[Herald1360]

Mildmartin's comment about asymmetric waveforms on top of other residual leakage from equipment on the same circuit sounds promising. The inrush current to a transformer primary at switch on is very asymmetric- huge spikes on top of one side of the sine as the core saturates until the the magnetic circuit settles down after several trips round the hysteresis curve!

It would be well worth connecting this setup, via a plug in dedicated RCD, to an unprotected feed. That will establish whether it's the set and its limiter combo or simply too many small leakages that is the problem.

FWIW, though, the lamp limiter will act as a reasonable "soft start" resistor to mitigate inrush current spikes down to maybe 10x FLC from possibly 100x or more.