Transfer to 21C Network

[#1]

He did an excellent job and tbh, having had problems with BTOR engineers in the past, this is the first one I can praise

[#2]

I am going to pass on the praise back to our suppliers and that should filter through
Just grabbing an update as to when we're going to see some movement on the REIN Case.

[#3]

I've just popped an update on your ticket.

[#4]

Ta chris, I've answered
[Update]  Everything has been rock steady since engineer left, all graphs except for FEC have been horizontal then.....
17:45hrs  massive FEC spike, SNRM has dropped from 6.0 to 5.3dbs . Nothing showing on bitloading graph though
massive CRC spike which actually disconnected my DSL while I was typing this
Sync speed dropped
DSLAM/MSAN type: IFTN
DSL mode:      G.992.5 Annex A
Status:        DSL Link Up
0 days, 0:02:02

Downstream Upstream
Attenuation (dB):      27.5 14.3
Connection speed (kbps): 16079 888
SNR margin (dB):        6.1 15.4
Power (dBm):            0.0 12.1
Interleave depth:      128 0
RSCorr/RS (%):          2712.3077 0.0000
RSUnCorr/RS (%):       
ES/hour:               
INP:                   

[#5]

Just a bit of information and just to clarify things for you, the missing tones 110, 127, 188, 191, 243, 291 & 348 are standard missing tones with most Broadcom based Modem/Routers and 21CN IFTN MSANs and are not related to the REIN issue. Those missing tones are a standard "feature". One other "feature" is a tone with only 2 bits loaded, which can be a different one after a resync. looks like Tone 219 in reply #138.
In reply #133, the big depression and missing tones in the bit loading from circa tone 358 upwards (as this was before you new drop-wire) could well be primarily due to the defective cable.
If the Bit loading in reply #138 was after the Interference started, then another Bitloading for comparison from tomorrow daytime when the interference is not there (with hopefully no loss of sync since reply #138) will then indicate the tones that are being affected by the Interference. (ie.you need 2 daytime graphs with the same sync, one with Interference & one without).
The depression around tone 384 and similar on higher tones could well be related to the REIN. As already mentioned, a comparison to a daytime Bits/Tone without interference at the same sync will help clarify that. Some of the very slightly reduced bit loading below tone 384 eg. around tones 200 and 250 might be due to MW broadcast interference and is not uncommon.

[#6]

The first 3 graphs are from last night when the interference was noticeable, the 4th graph is from this morning while FEC and CRC graphs are both at zero.
weird one last night, because I wasn't seen much on the graphs I decided to test the radio to see if I could pick up any interference. Somehow I actually picked up some french radio station broadcasting around the 612khz mark.
Anyway, as I said seems have been quiet since the weekend, so I'll keep monitoring and just hope I get mega spikes again for the reins engineer to hunt down. Added DSL stats from the router to show the error rates for 13 hours uptime, seem quite high although the graphs don't show much.
DSL Connection
Link Information

Uptime: 0 days, 13:39:20
DSL Type: ITU-T G.992.5
Bandwidth (Up/Down) [kbps/kbps]: 888 / 15,989
Data Transferred (Sent/Received) [GB/GB]: 3.68 / 14.08
Output Power (Up/Down) [dBm]: 12.0 / 0.0
Line Attenuation (Up/Down) [dB]: 14.3 / 27.0
SN Margin (Up/Down) [dB]: 15.6 / 6.9
System Vendor ID (Local/Remote): TMMB / ----
Chipset Vendor ID (Local/Remote): BDCM / IFTN
Loss of Framing (Local/Remote): 96 / 0
Loss of Signal (Local/Remote): 11 / 0
Loss of Power (Local/Remote): 0 / 0
Loss of Link (Remote): -
Error Seconds (Local/Remote): 1,087 / 0
FEC Errors (Up/Down): 0 / 10,252,459
CRC Errors (Up/Down): 0 / 17,407
HEC Errors (Up/Down): 0 / 215,595

[#7]

That 2156 Bitloading from last night is a good one, clearly showing the REIN. Was the CRC and the FEC at a peak at that moment as well?
At say 2130, when the FEC and CRC is very low to zero, is the Bitloading pattern still the same or do those rough U shapes disappear?
As far as the Modem/Router DSL stats go, obviously the numbers show the error rates during the uptime to be somewhat high. What is probably more useful is the stats say at a 15 minute interval when no interference is present, and likewise when it is present (as you did before). One can then look at the change in the figures over the interval.
OK a couple of thoughts for you to eliminate (or add suspicion to) some possibilities -
Do you have any MW Pirate Radio stations nearby?
Do you or any of your close neighbours have old type Analogue Cordless phones?
Do you have any Radio Amateurs nearby?

[#8]

DSLstats is set to take snapshots of the graphs every 4 hours atm, I can reduce it to every 1 hour but any less and I'd have to take the snapshots manually. Let me know if the hourly snapshots would be better. Unfortunately there doesn't seem to be an option to save the stats at the same time, unless I'm missing the setting
As to the Pirate station or Radio Amateurs I've no idea. There is a radio mast just around the corner for taxi cabs, and about 100 meters away is the local Police station with all their masts/mobile phone masts etc but since they would be in use outside the hours the interference appears I ruled them out as a source.
As to the phones I've no idea. I have a corded phone with answer machine connected atm but no idea what the neighbours use.
I've posted the latest stats from dslstats. I've been trying to do some research on interleaving which seems straight forward but most of what I've seen suggests a depth of between 1 and 64. But if you look at my stats the interleave depth is at 192, not sure if thats something to be concerned about or not.
DSLAM/MSAN type: IFTN
DSL mode:      G.992.5 Annex A
Status:        DSL Link Up
0 days, 21:57:30

Downstream Upstream
Attenuation (dB):      27.0 14.3
Connection speed (kbps): 15989 888
SNR margin (dB):        6.8 15.5
Power (dBm):            0.0 12.0
Interleave depth:      192 0
RSCorr/RS (%):          17328392.3077 0.0005
RSUnCorr/RS (%):       
ES/hour:               
INP:                   

[#9]

I wouldn't have thought setting a shorter snapshot interval would be worth while. But when you know the Interference is on, doing a manual snapshot of the Bitloading when the FECs are peaking and one when they are down near zero could be useful, if you happen to be able to monitor it at the time.
A Radio Amateur would tend to have quite an array of fairly large antennae, so you'd probably see them if there was one in the vicinity. Unlikely to be Taxi or Police Radio for the reasons you've mentioned and the much higher frequencies used for such transmissions.
A MW Pirate Radio station might be fairly obvious once you start looking. It would obviously be on air when you were suffering the interference because it would be fairly local (and so it would be quite a strong signal), they are usually run by some minority group, broadcasting just one particular genre of music, the quality is normally fairly poor and not well presented. Depending on the minority group, it may not be English language being used.

[#10]

There you go Anotherone, all 3 graphs taken at the same time, first spike of the evening

[#11]

Yeh, a nice repeatable pattern on the Bitloading.
Now have a tune around on your Radio, see if you can find something anywhere around about 1070kHz to 1100kHz (280/290m. ish) also somewhere down around 1600kHz (188m.)

[#12]

Took a turn with the radio and definitely getting something which is stronger out back.
Latest graphs, CRC has dropped to zero so not included

[#13]

That's interesting, how strange. Notice how all the tones above about 370 are depressed with less bitloading, but the dips around 240 and 310 have disappeared. One has obviously to assume this is all related to the same source, but it's doing something slightly different at present.

[#14]

Yeah, I'm suspecting that when it is switched on, like all appliances there's a sudden surge, then it settles down to hum and cause steady interference
Then again I'm on a steep learning curve with this stuff (steep more like vertical) so I could be talking bs
Just had another spike, CRC aswell and the bitloading changed again

[#15]

23:25 hrs Both FEC and CRC dropped to zero, bitloading graph below