bad.
-
Will BT sort that for free, or will they say it's my issue do you know?
-
This is what I found when I moved in!
(https://forum.kitz.co.uk/proxy.php?request=http%3A%2F%2Fthumbsnap.com%2Fs%2F8EvByVEe.jpg&hash=9e020f81e441f427f5b47e778f4ac9ac96eb88e4)
I've since changed the faceplate and cable tied it together so it's not all exposed!
-
If the fault is found outside of the premises (in Openreach's network) then there will be no charge.
-
But if the faults are due to the obviously dodgy looking 'master' socket above?
I'm within my BTW speed parameters, there's no quiet line issues from their tests, so they think I have no issues.
-
If you've made amendments to the master socket then yes they are within their "apparent" right to charge you, stupid as it is...
Unfortunately, as your line speed is within the estimate then there's not really much you can do to increase your speed accept trying to do it yourself...
-
They're on mydslwebstats
-
I noticed my line did a re-sync an hour ago, attenuation seems the same but SNRM dB seems much lower and sync speeds have jumped (particularly upload).
DLM resetting??
-
Looking at the back of that NTE are there two wires terminated to each screw A and B?
It looks like two drop 11 cables are terminated.
Also how many extensions are connected to the IDC blocks on the front of the socket?
-
Now into the master socket main plug on the face there's an extension lead that runs to a splitter near my DECT base station and alarm panel.
The modem is plugged into the DSL socket on the faceplate via a 5m RJ11 cable: https://www.amazon.co.uk/MCL-RJ11-Cable-ADSL-Lines/dp/B001DA34XG?ie=UTF8&ref_=pe_385721_37986871_TE_item
Cable has to be that long as the socket is in the garage loft.
-
Do you still have the bridge tap on the back of the master socket though, where the BT line coninues after the master socket to some unknown destination?
-
Yes although I've booked a local telecoms engineer to come and fix it - as I know where the cable goes to, and it's not used anymore there (my porch).
-
The line looks really messed up now, the SNR Margin is too low.
-
You could get the engineer to move the NTE on some new internal cable and relocate it next to your modem and phone :)
-
Is it acceptable to remove the bridge tap and re-cover the existing wire, or would it need to be jointed to new cable?
-
At the point where the current NTE5/A is situated I would remove the excess length external drop-cable, thus removing the bridging tap. Then I would connect CW1308 specification cable onto the end of the drop-cable via a BT80 and run the new length of CW1308 specification cable to wherever you would like the new NTE5/A to be located.
-
Is there a good or bad way to turn off the modem when the Engineer arrives so as not to cock up the DLM?
I read stuf before about unplugging the cable vs power off etc.
-
Power off the modem, then remove the patch cable connecting it to the filter. :)
-
Make sure the modem is off if you want to make sure nothing untoward happens.
-
Master socket relocated - things seem to have improved. I seem to be maxing out my profile on the down, how do I force a profile reset?
-
Try a reboot of the router but not the modem
-
Rebooted the HH5, doesn't seem to have changed anything.
I'm reasonably happy with my DS speed, was just expecting a bit more of an improvement in the US.
Do the rest of the stats look ok now? I've looked through the graphs but not really sure what I'm looking for.
-
Line looks quite unstable, lots of undulations/spikes in the SNR Margin especially on Downstream but I'll monitor it and see what happens.
-
What's been done? Has the bridge tap been removed?
Do you have a new master socket?
-
Yes the drop wire was cut and terminated into a BT80 - then internal cable run to a VDSL master socket next to the modem.
-
Is there anything else connected to the phone line?
-
Cordless phone and alarm.
Both connected to the master socket - although I didn't have a two way splitter so the alarm runs into the DSL side socket of a ADSL filter and the phone in the phone side.
-
I think you should try disconnecting those and see if the stability improves on mydslstats
-
Those have been unplugged for the last 20 minutes but seems to make no difference.
There's been a massive increase in FEC errors since the work was done, but the internet seems no different to use than before (just slightly faster)
-
Just wondering if old DLM settings are still applied, and the line has been improved - would that cause issues until the DLM re-adjusts itself?
-
Not really, this looks like noise interference on the SNR margin.
Can you post a picture of the master socket?
-
(https://forum.kitz.co.uk/proxy.php?request=http%3A%2F%2Fthumbsnap.com%2Fs%2FtrcnLeL4.jpg&hash=4d5a37a6017dc3db97856afd8ad7030188932697)
-
Hi
Is your line cable running parallel with your power cable and are both cables using same entry hole
Also, your alarm line feed from your filter is unfiltered, which will not help as you state and pic shows, is using dsl socket
If your line feed is running parallel to power cable, you maybe picking up noise from magnetic field of power cable
Many thanks
John
-
Thanks. Looks a bit close to the mains though.
If you look at the SNRM/Band DS graph you can see the fluctuations on the D2 band
-
An increase in FEC errors with no physical changes to the line made by DLM isn't exactly a good thing, let's just hope G.INP can stop bursts of CRC's getting through.
Just seen your photo, you shouldn't have a microfilter connected to the phone line part of the SSFP.
-
The phone cable runs separate to the power cable - the phone line goes straight through the wall into the garage and runs up the wall (not near anything else).
I've noticed when logging into the HG612 that the HEC errors are mostly only on the up:
Path 0
Path 1
Downstream Upstream Downstream Upstream
Line rate (kbit/s) 28306 3378 0 0
CRC errors 254 108 18 0
FEC errors 0 0 0 0
HEC errors 90 4294967290 32 0
-
That's probably because G.INP is fixing the downstream errors
-
The stats from the HG612 interface seem to conflict with whats being sent to Webstats via the GUI.
-
I am using AV500 Powerline adaptors - saw a thread elsewhere on here where someone said their AV500 powerline adaptors were causing errors (but AV200 didn't).
-
Why have you got a microfilter in the phone part of the SSFP? That's bound to cause issues which no one has spotted.
-
It's not causing the problems which happen when that's disconnected too. In fact I can't see what problems it could cause anyway.
-
Didn't you know that you're not supposed to connect a microfilter into the SSFP? He's got a dual phone line going into one socket if you look closely, can't see how that would work well.
-
Works fine for me
-
Disconnected the alarm and errors have gone - the alarm cable does run next to the main incoming 240v feed to the fuse board so probably that!
-
Also, your alarm line feed from your filter is unfiltered, which will not help as you state and pic shows, is using dsl socket
I thought the phone and alarm are plugged into that plug-in DSL microfilter. The phone will be being double filtered, once by the Mk3 faceplate and again by the plug-in filter. The alarm will be filtered only by the Mk3 faceplate.
I've heard Plusnet staff say that things being double filtered "will cause issues", not specifying what issues or why. In the BTWholesale FTTC handbook it used to say it didn't matter. The "issue" might be that a visiting Openreach technician will object to it and make the visit chargeable.
-
Is the phone cable actually UTP (unshielded twisted pair) from the BT80A to the master socket?
-
The line still looks unstable and can't remember seeing any actual errors on your line.
-
Not sure - it was an existing cable that he used but he reckoned was ok - it looks more like flat-ish telephone extension cable than Cat 5 cable if that's what you mean
-
I've disconnected the Powerline adaptors and Alarm from the main telephone socket (just the DECT phone into the socket) - the errors seem to have dropped right down
-
The stats from the HG612 interface seem to conflict with whats being sent to Webstats via the GUI.
that's because the gui is wrong. fec/crc shows incorrect, only hec is right. I never login to the web interface. use telnet to get your stats or 1 of the stats programs.Why have you got a microfilter in the phone part of the SSFP? That's bound to cause issues which no one has spotted.
it's unnecessary, but can't see what problems it would cause.
-
I'm not sure what problems either, but it's been told that you shouldn't connect a microfilter into that socket...
-
Not sure - it was an existing cable that he used but he reckoned was ok - it looks more like flat-ish telephone extension cable than Cat 5 cable if that's what you mean
That's what I meant and it should look like CAT5. In fact CAT5 would be better in my opinion as long as it is solid copper.
-
So with everything disconnected it seems like my line is solid.
Only problem is now to try and get devices to connect to the LAN if I can't use Powerline.....
I've got a Unifi UAP-LR currently serving as the WAP (HH5 has WLAN turned off) - which seems to make performance much better, but I need to run more throughout the house without cables everywhere (else Mrs will go mental)!
-
I wish we could have seen (1) the cable that has been reused to link the BT80 to the new NTE5/A (2) the proposed new location for the NTE5/A before the work was done.
Having the NTE5/A that close to a mains power socket is something that I would not recommend and -- sudden thought time -- I wonder if there is some form of feed-back loop, which is coupling the signal output from your "home plugs", via the mains wiring to that 13A socket, onto the telephone line and, thus, into the analogue front-end of the modem? :-\
Looking at you circuit, via MDWS, I see that it is not particularly "happy" and will be interested to see if there are any DLM induced changes in the forthcoming hours. The first 24 hours, post telephony wiring modifications, will be interesting viewing.
-
It's a difficult one - you pay professional tradesmen for their experience, this guy said it would all be fine - but then he also reckoned that the HG612 and HH5 wouldn't work together so it made me a bit wary!
-
To be fair, only Openreach are allowed to do that work
-
Saw you had a spike of 200 Downstream CRC's earlier on this morning. ???
-
Reviewing the various plots over the last 24 hour period, we can clearly see the --
- pre-modification phase
- disconnected phase
- reconnected phase with the "home plugs" operating
- reconnected phase without the "home plugs" operating
I know that N*Star will recognise phases (3) & (4), as he has had personal experience of the same. ;)
Overall, the circuit is recovering and a quick look at the Hlog plot tells me that there does not appear to be any significant defect in the metallic pathway. The limiting factor appears to be the loop-loss but at least the US0, DS1, US1 & DS2 bands are present in the medley ("showtime") phase.
-
So is the current up and down sync the best I can expect?
Seems to correlate with BT DSLChecker but the bloke yesterday said that's rubbing generally
-
Well, the synchronisation speeds are the best you can get at this moment in time . . . Whether or not there may be potential for further improvements will require close scrutiny of a few days of statistics by one of those more versed in the intricacies of G.993.2 circuits, such as Bald_Eagle1.
-
You've already done more than you're supposed to, as everything before the NTE5 belongs to OpenReach. I wouldn't pay too much attention to what the "engineer" said. Anyone who puts a master socket so close to a power socket that they are practically touching shouldn't have too much attention paid to them.
The answer to your question is probably yes. the line is much more stable now, without any faults showing and within dslchecker range, chances of OR being able to do anything that would make any worthwhile difference are practically nil.
-
Your downstream looks to be in the right area say about 2 Mbps off but your upstream is a bit low, I'd expect about 5-6 Mbps. I think that is down to the poor quality cable between the BT80A and the master socket which is probably preventing it syncing at those figures.
But given your estimate is VDSL Range A (Clean) 29.7 22.4 4.9 3.2 -- Available -- --
VDSL Range B (Impacted) 22.3 10.4 3.8 1.1 -- Available -- --
you're still in Range A :)
-
Ok, so would you say my upstream is low, too? I get 42.7/7.9.
-
Are your stats on MDWS?
-
you're clean high estimates are 34.2 / 7.4
wish I got above my estimates :no:
-
Yeah, it is pretty cool. I just throught the upstream seemed slightly low.
-
Just had a re-sync, the sync has shot up past the attainable - but I've noticed a huge increase in FECs, are these actually a problem or normal with interleaved line?
Last night I moved my DECT phone base station away from the power line, and in the past couple of evenings have been experimenting with Powerline adaptors on/off (as the Mrs needs them on to watch iPlayer).
-
Woah, something is drastically wrong there, your Downstream SNR Margin is now at 0.4 dB which is extremely low, I think the Downstream FEC figure has got stuck because of this, I'd do a manual resync immediately.
FEC's aren't an issue especially with Downstream G.INP enabled, but your router is working incredibly hard to keep the line stable and surprised to say the line hasn't lost sync yet.
-
What do the three traffic lights mean on MyDSLWebStats?
I'm not at home so can't do anything, but interesting to see how it's jumped so high (maybe too high?).
-
Currently, your line is perfoming very well to say how much strain it's under. You could leave it, but I wouldn't.
-
. . . and in the past couple of evenings have been experimenting with Powerline adaptors on/off (as the Mrs needs them on to watch iPlayer).
Usage of the "powerline adaptors" ("home plugs" / "whatever the current name") can clearly be seen by the sudden upwards jump in the FECs when examining the circuit statistics via MDWS . . . :-X
-
Will try again tonight, but seeing as I now have eleventy billion FEC errors on my line, do a few more make any actual difference if not causing CEC etc?
-
It'll be making a slight difference, especially to the HG612!
-
Just checked your line, CRC's appearing now and lots of them, the Downstream SNR Margin dropped to 0.1 dB, not sure how you've still got a working service.
-
. . . the Downstream SNR Margin dropped to 0.1 dB, not sure how you've still got a working service.
In the past, I have seen a HG612 maintain synchronisation with the DSLAM when the SNRM has gone negative. I watched it go down to -1.9 dB and it was only when it reached -2.0 dB that it finally "let go".
-
Just had re-sync
-
Speeds lower, errors lower - just need DLM to find the sweet spot in the middle now.
US seems slow.
-
I think there's still a fault with your line.
I wish we could change SNR Margin on FTTC, I'd have it at 0 dB.
-
Puzzled were the 10Mbps came from on the downstream it didn't come from the attainable as it stayed around 28.9Mbps even when a disturber has turn off their modem the SNRm would increase and so would the attainable rate. and why was the modem set at such a low SNRm in the first place ???
-
Think the modem synced when there was a noise spike.
-
Sounds plausible the DS & US SNRM went low just as the modem finsihed it's resync and the result is higher sync rates with very low SNRM.
-
Now, that gives me an idea, I need to time my next manual resync... :lol:
-
I reset the modem and seem to have restored my US to where it should be at least.
I'm happy enough (for now) with 28/3 if it's stable - once I've sorted my LAN then hopefully will resolve the AV500 FEC issues.
Anyone else had experience with the Unifi products? I'm impressed with the WAP I've got, so thinking about adding a second upstairs and replacing Homehub 5 with the Unifi Security Gateway.
-
I've been away for quite a while so I was just reading the early post on this thread.
You said you "just plugged it in to the HH5".
Is the HG612 plugged in to the phone socket?
Then the HH5 is behind the HG612?
-
Yes - everything seems to have stablised a bit now, I'm still using the AV500 powerline adaptors but they don't seem to be causing massive problems. I've also tried separating power/data cabling as much as reasonably possible.
The order of things is:
Master socket - HG612 - WAN socket on HH5 (with Wifi off) - Unifi WAPs
My original plan was to replace the HH5 with another router, but not found out yet if the Ubiquiti range is fully compatible with BT Multicast yet.
-
Yea, thats the problem with 3rd party routers, the BT Multicast function.
I'm glad to hear things have stablized.
I'm not a fan of powerlines but I can see how useful they are. As long as your getting a good connection then if you want to make any changes then at least its in your 'realm' to do them.
-
I'm trying to get rid but in the short term there's no other option without resorting to wireless bridges but that may then just reduce overall Wifi performance for little gain.
Unfortunately in an existing older house - running data cabling everywhere isn't that easy!
-
My original plan was to replace the HH5 with another router, but not found out yet if the Ubiquiti range is fully compatible with BT Multicast yet.
If you take a look at this post (http://forum.kitz.co.uk/index.php/topic,17428.msg326062.html#msg326062), it might help you to decide if an Ubiquiti EdgeRouter Lite would be appropriate for your needs.
-
If seen that one but there was no mention of whether he was using it for Youview/BTTV so assumed not.
The ERL and USG seem to be the same hardware but with different UI - advantage of USG is that it's managed from the same portal as WAPs whereas the ERL is more from command line and is less 'home' user friendly.
-
If seen that one but there was no mention of whether he was using it for Youview/BTTV so assumed not.
Understood.
The point I failed to make is that the device can be configured exactly as required. Assuming one knows the requirements for such a service (as above), then the ERL can be configured for its usage.
-
I found the info about the test streams etc so was trying to get someone on the Ubiquiti forums to try it out, but not heard anything back yet.
Unfortunately my networking skills are limited, I can do basic port forwarding etc, but when it comes to CLI work I end up using copy/paste from Google hits!
-
Looking at the stats, it's mostly settled down now - but I'm getting error spikes around 2am each night - and I have no idea why!
-
Do you mean the OHF Header Errors/min [OHF] ?
-
FEC and CRC I was looking at
-
Looks like they only started on the 15th
-
FEC has been going on for a while (overnight peaks) but you're right that CRC has been recent, which is even more odd as I was away this weekend!
-
Maybe an electrical device starting up at 2am - air conditioning perhaps?
If you still have that flat phone wire instead of UTP, this will pick up electrical interference that UTP would reject.
-
I double checked and it's not flat actually - think it's UTP.
The immersion heater comes on overnight all the time and dishwasher we set to night (when home) so must be that!
-
It looks like one event at 2am each night so I would guess it's a thing that turns on (or off) at that time and stays on (or off). But it doesn't appear to be a problem, just one of those things you see if you record stats 24*7 :)