Every time you change the modem you should follow a set of steps to minimise the potential for DLM interference.

In reply to a post by MHC:

When you want to change over do the following:

Turn off the modem power
Wait 2 minutes (longer than needed however it will make sure)
Disconnect the RJ11 cable from socket to modem (HG612 or HH5)

Power up the replacement device - without it being connected to the line.
Wait 2 minutes for it to fully boot and stabilise.
If using an HG612, connect the router.

Connect the RJ11 cable from line to modem.

Wait for sync, read stats.

Doing it that way should ensure that each loss of connection is seen as a deliberate act rather than a loss of sync due to noise, attenuation or corruption.

Can I suggest doing te above and leaving it OFF for maybe 30 minutes. Turn off anything close by in the house - fridge/freezer, central heating or boiler, TV, HiFi, Xmas lights, &c. And then doing the reboot. Give it 5-10 minutes and then turn everything on again. There is always a chance that a local noise source is giving problems.

Is the sync speed still around 1Mbps? And has the noise margin changed at all? The IP Profile will probably be stopping actual speeds getting above 1Mbps for a while even if the sync speed has increased.

Thanks. I've tried the reset and will let things settle for a while.

Latest stats (Dlink DSL-504T) are here.

Going back in my memory - about 15 years!

1152 kbps - that may well be the FIXED 1Mbps downstream rate. Having the same speed on two different modems suggests you have been given a fixed speed profile - could be due to too many disconnect/connect cycles. I know that 0.5 Mbps was 576kbps so 1Mbps could be 1152kbps.

Don't forget this post.

You are correct that BT Wholesale fixed speed 1Mbps is 1152kbps sync, but I think it is just coincidence, as the upstream would be 448kbps.

Those D-Link stats are wonky.

If people look back at your earlier stats, the attenuations are in step. The new ones have an almost impossible upstream value.

Which then raises the question of which of the two is correct about the downstream noise margin. The first at 19.2dB or the new at 9dB.

I think you may have a highly variable noise source, with the extra noise being a fixed amount when present. That would account for the earlier results from two modems, but not these latest ones.

Edited by RobertoS (Tue 01-Dec-15 12:24:23)

Oh, yes it could. If the noise is still present when you took the D-Link stats. Try taking them every hour, without a re-sync, and see if the noise margin jumps.

I don't know if RouterStatsLite works with the D-Link. It does with the Netgear 834. The graph of the downstream noise margin could be interesting. Sampling at 7 or 10 second intervals.

Edited by RobertoS (Tue 01-Dec-15 12:28:20)