Primary impedance with pulse skipping

[flyingperson23]

I'm planning a large coil with hopefully 1ms on time for playing music, and am considering high vs low impedance designs. Most of the information I've seen about impedance was in regards to non pulse skipping design, so given that I am using pulse skipping and long on times, what difference will the impedance have? My current understanding is that low impedance will have higher MMC energy, so brighter (and louder?) arcs from faster current rise, but will be consistently skipping a lot of the pulses, reducing overall power. Would this be overall desirable? Also, could I add a small bank of film caps on the bus in addition to electrolytics to speed up current rise with a higher impedance design?

[davekni]

Would this be overall desirable?

Excessive pulse skipping (excessively low impedance) will reduce coil power for a given IGBT current rating.  Also, pulse skipping causes more power dissipation within IGBT diodes.  If you are using IGBT bricks, diodes are usually capable enough.  If using an array of smaller IGBTs, diode power may be limiting.
So I think best is just low enough primary impedance for pulse skipping to occur for perhaps 10% of cycles.

Also, could I add a small bank of film caps on the bus in addition to electrolytics to speed up current rise with a higher impedance design?

Film capacitors (often called snubber caps) are needed at H-bridge (at IGBTs) to handle switching current locally with minimal inductance.  I don't see any advantage in more film capacitors than necessary to minimize spikes at IGBTs.