In reply to a post by timh:

by itself at about 5am this morning

This could indicate a problem.

With the faceplate removed, is the alarm still connected?

With faceplate removed, all the extensions are now disconnected - I have put our phone directly into the 'test' socket.

I assume I'm still in the training period, and thought some of these disconnects would relate to that?

Is there any way to get error stats from the home hub 5?

Looking at the event logs, it appears the disconnect/reconnect was being done as part of a f/w update attempt?

05:49:30, 05 Dec. (19421.650000) CWMP: session completed successfully
05:49:30, 05 Dec. (19421.430000) CWMP: HTTP authentication success from https://pbthdm.bt.mo
05:49:25, 05 Dec. (19416.800000) CWMP: Server URL: https://pbthdm.bt.mo; Connecting as user: ACS username
05:49:25, 05 Dec. (19416.790000) CWMP: Session start now. Event code(s): '4 VALUE CHANGE'
05:49:23, 05 Dec. (19414.350000) WAN operating mode is VDSL
05:49:23, 05 Dec. (19414.350000) Last WAN operating mode was VDSL
05:49:22, 05 Dec. (19413.460000) PPP IPCP Receive Configuration ACK
05:49:22, 05 Dec. (19413.440000) PPP IPCP Send Configuration Request
05:49:22, 05 Dec. (19413.440000) PPP IPCP Receive Configuration NAK
05:49:22, 05 Dec. (19413.440000) PPP IPCP Send Configuration ACK
05:49:22, 05 Dec. (19413.440000) PPP IPCP Receive Configuration Request
05:49:21, 05 Dec. (19412.930000) PPP IPCP Send Configuration Request
05:49:21, 05 Dec. (19412.930000) CHAP authentication successful
05:49:21, 05 Dec. (19412.350000) CHAP Receive Challenge
05:49:21, 05 Dec. (19412.350000) Starting CHAP authentication with peer
05:49:21, 05 Dec. (19412.340000) PPP LCP Receive Configuration ACK
05:49:21, 05 Dec. (19412.320000) PPP LCP Send Configuration ACK
05:49:21, 05 Dec. (19412.320000) PPP LCP Send Configuration Request
05:49:21, 05 Dec. (19412.320000) PPP LCP Receive Configuration Request
05:49:20, 05 Dec. (19411.520000) CHAP Receive Challenge
05:49:20, 05 Dec. (19411.520000) Starting CHAP authentication with peer
05:49:20, 05 Dec. (19411.520000) PPP LCP Receive Configuration ACK
05:49:20, 05 Dec. (19411.500000) PPP LCP Send Configuration Request
05:49:20, 05 Dec. (19411.500000) PPP LCP Receive Configuration Reject
05:49:20, 05 Dec. (19411.500000) PPP LCP Send Configuration ACK
05:49:20, 05 Dec. (19411.500000) PPP LCP Receive Configuration Request
05:49:20, 05 Dec. (19411.490000) PPP LCP Send Configuration Request
05:48:29, 05 Dec. (19360.480000) CWMP: session closed due to error: Could not resolve host
05:48:29, 05 Dec. (19360.460000) CWMP: Server URL: https://pbthdm.bt.mo; Connecting as user: ACS username

Same disconnect this morning at 9:00am

Googling suggests this is a common problem - but I don't understand why the BT agent needs to disconnect/reconnect the PPOE when the f/w isn't actually updating (as I seem to have the latest f/w already).

09:00:44, 06 Dec. (59023.040000) CWMP: session completed successfully
09:00:44, 06 Dec. (59022.820000) CWMP: HTTP authentication success from https://pbthdm.bt.mo
09:00:39, 06 Dec. (59018.200000) CWMP: Server URL: https://pbthdm.bt.mo; Connecting as user: ACS username
09:00:39, 06 Dec. (59018.200000) CWMP: Session start now. Event code(s): '4 VALUE CHANGE'
09:00:37, 06 Dec. (59015.610000) WAN operating mode is VDSL
09:00:37, 06 Dec. (59015.610000) Last WAN operating mode was VDSL
09:00:36, 06 Dec. (59014.730000) PPP IPCP Receive Configuration ACK
09:00:36, 06 Dec. (59014.710000) PPP IPCP Send Configuration Request
09:00:36, 06 Dec. (59014.700000) PPP IPCP Receive Configuration NAK
09:00:36, 06 Dec. (59014.700000) PPP IPCP Send Configuration ACK
09:00:36, 06 Dec. (59014.700000) PPP IPCP Receive Configuration Request
09:00:35, 06 Dec. (59014.390000) PPP IPCP Send Configuration Request
09:00:35, 06 Dec. (59014.390000) CHAP authentication successful
09:00:34, 06 Dec. (59012.780000) CHAP Receive Challenge
09:00:33, 06 Dec. (59012.090000) Starting CHAP authentication with peer
09:00:33, 06 Dec. (59012.090000) PPP LCP Receive Configuration ACK
09:00:33, 06 Dec. (59012.070000) PPP LCP Send Configuration ACK
09:00:33, 06 Dec. (59012.060000) PPP LCP Send Configuration Request
09:00:33, 06 Dec. (59012.060000) PPP LCP Receive Configuration Request
09:00:31, 06 Dec. (59009.580000) CHAP Receive Challenge
09:00:31, 06 Dec. (59009.580000) Starting CHAP authentication with peer
09:00:31, 06 Dec. (59009.580000) PPP LCP Receive Configuration ACK
09:00:31, 06 Dec. (59009.560000) PPP LCP Send Configuration Request
09:00:31, 06 Dec. (59009.560000) PPP LCP Receive Configuration Reject
09:00:31, 06 Dec. (59009.560000) PPP LCP Send Configuration ACK
09:00:31, 06 Dec. (59009.560000) PPP LCP Receive Configuration Request
09:00:31, 06 Dec. (59009.550000) PPP LCP Send Configuration Request
09:00:29, 06 Dec. (59007.970000) CWMP: session closed due to error: Could not resolve host
09:00:29, 06 Dec. (59007.950000) CWMP: Server URL: https://pbthdm.bt.mo; Connecting as user: ACS username
09:00:29, 06 Dec. (59007.950000) CWMP: Session start now. Event code(s): '4 VALUE CHANGE'
09:00:29, 06 Dec. (59007.750000) CWMP: Initializing transaction for event code 4 VALUE CHANGE
09:00:22, 06 Dec. (59001.140000) PPP LCP Send Termination Request [User request]

Could possibly be the modem quality.

The BT tester (meter) will have a well made, tested and calibrated front end (modem) which will include filter (high pass) and and terminations. Your modem will be produced on a production line and whilst the design works, keeping costs to a minimum will mean that tolerances can be wider, testing less (if any) and calibration unlikely.

Can you borrow the modem from your neighbour and see what it gives on your circuit and try your modem on his? See if the results are line dependent or modem dependent.

I have two "identical" modems here - on one the attainable upstream rate is around 26.5 Mbps and the other 22Mbps - they also have different downstream rates too.

If he already had an ADSL faceplate fitted WHY change anything at all, mine was never changed, all my engineer did was check that it was wired up correctly, and I have had a perfect connection ever since

In theory yes a well working ADSL faceplate should be fine for VDSL2

Depends a lot on the local training, and what bosses have told the engineers.

I've still been getting these disconnects, I'm wondering if the f/w check is just part of the normal reconnect procedure - rather than actually causing the reconnect itself? Would the exchange tell my router to reconnect or just force the reconnect from the exchange end?

I've also put back my original ADSL faceplate and the upload speed is now about 3.5Mb, which is higher than it has been. Not sure if this will improve over time, but will wait and see what happens.

Fingers crossed, but since changing back to my original filtered faceplate, I've not had any disconnects for about 26 hours.

This adds to my suspicion that there may have been a fault with the new filter the engineer installed.

Although things had improved slightly with the original filtered faceplate, I still couldn't see the full connection stats from the home hub 5, so I purchased an HG612 modem from ebay for about £3.50! I have installed the unlocked f/w and can now collect full stats to see what my connection is up to. I have included some graphs here if people could comment? Interestingly, with this HG612 modem it is now showing a much higher max upload speed (more in line with expectations) although I also see the upload SNR has risen to 9db. Will my profile now increase over time to get closer to the max rate ?

Max: Upstream rate = 5924 Kbps, Downstream rate = 29516 Kbps
Path: 0, Upstream rate = 3693 Kbps, Downstream rate = 25209 Kbps

Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 6.2 9.6
Attn(dB): 0.0 0.0
Pwr(dBm): 12.2 5.2


Downstream Upstream
Attenuation (dB):
Connection speed (kbps): 25209 3693
SNR margin (dB): 5.7 9.7
Power (dBm): 11.9 5.2
Interleave depth: 399 1
INP: 3.00 0

RSCorr/RS (%): 0.0917 0.0000
RSUnCorr/RS (%): 0.0017 0.0000
ES/hour: 11.5 0


https://lh4.googleusercontent.com/-puP_w1peEmg/UqiLE...

Edited by deleted (Wed 11-Dec-13 16:06:28)

The band plan found under Telnet data ¦ pbParams may shed more light on the connection. Looks like DLM will raise your upstream given time.