On Sun, 08 Jan 2006 12:18:35 +0000, wrote:


The direction of the power flow through the phantom phase
link then depends on the relative mechanical loading on the two
rotors. In the two limiting conditions of both motors idling or
both motors loaded to the same fraction of their rated HP, each
motor is self sufficient and little or no third phase power flows
through the link. In the normal case of one idling and one
loaded, the power flow is from idler to load motor.

It is equally valid to regard this current as a circulating
current within a parallel connected system or as the current
developed in the series connection of an idler generator to its
load motor

See my later post that addresses this issue of series vs parallel.
The situation is quite clear if idler and load are delta-connected:
there is no neutral and phases are clearly in parallel however they
may or may not be connected to singlephase or 3phase mains.

You know of and about Delta - Y transformations of three-terminal
networks, particularly symmetrical threephase machines.

Everyone seems to agree that any neutrals that might exist in an RPC
system are not connected. If neutrals are not connected or
don't exist, then I say they can and should be ignored. Absent
neutrals, the question of phase-to-phase and device-to-device
parallelism seem quite clear to me.