'Normal' evening SNRM drop/flat twisted pair RJ11 cable.

[#1]

OK, a new day, an old Netgear router reconnected since 1515 yesterday afternoon:

Router log:
Thu, 2014-11-27 14:54:20 - Receive NTP Reply from time-g.netgear.com
Thu, 2014-11-27 14:56:04 - Administrator login successful - IP:192.168.0.2
Thu, 2014-11-27 15:50:20 - Router start up
Thu, 2014-11-27 15:15:35 - Initialize LCP.
Thu, 2014-11-27 15:15:35 - LCP is allowed to come up.
Thu, 2014-11-27 15:15:36 - CHAP authentication success
Thu, 2014-11-27 16:42:35 - Administrator login successful - IP:192.168.0.2
Thu, 2014-11-27 19:58:12 - Administrator login successful - IP:192.168.0.2
Thu, 2014-11-27 21:27:42 - Administrator login successful - IP:192.168.0.2
Thu, 2014-11-27 21:37:54 - Administrator login successful - IP:192.168.0.2
Fri, 2014-11-28 02:05:23 - UDP Packet - Source:89.163.239.65,5086 Destination:xx.xxx.109.218,5070 - [DOS]
Fri, 2014-11-28 02:05:24 - UDP Packet - Source:89.163.239.65,5086 Destination:xx.xxx.109.218,5071 - [DOS]
Fri, 2014-11-28 02:05:24 - UDP Packet - Source:89.163.239.65,5086 Destination:xx.xxx.109.218,5072 - [DOS]
Fri, 2014-11-28 02:05:24 - UDP Packet - Source:89.163.239.65,5086 Destination:xx.xxx.109.218,5073 - [DOS]
Fri, 2014-11-28 02:05:24 - UDP Packet - Source:89.163.239.65,5086 Destination:xx.xxx.109.218,5074 - [DOS]
Fri, 2014-11-28 02:05:24 - UDP Packet - Source:89.163.239.65 Destination:xx.xxx.109.218 - [PORT SCAN]
Fri, 2014-11-28 02:05:24 - Send E-mail Success!
Fri, 2014-11-28 08:03:26 - Administrator login successful - IP:192.168.0.2
Current DSL stats:
Downstream:
Noise Margin:    10.6  dB
Connection Rate:  8242  Kbps
Line Attenuation: 38.0  dB
Power:            0.0  dBm
Max Rate:        9220  Kbps

SuperFrames:      3856343
SF (CRC) Errors:  26
Reed Solomon:    501324594
RS Corrected:    80983
RS Un-Corrected:  274
HEC:              22
Errored Seconds:  18
Severe ES:        0
Interleave Depth: 64
Bitswaps:        18003
Upstream:
Noise Margin:    5.7  dB
Connection Rate:  1144  Kbps
Line Attenuation: 16.1  dB
Power:            12.4  dBm
Max Rate:        1148  Kbps

SuperFrames:      3476853
SF (CRC) Errors:  0
Reed Solomon:    3025812
RS Corrected:    0
RS Un-Corrected:  0
HEC:              0
Errored Seconds:  0
Severe ES:        0
Interleave Depth: 4
Bitswaps:        361
Totals:
Total Uptimes (From SF counts):
WAN:  0 days, 18:15:47
LAN:  0 days, 18:15:47


CRC:  26
LOS:  0
LOF:  0
ES :  18
Bytes Transferred (ppp0):
  Tx: 83182038 (79.3 MiB)
  Rx: 2645128326 (2.4 GiB)

So, no drops, and only 26 CRC's reported - the higher than average (for me) Rx totals are because I downloaded 3 X 1 hour iplayer programmes at 1826, 1904, and 1924 (in HD) just to see what happened.
The graphs seem to show the odd hiccup though - will post some up in a new post when I've had another cup of coffee and a chance to look through them.

[#2]

Morning JayG, very wise, I'm about to do the same.

[#3]

As can be seen from the DSL stats, there doesn't seem to be a need to show the CRC or errored  seconds graphs, so I'll concentrate on Rx and Tx SNRM and sync. What looks to my untutored eyes like a dropped connection at 2306 doesn't seem to be reflected in the overall stats, nor in the ping graph covering that period of time.  The SNRM has been down to a minimum of 8.2dB overnight but has slowly climbed back up to 11.9db as of now - I'm assuming the DLM target is still 12dB.
Edited to modify selection.

[#4]

Ah, there was me trying to look at the graphs and getting an error message - decided to to refresh the post and hey, they're gone
No wonder I got the error message  
And now they are back   

[#5]

Sorry about that - I was still grappling with Routerstats trying to decide what to capture and what not to yesterday evening, so a couple of the graphs turned out to be duff. 

[#6]

It happens.
That 2306 event looks like what's called a "polling" issue, when the RS program can't get the data for whatever reason. More shortly.

[#7]

Nothing later than 0328? I'd be keen to see all the Rx and Tx SNRM graphs right upto now.

[#8]

Only allowed 9 attachments per post, so tried to select the most 'relevant.' 
Will post them all together for the sake of clarity, in two posts, starting with Rx SNRM:

[#9]

Looks good, nothing nasty there, and supports the supposition that the Target SNRM is 12dB.

[#10]

Tx SNRM:

[#11]

The Tx Noise Margin is not as clean as one might hope to see, but currently not a specific issue. Whether it's just indicative of things getting switched on & off, or maybe of a possible line problem is difficult to deduce at the moment. It's currently no different  from many lines not reporting specific problems.

[#12]

I suggest post up some more stats shortly maybe around 1130 when you'll have more Scheduled Captured Graphs and again later on. With this low error rate and current graphs, if this continues (which I think we expect it might) that would indicate that the TG582n is extremely likely to be faulty.

[#13]

Will continue to monitor - the worst scenario would be if the TG582n isn't faulty as such, but just isn't good enough to cope with my line conditions, although that doesn't seem to explain the huge numbers of CRC errors reported when using it compared with the Netgear at the same target SNRM (I'm guessing, of course, as ever!  ;))

[#14]

Quote from: Anotherone

It happens. That 2306 event looks like what's called a "polling" issue, when the RS program can't get the data for whatever reason. More shortly.

JayG,
RouterStats has an intermittent bug related to an internal message sequencing error.  The symptoms are zero or bizarre plots.  The cause is the "wrong" data being captured in the SUMMRY tab.  Usually one can see the bits per tone data in there rather than the DSL stats.
The problem is readily "fixed" by clicking stop / start logging buttons.
I use the alarm "synch fallen below X" to alert me to the condition so that I can apply the fix.
I do not know if the issue is particularly associated with the tg582n or might be seen with any router.  I've had it occur several times a day and never in a month.
Kevin

[#15]

@JayG
I can assure you that a correctly functioning TG582n (which has a Broadcom chipset btw, as does your Netgear) is more than capable of coping with some quite horrible lines, and yours, so far seems quite good. Post up later and I'll look back.

Quote from: Townman

I reiterate my opinion again (this should not be a battle of opinions but a consideration of wisdom).

Sorry Kevin, your "opinion" is not a matter of fact, nor in this case, wisdom.

Quote from: Townman

A straight line plot of increasing CRC simply shows that it is increasing, but not how it is increasing.  Examination of the series of plots from a few posts back shows differing gradients from one graph to another, unless careful consideration is given to the changing scale of the axis, one could be fooled into assuming the rate of CRC generation was changing.  It might be, but as the axis changes it is hard to tell.

I'm afraid you need to go back to school and do some revision of basic 'O' level/GCSE algebra and calculus.
A straight line graph shows precisely how something is increasing - it is increasing at a constant rate.
The rate of change in y with x ie. dy/dx is a constant for a straight line graph. (Also the gradient/slope of any graph at any point gives you the rate of change at that instant).
In this particular case, the actual rate isn't even relevant. The way & when it's doing it, in conjunction with the changed Target SNRM clearly indicates a fault. The history (ie prior to the 582n, see OP) suggests a stable connection, albeit with minor issues. This does suggest that the TG582n should be included in any consideration as to what is fault, going back to the Netgear currently indicates that inclusion was correct.
For your maths revision, here are some useful links
http://www.mathsisfun.com/gradient.html
http://www.mathsteacher.com.au/year10/ch03_linear_graphs/02_gradient/line.htm (note definition at the end of the page)
http://www.mathsrevision.net/advanced-level-maths-revision/pure-maths/calculus/differentiation-first...
Edit: definition remark was on wrong line!