The ADSL standard incorporates a dying gasp signal on power-down. I vaguely remember some discussion on this topic and I think the consensus (for what it's worth) was that BT make no use of that signal.

And neither is your explanation. Perhaps I should have said a lot of myths exist about what actually happens during the 10 day period.

Over the first 10 days all that is set is the MSR for the line and I am sure I do not need to expand on what that is.

Where you go wrong is that you assert that any drop of connection will result in a noise margin increase, multiple reboots are needed for this to actually happen in a short period. Why? So people with USB modems don't get margins at 15dB all the time, and to allow people to turn off kit at night.

If leaving the modem on solidly for 10 days to avoid margin changes, what happens at the end of the 10 days. The processes that alter the margin still continue don't they. Just that the MSR and FTR are set.

Your suggestion will leave people with extensions not working also, i.e. surely people should either tidy up the wiring or just use the kit in the same place as they normally would.

Maybe i should make myself clearer for you.
What i meant is that if you reboot your router or modem then when you turn it on it will more than likely sync at a lower speed and then this will be taken into consideration by rambo and max as a lower sync rate and then what will happen is the msr and ftr may then be calulated lower than they should be and then if you try to call your provider to say that you have a slow speed they may not be able to help you.

For instance if your sync rate is 1mbps but the ftr is 500kbps but you now that your line can handle the full speed then your speed would be considered by bt and your provider to be normal at 1mbps.

So unless you want to risk a drop in sync rate and a lower ftr then do all the rebooting you want.

And another thing is that extensions are made for telephony equipment and analog signals not broadband equipment and digital signals so people should not really be using them for their broadband.

On good lines I doubt it makes any difference if you power cycle your modem or not (unless you do it more than 10 times an hour). A good line will generally always sync at the same rates. On a poor line it may be beneficial to leave a modem on all the time. I'm not convinced of any real benefit though, for instance you may well be better off powering down the modem if a thunderstorm is expected.

Well obviously you would be better powering down the router/modem down during a thunderstorm but unless you have to what are you going to gain from turning it off anyway.

I get the impression from reading these forums that a lot of poor sync events happen during non-use of the connection. Each line is unique and may need experimenting with the power cycle regime to decide what is best in practice.

"And another thing is that extensions are made for telephony equipment and analog signals not broadband equipment and digital signals so people should not really be using them for their broadband. "

But to recommend just using the test socket for the first ten days is daft, since after that period people will plug all their kit back in and see a drop in sync, better to suggest the various methods of improving the connection in the first place.

For those using their broadband for 3 or 4 hours just in the evening it makes sense to switch the modem on once when they come home and off before going to bed.

a) Saves a fair bit of power over the course of a year
b) Avoids the resyncs for the many hours they are not using it.

I tend to forget there are people who work for a living.

In reply to:

---

What i meant is that if you reboot your router or modem then when you turn it on it will more than likely sync at a lower speed...........

---

Pure twaddle ! The router will sync at a speed which gives a pre-assigned noise margin. If my router was last sync'd at 3AM, when HF noise is at its highest, and I sync it again at noon, there's an excellent chance that the sync speed will be considerably higher. That's why there is a noise margin target set - to allow for variations in line noise over a 24 hour period.