On Mon, 30 Aug 2004 10:58:19 +0800, Old Nick wrote:
I have a rough (foggy with age and 25 years of little- to non-use)
idea of starting currents and no-load synch currents of sync motors. I
know what happens as you start to load them and they slip a bit and
start using power.

But the stopping point for me is that an idler basically MISuses a 3PH
motor, and all the rules come under question immediately.

Thinking of it as a motor just confuses the issue. Think of it as a rotary
*transformer*. That's the function it actually performs. Like any transformer,
as long as it is big enough (at least 1.5x the hp rating of the largest single
load), and is designed correctly, it does its job with very good efficiency.

But after all the tweaking, the RPC _still_ appears to suit a certain
load; not given the cost, or the size of idler making it easier, but
because if you tweak for no load, then use various loads, you get
efficiency losses. Again, this is not a cost issue, so much as limited
power supply and making that work. So I realise it takes a lot of
power costs to _pay_ for the thousand dollars a motor gen will cost me
to set up.

No! Tuning has very little effect on RPC efficiency. What tuning does
is improve phase to phase *voltage balance*. As long as phase balance
is reasonable (and reasonable can be a pretty rough approximation
when you're only driving motor loads, which is why so many people are
able to use RPCs with no tuning), efficiency will be good if the RPC had
good efficiency *as a 3 ph motor*.

The primary factors which determine RPC efficiency are the winding
resistance (wire size), the magnetic path lengths, and the magnetic
permeability. The people who designed the motor you use as a RPC
determined those things. If it was a good motor, it'll be a good RPC.
End of story.

The main reason to tune a RPC is if you have sensitive loads, like
control electronics, running off of it. Some (by no means all) electronic
loads are picky about phase to phase voltage balance. An example of
this sort of equipment would be a CNC machining center. Some, again
not all, welders will bitch about poor phase to phase balance. In those
cases, balancing the converter, or making sure the sensitive loads are
fed by the stiff legs (L1-L2) will resolve the problem.

Gary