I am now in the process of sticking all modems onto the wall, with very short cables to wallsockets. My wife Janet did all the work of course, so the thanks go to her. We using a certain type of velcro sticky pads, which take 12 hours to 'cure' and are not sticky initially. There is a velcro pad stuck to the wall and one stuck to each modem. One initial snag we encountered was that the feet on the modems are too deep, deeper than the thickness of a pair of the velcro pads. So as a kludge, we made a spacer out of a pair of back-to-back velcro stickies, to increase the thickness enough to clear the feet. This worked really well though. The stickies are rated at 5kg weight so miles more than enough.
Thanks to a suggestion from Burakkucat, I am using a BT ordinary telephone-only front with the one NTE5C wallsocket that I have with a specially made 0.25m BT431-to-RJ11 cable from an Ebay custom supplier. So we are now finally filterless on this wretched NTE5C that was fitted by the BT man despite my wishes - because I was in bed not overseeing matters - and the nuisance BT Mk4 SSFP that was in there just to make the connector the right fit is gone.
Despite Burakkucat warning me not to use CW1308 cable for such a link, I screwed up
and indeed somehow ordered just such an item, having got confused through all the pain drugs. So I am duly warned that the plugs might fall off the cable.
The other wallsockets are NTE5As with single RJ45 sockets provided by
Andrews and Arnold shop-supplied DSL-only straight-through fronts, no SSFPs, so for these I am using RJ45-to-RJ11 cables from the same custom supplier, again a mere 0.25m. These have a nicer snap than my beloved Tandy/ADSL Nation cables because having RJ45 plugs not RJ11 at at the wallsocket end they are the correct fit.
I am putting TDK split ferrites on all the modems’ DC cables, rest assured with the required dose of skepticism. We can only try to measure and see if it improves things but such measurements will be very difficult and in any event I would imagine any effect would be ridiculously small if non-zero.
Janet had done an excellent job of routing signal and DC cables well away from all mains. Now the DSL signal cables to the modems are ridiculously short and they and the modems themselves which are plastic-cased, not metal and not conductive-coated, are even further away from mains. So every little helps, cannot say by how much. At 65dB downstream attenuation, that is not much signal voltage at all at our end, that is my thinking. I do not know by how much the tx power might be increased for us at the exchange end, if at all, to help us out with our extremely long line at 7300m, but according to my calculator that is a voltage reduction by a factor of 1778, so that each 1V becomes 0.56mV.