Kv is the number of rpm the motor can turn per volt, right? So the battery drains down so you have to set your governor rpm below that number.

Motor 700 Kv
6 cell Battery 25.2v at the start
17640 rpm possible

19.2v after flight (3.2 cell)
13440 rpm possible

So how much head room do you want if you want a 10,000 flight rpm? Someone told me the same 6 cell with a Kv of 820 would be less efficient. What makes it less efficient?

Thanks,
Motorman

So, you got it all simplified, and now you want it complicated right back up.  OK, I can do that.

The rule of thumb is to give yourself about 25% headroom -- meaning, multiply the nominal voltage of 3.7V/cell by about 75%.  That accounts for the resistance in the motor and wiring, ESC inefficiencies, and the actual voltage drop in the battery under load.

The reason that using more voltage than necessary is inefficient has to do with losses in the system.  When an ESC is running at part throttle, it's switching the voltage to the motor (at around 10-20kHz these days -- the exact number depends on the ESC, and I don't follow who does what).  So at part throttle, the active pair of wires into the motor is seeing the voltage pulsing between +battery voltage and -battery voltage (or, depending on the ESC, between +battery voltage and shorted).  Mechanically, the motor reacts to the average of the voltage it sees, but the more pulsing needed to get the same effect, the more losses there are.  The most efficient setup would be one where the motor is running at full throttle, so that there's no pulsing at all -- but this would mean you're running at zero overhead, so you would hope that doesn't actually happen.

Thanks Tim, interesting stuff. I guess that's a good reason to keep the leads as short as possible

You can always go overboard.  I'd keep them as short as is reasonable, then add an inch or so because I always, always underestimate the length that a cable needs to be.  If you get a note higher than a middle "C" when you pluck a battery lead, it's too short.