Has anyone ever heard of using a "AC curve test" to set neutral on a DC machine. We have a client who has requested this procedure to set the neutral on (2) 1500 HP DC main propulsion motors that power a research vessel. They have been experiencing sparking on the commutator at full load on the starboard motor that is eliminated after the comm has been stoned but recurs after approximately a month in service. We suspect loose commutator bars (it's an overhung 3 segment glass band comm) but they suspect the armature may be set off neutral. They are returning to port on Monday and would like us to perform this test. Any thoughts?