Gary Coffman wrote in message . ..
On Wed, 24 Mar 2004 20:03:28 -0800, "Robert Swinney" wrote:
gradstdnt sez:

". . .currents for each leg are not as nicely
balanced. . . . "

You should not be measuring current in the 3 legs when attempting voltage
balance. It will totally confuse you! The current flowing in a 'rotary
phase converter / parallel load' is very complex. Some of the current looks
like it should be flowing in 2 directions, against itself. Of course, this
doesn't actually happen, but the reactive currents combine, add, and
subtract in the 3 legs in ways that make them an extremely unreliable
indicator of balance. Stick with voltage balancing techniques and leave the
clamp-on ammeter for the final power factor adjustments.

Bob's right. Balance the phase voltages, not the currents. The only time
you want to measure currents is when tuning out power factor on the
primary.

Gary

I would like to clarify my three phase amp measurements. I was not
measuring currents to/from the converter. I was masuring currents in
each leg to the load motor while under power. I have measured both and
they are different. My initial thought was to better understand how
much current I was providing in the generated leg. I measured the
other legs for reference and found the results interesting. When the
voltages were blanced to within 1%, the currents were also nicely
balanced. I thought this was the goal of balancing but others have
pointed out issues with idling currents.By removing capacitance, thus
unbalancing the loaded system, the currents in each leg to the load
motor were also thrown out of balance. One let got hotter, as two
others got weaker.
In general, any further inbalance I have in the converter decreases
the current provided in the generated leg.

Thanks again for the input. I will work with balancing the converter
for my min and max loads. This should provide me with the boundary of
my operating conditions. I can then better evaluate the tradeoff
between loaded and unloaded voltage balance.