Dsturzy wrote on Mon, 03 January 2005 20:33 
This certified electrician is the same guy that has accedently forgoton to tie in a nuetral when other sound companies have come into our house.

Oooh, thats not good  that could seriously ruin your day, not to mention the rig.
Anyway, there are two facets to your problem.
a) You have an unbalanced three phase system. Although you are only using two phases, its still a three phase system, just with zero amps in the third phase.
So, if you just had lamps instead of amplifiers, you would have phase currents of P1=15, P2=15, P3=0. Plug those numbers into a formula for calculating neutral currents in a three phase resistive system, you you get a neutral current of about 15A.
See the math
http://www.electrician.com/electa1/electa5htm.htmYou were hoping for zero amps. You dont get that  if you had a perfectly balanced three phase resistive load, you can get zero amps, but you aren't running three phases, nor a resistive load, so there will be a return current through the neutral.
Thats all stright textbook three phase, nothing contoversial here.
b) Now it gets worse. I'm doing some extrapolating from things I know to things I haven't measured, so its a bit speculative.
The big issue is harmonic currents and nonlinear loads.
I'm going to make an unproven assumption here that a switchmode PA amp works  in practice  much the same way as any switchmode power supply, for example, the one in your PC. Switchmode power supplies (SMPS for short) have a particular characteristic, in that they dont draw current linearly, but as a sharp edged pulse.
Have a look at this link, and check out the graph with the thick red line:
http://www.marcspages.co.uk/pq/3230.htmNow this nasty current waveform generates harmonics, and the real sod is that if you overlay the harmonics waveforms on each of the phases, particularly the third harmonic at 180Hz (assuming USA 60Hz) the currents in the neutral add, not subtract. Again, to save me typoing and drawing, heres another link:
http://www.marcspages.co.uk/pq/3240.htmThus in a power system with an unbalanced load and nonlinear current draw, you get neutral currents that frequently exceed the phase currents. Despite what Marc (whose pages I link to above) says, and I dont disbelieve him, and his logic appears sound to my nonmathematical mind, I've never seen N current greater than twice the biggest phase current.
I can also say that it is standard computer room practice to build power distribution with 200% N current capability. So if you have a 200A per phase service, your neutral capability is uprated to 400A. Otherwise, you can get overloaded neutrals, and thats when the fires start. That 200% neutral has to go right back to the transformer, and the transformer will dissapate the excess N current as heat. To do this you need specially designed transformers, known as K rated transformers. Otherwise, the transformer get overheated.
So, getting back to rock and roll.
a) If you can, run your system single phase. Then you cant get all this bad stuff happening.
b) If you must go multiphase, use all three phases. Derate your power system by 50%, ie if you intend to connect to a 50A supply, draw no more than 25A per phase. Its no use putting 200% neutrals in your distro, and then plugging into the house power, which only has 100% neutral, and running full current from the house supply, its neutral will be overloaded. Remember what will happen if the neutral melts....
So, thats my theory. I haven't found any data on switchmode PA amps causing nonlinear harmonic current problems, its just my suspicion, but it fits the facts you've measured.
I'd be interested in comments, particularly from anyone who has a really big PA with switchmode amps on three phase, who has a clampmeter...
(and yes, lighting rigs suffer from the same phenomenon)