Thanks for the welcome pete.
Yes, I am a bit of a johnny-come-lately here - I've worked my way through several pages of the thread but alas not all.
I'll try the parallel connection - I never thought of that, I was just going to ground both ends of the coil and hope that the resulting low current loop didn't cause too many problems, using the driver cct as a low-impedance load will do the same thing - it should certainly tidy-up some of the stray noise/feedback issues. An open-circuit secondary coil can result in core saturation if the primary is driven too hard, but I think (hope) you'd have to drive it really hard for that to be a problem here.
As an aside, I'm currently using the trusted LM386 amplifier to drive it, but I am also developing a Class-D amplifier (kind of inspired by the sustainiac patent - but without the deliberate errors ) - the prototype works okay on the bench, but it's too big to fit inside a guitar cavity - I'm going to investigate using a switch-mode psu control chip for this sometime, (circuit-wise it's basically the same thing), which will drastically reduce the cct board size. If it works I'll post it here.
dean©