Here's the
schematic I posted. Off the cuff a modified panel drive circuit would include a large series cap like C2 170uF (which is about 6 ohms at 150Hz) and possibly some series R as shown. To really give an authoritative answer on what would be electrically correct, I would need some guidance on the transformer turns ratio from ML. Then a SPICE simulation would be run to fine-tune, really not a big deal at all.
The goal would be to reduce the high-pass function of the series cap below 150Hz so as to not add another pole to the digital filter. Really all I would want to do is DC-block the transformer. This could mean the pass band to the panel would (theoretically) extend pretty low, probably less than 70Hz. But of course signal content from the digitally crossed-over amp would only contain higher frequency content.
I did notice that, according to this schematic, the stators of the panel are both driven differentially even with a single-ended input. One stator is
not ground.
