Is this SNR fluctuation normal?

[#1]

Hi,

Attached a graph showing my SNR plot over the last week. As you can see there is a regular fluctuation pattern. It's been like this for months now - not causing any particular issue but I'm just interested to understand.

I also have another Plusnet bband link that shows a similar pattern.

Any advise if this is normal would be appreciated.

Cheers 

[#2]

Hi there,

I'm guessing it's not a significant fluctuation if it's not causing any side effects. It's difficult to say if that's "normal" as every line operates within it's own means. It's possible there may be something causing interference on the line so it may be worth seeing if there's anyway you can set up monitoring which will record the time of the fluctuations.

[#3]

Hi,

Thanks for the response. From the monitoring I use I've zoomed into a couple of days of the SNR data, graph attached. This shows the time of day in more detail.

I'm curious if as the pattern is almost identical 7 days a week if it could be DLM or some other throttling technique being used. My other working theory is some type of REIN but a bit stuck to know what it maybe given the time of day pattern - ie unlikely to be street lights.

Cheers

[#4]

That looks entirely normal

Here is mine about 1.5 Km overhead before heading underground to exchange 

[#5]

@atc

@ian007jen  Both traces look perfectly normal.

Most ADSL lines suffer to a greater or lesser degree from MW radio interference, which becomes stronger

during the hours of darkness.

In both these cases it seems the DLM is just doing its job of reducing the SNR margin at night to counteract the

interference.

[#6]

The above answer is both correct and wrong!

SNRM is the margin of the signal above the back-ground noise.  During the hours of darkness the increased impact of interference from MW continental radios stations (the correct bit) reduces the the margin of the signal over the noise.  The DLM has no part in this observation.  If the margin reduces to zero and there are consequentiallyreoccurring disconnections, the DLM will interview to raise the target SNRM in 3 dB steps.

One might see as much as 2.5dB variation between day and night SNRM levels due to this.  My line sees about 0.5dB.  The amount an individual will see depends on where they live and the quality of their internal wiring.

Though the above explains the day-night variation of those SNRM plots, it does not explain the persistent jitter.  It might be worthwhile you looking at going around the hime switching off various electrical devices - especially ones using 'power packs' (unplug them) to see if the jitter disappears.  Switched mode power supplies as used by laptops, routers... etc can deteriorate over time,giving off electrical noise whilst otherwise appearing to be working OK.

[#7]

---

townman wrote: The DLM has no part in this observation.  If the margin reduces to zero and there are consequentially reoccurring disconnections, the DLM will interview to raise the target SNRM in 3 dB steps.

Aye, you're right - thanks for the clarification.

---

townman wrote:
Though the above explains the day-night variation of those SNRM plots, it does not explain the persistent jitter.

Hmmm, not so sure about that. Even on the expanded scale of the second graph (Message 3), it's very compacted for the period covered. A more 'normal' scale would not look so 'jittery', I feel.

[#8]

Looking at the graph on message #3 I would say there is too much jitter on the daytime plot. Either it's a very noisy environment (from an electrical/electronic point of view) - noting it's a business? - or maybe the wiring is not upto scratch, even the drop-wire perhaps!

The day-night variations are pretty typical and can be greater than 2.5dB even upto 4dB on occasions even with a reasonably balanced line with wiring to the correct standard. Wiring that isn't upto the right standard will just exaggerate the problems.