Attenuation Changes - an observation

I needed to test the RCBO on the dedicated circuit for the small comms room I have - it protects just that room with just 6 or so double sockets. Turned off all devices, disconnected the line from the modem and then carried out my checks.

I left DSL stats running to see if there would be any significant changes. Re-booted all the devices, modem, router, switch &c and waited a few minutes before reconnecting the line - all cable remained the same.

On DSL stats all looked fine as I scrolled though although there were a few minor changes. The sync had increased by around 0.5Mbps and a few missing tones had returned.

I then looked at the attenuation figures: Actual LINE attenuation had decreased by 0.8dB in U1, 1.2dB in U2, 0.3dB in D1, 1.0dB in D2 and 1.3dB in D3. The signal attenuation has remained constant on U0, U1 and U2 whilst dropping bu 0.5, 0.9 and 1.2 dB on the Downstream.

Overall looking at the graphs the SNR Margin on Upstream has risen form around 8dB to over 15dB.

How long had the modem been up beforehand? I'm thinking about the effect on reported line and hence signal attenuation of bit-swapping and ultimately tone removal over a period.

Alternatively, simply a noise source not present during this sync that was present whenever the previous one took place.

Modem has been up for several months ... Max achievable and sync have not changed (significantly) for quite a while and like you I think it is just down to tones that had been masked out being re-introduced.

As I said - just an observation although it may show that an occasional nice clean reboot of everything can tidy things up.