Optimistic SNRM causing frequent disconnects?

[#1]

Quite some time ago, I had the upstream speed cap removed, and ever since had problems with "often enough to be annoying" disconnects. I'm fortunate to live near enough to the exchange to get very good line sync speeds, but this instability is starting to drive me nuts.
Am I right in thinking this is probably due to an overly optimistic SNRM, and if so, am I also right in thinking that this needs to be adjusted by PlusNet? 
TG582n is direct into the NTE5 via a microfilter, with one other extension wired from the NTE5 faceplate also with a microfilter, no bell wire.  I have an vDSL Interstitial faceplate on it's way too, in case I have a dodgy microfilter.

Uptime: 0 days, 3:33:53
DSL Type: ITU-T G.992.5
Bandwidth (Up/Down) [kbps/kbps]: 1.223 / 20.335
Data Transferred (Sent/Received) [GB/MB]: 1,67 / 866,69
Output Power (Up/Down) [dBm]: 12,4 / 0,0
Line Attenuation (Up/Down) [dB]: 6,6 / 16,0
SN Margin (Up/Down) [dB]: 6,2 / 3,1          <-- Is this the problem?
System Vendor ID (Local/Remote): TMMB / ----
Chipset Vendor ID (Local/Remote): BDCM / IFTN
Loss of Framing (Local/Remote): 262 / 0
Loss of Signal (Local/Remote): 30 / 0
Loss of Power (Local/Remote): 0 / 0
Loss of Link (Remote): -
Error Seconds (Local/Remote): 2.957 / 11
FEC Errors (Up/Down): 0 / 0
CRC Errors (Up/Down): 0 / 5
HEC Errors (Up/Down): 0 / 18

IP Profile for your line is - 17.94 Mbps
Current line speed: 17.8 Mb

TIA!
Phil

[#2]

What firmware version is the 582n running?

[#3]

Hi Phil,
Welcome to the forums.
Looking at all of the figures, I would say that something else is wrong.  A higher SNRM might improve stability however that would be at the sacrifice of yet more synch speed. Give the line's attenuation, you should see 21.6Mbps at 6dB - you are seeing less at 3dB.
Is the phone line noisy?  Dial 17070 option 2.
What do the stats look like if you remove the NTE5 face plate and connect to the test socket?  In this configuration is there a difference between the two filters?
The VDSL face plate sounds like a useful activity.
How long has the router been powered up?  The loss counters and error seconds are high, however I think they span a longer time than the current DSL session.  A power down to clear them would be useful.
After checking out the above, raise a fault report here - http://faults.plus.net - to have the line tested.
Kevin

[#4]

Firmware is 8.4.4.J,
Router uptime 23 days
17070 line test is quiet

I'll repeat some stats via the face plate later today...

[#5]

Might as well upgrade to the 10.2.2.B firmware (assuming you have a computer that can run the firmware upgrade software).

[#6]

Phil,
Ejs makes a good point - can you take a look at the router logs and see if there are any references to Kernel Warm Restarts?
I've seen that version of firmware throw inexplicable disconnects in the past.
You might be able to get PlusNET to do a remote firmware upgrade, but that will not be today.
Kevin

[#7]

Thanks all; connecting directly into the NTE5 test socket hasn't really helped, even a short while later;
Uptime: 0 days, 1:29:51
DSL Type: ITU-T G.992.5
Bandwidth (Up/Down) [kbps/kbps]: 1.191 / 20.767
Data Transferred (Sent/Received) [MB/GB]: 141,44 / 1,17
Output Power (Up/Down) [dBm]: 11,8 / 0,0
Line Attenuation (Up/Down) [dB]: 6,5 / 15,5
SN Margin (Up/Down) [dB]: 6,2 / 3,1
System Vendor ID (Local/Remote): TMMB / ----
Chipset Vendor ID (Local/Remote): BDCM / IFTN
Loss of Framing (Local/Remote): 0 / 0
Loss of Signal (Local/Remote): 0 / 0
Loss of Power (Local/Remote): 0 / 0
Loss of Link (Remote): -
Error Seconds (Local/Remote): 3 / 0
FEC Errors (Up/Down): 0 / 0
CRC Errors (Up/Down): 0 / 8
HEC Errors (Up/Down): 0 / 13
I'm Mac based (with Windows 8 VM), so a firmware upgrade won't be plain sailing.
I'll log a fault, and see where we go...
Thanks all for your help,
Phil

[#8]

Just to wrap this up, PlusNet support agreed that whilst there was no obvious line fault, the low SNRM was likely causing an overly optimistic, unstable, connection.
They ordered a reset (to 6db), and things have been much improved since, albeit at the slight loss of downstream speed (currently at 1.203 / 18.467 at 6,1 / 5,8 SNRM).
Phil