When I was working with flooded lead acid batteries, one of the design considerations was the system voltage. We generally tried to get the voltage as high as possible since in a DC setup this reduces the current. Lead acid cells have an inverse relationship between current and capacity that's called the Peukert curve. The higher the current, the less actual capacity (amp-hours) the battery yields. At electric vehicle currents (say 300A!) the effect is significant.

So, does this effect appear in our LiPo batteries? It would be an argument for connection in series to raise the system voltage. Anyone for 32V?

John Witt

Yes, we can use 8 cells instead of 4, but the same weight will need half the capacity and the same power will need half the current - what leads us to the same situation as before - current ~7C in both variants, means the same situation and no advantage.

John,
Except for the hookup wire and ESC current capability, the efficiency of a low voltage-high current-high kV motor vs the high voltage-low current-low kv motor is identical.

Let's examine my earlier setup (because the math is easier!).

Also, to keep things more or less the same, we will assume the same motor family--in my case since I use Scorpion motors, we are dealing with the same 3020 motor, just varying the motor windings to adjust the kV.

If you are flying a particular plane with a particular prop, your power needs (as you like to fly it) can be satisfied by many different combinations of the basic same weight motor and battery choice. So what I mean here, is that you can hook up things as you like.

So lets start with the motor. With a 4s setup, you want a kV in the ~ 800 rpm/volt range. In the Scorpion 3020 case, this will be a motor with about 8 winds per tooth (nominally called a 3020-16). Lets say you draw 20 amps in level flight with this setup (a little higher than I draw with a 4s setup with this 3020-16 in my Nobler). If I rewind this same motor with 4 winds/tooth (now a 3020-8), I would get a motor with a kV of 1600 rpm/volt, and with a 2s battery, I would now drive 40 amps through this version, and get exactly the same torque as before. The heating in the windings would be the same, since the lower wind would be 1/4 the resistance of the higher wind (1/2 the number of winds, but also twice the cross sectional are of wire since I could use a gauge 3 units lower as the fewer # of winds of the original gauge takes up less space). This 1/4 resistance exactly balances the I² factor of 4 so the energy lost to I²R in the motor is the same.

So the efficiency of the high kV-high current setup vs the low kV-low current setup is identical with respect to the motor.

Since as we said before the lost energy due to internal resistance of a 4s2100 mAHr pack vs a 2s2p4200 pack (both made with the same 2100 mAHr cells) is the same. In the first case we are drawing 20 A from each cell (20Amp draw), and in the second case we are drawing 40A from the pack, but since the we have two 2s2100mAHr packs in parallel, each 2s pack, and therefore each cell in the 2s pack must be supplying 20A, again the heat in each of the 4 2100 mAHr cells is again exactly the same (as well as weighing the same --obviously).

So the only difference is you need to have a 40A ESC with the low voltage option, vs the 20 amp ESC for the high voltage case. The larger ESC does cost you more. But if you try to extend the high voltage to a higher cell count, at some point, you will need to have a HV ESC, which will also cost you more.

If you are worried about the heating loss in the wires, the easiest option is to increase the hookup wire gauge, but for the currents we are dealing with, this isn't a really big deal in terms of inefficiency.

So what I am saying, as long as you can adjust kV to the battery, there is no intrinsic efficiency gain. So with my eVector, I will start with a 3s2p4200 pack (two 3s2100 mAHr packs in parallel) and a ~1000k motor. I could also hook up the two packs in series as a 6s1p2100mAHr pack and a 500 mAHr motor and have basically the same weight and efficiency--only the ESC (40A or a HV 20A) would be needed to be compatible with my choice.

I hope this makes sense.
---edited to remove some typos (probably not all!)