Kitz Forum
Broadband Related => FTTC and FTTP Issues => Topic started by: matt12321 on November 10, 2017, 09:14:52 AM
-
Hi,
Stats are on MyDSLWebStats under my username, but I see nothing obvious there - apart from DS SNR seems to decrease gradually and then triggers a DLM reset.
Line has been rock solid for ages, now all of a sudden over the last few days I've had annoying dropouts. Nothing has changed that I'm aware of that could cause this.
Anyone got any ideas?
Thanks!
-
Perhaps dlm is decreasing snr on the ds as you are on huawei cab so can have lower snr profiles.Then you get too many errors so dlm resyncs increasing the ds snr so the Cycle starts again..
-
US SRN is holding steady at 6.4/6.5dB, but when I look back at say the last 5 days DLM has been triggering at random times.
There's no obvious link I can see between DLM triggering any any type of error spike - it also seems to do it when DS SRN has reached different threasholds, so it's not a consistent 3dB cut off for example.
-
It doesn't look like DLM is causing these resyncs to me.
The strange pattern looks to be the "decay curve" on the SNRM graphs. From this, I'd suspect a fault on the line, or in the modem driving the line.
Looking back a longer period, I'd say there was a noticeable change around 10th October. Using the 90 day graph, you can see that was the end of a long period of uptime (which is otherwise not that stable), and a shift in the SNRM patterns, and a shift in the bit-swapping graph.
If you look at the graphs that show behaviour per tone (Quiet Line, SNR/tone and bits/tone), and specifically check the historic ones for the 10th and 11th October, you can see marked changes between tones 300 and 835 between those two dates. Notionally, the "after" graph looks like it gets less interference, but the "before" graph looks an awful lot smoother.
I can't say what those changes mean, though.
-
I tend to look at the immediate preceding 24-hour time-frame for a "snapshot" overview of a circuit's behaviour.
Of the five plots, upon which I concentrate, the one that I find abnormal is the QLN. The shape of the DS1 band appears to be rather different from either of the two typical forms that we regularly see and the plethora of spikes in the DS2 & DS3 bands is most undesirable.
Although the Hlog plot is perfectly acceptable, my feeling is that the circuit, as a whole -- that is from the transceiver of the xTU-C to transceiver of the xTU-R -- has some defect.
-
Could you attach Hlog and QLN?
-
Could you attach Hlog and QLN?
I've sent you an e-mail message with copies of the Hlog, QLN & SNRM plots for Matt12321's circuit attached.
-
Thank you. :)
-
Could you attach Hlog and QLN?
Are you not able to access https://www.mydslwebstats.co.uk ?
Stats are on MyDSLWebStats under my username
Hlog/QLN are available to view on there, along with nearly every other stat available.
-
Question is, what can I ask BT to do about it?
I hate their standard response of "we can send an Engineer but it will cost you Eleven Billionty Pounds unless we find the fault is ours (which it probably isn't)".