I've been building hi-sensitivity electronic GFIs for years,
OK, but realize this, when a common mode signal comes along, that transformer
has a whole new set of specs., i.e. very low impedance and the inductance is
pretty much nullified. What's left is going to resonate higher.
Now for the stretch. . . . . .
Any large unit using a switched mode power supply near enoungh to cause a
transient on the AC line could trigger a reaction of some type. Now even if the
SMPS didn't radiate, I've seen some where there is no resistor going to the
rectifier. There is when you first fire it up, but then a relay closes and
shunts it out. If it has pretty healthy rectifiers that happen to switch fast,
that will put a transient on the line, especially under a higher load and
especially when they feed like an 820uF filter cap. Iv'e seen some things that
rectified the line into two 470s in parallel. If the diodes are switching fast
and the ESR of the main filters is/are low. . . . . . . . . .
I know this is far fetched, perhaps barely plausible, but explain it some other
way, I'll listen. This is however all that I can think of right now.
JURB