The chamber managers now better handle the recovery for the case of HEPI's master switch OFF but the ISI's master ON (ie. if HEPI tripped but not the ISI for some reason).

Years ago we added the ISI_DAMPED_HEPI_OFFLINE state to the chamber managers, but one small fallout from that is that if HEPI's master switch was off, but the ISI's master switch was on, when someone would untrip the HEPI, the manager would run through INIT which would immediately jump to turn HEPI back on. This would then potentially retrip the HEPI, frustrating users. I've changed the INIT state to just go to ISI_DAMPED_HEPI_OFFLINE for this case, as it should.

I tested this with SEI_ETMY and SEI_HAM6 by setting the nodes to ISI_DAMPED_HEPI_OFFLINE first, then tripping HEPI by changing the trip level, turning OFF the HEPI master switch, then untripping the HEPI. All tests successful and I loaded this into all other chamber managers.