I heard the call . . . and went to MDWS to see an absolutely appalling SNRM graph for
pianist.
Obviously the actual synchronisation speed graph reflects the bulk happening of the SNRM graph. I was also monitoring CRCs and FECs (in panes 3 & 4). The too showed events which "line up" with the distorted shape of the SNRM curve.
Swapping over to view the Hlog graph in place of the FECs, we can immediately see a circuit with infrastructure defects. I would certainly agree with
N*Star's tentative analysis -- it does look as if there is at least one bridging tap present, if not two.
I think the time is ripe for
pianist lloyd to give us a guided tour of the wiring from the point where the service feed enters the property. Photographs of each terminal (junction) box and socket found, both an exterior and an internal view (so we can see the wiring as currently connected), coupled with a descriptive sentence or two.
It is quite clear . . . there is one or more physical defects as to how the circuit is currently wired.
One final quick question, for now: Is there an alarm system wired into the circuit?