Story about an XLPower 4020 motor
Expert level of RC heli modelling: motor rewinding
I bought an XLPower 4020 motor to test it in my Logo 480 xxtreme instead of the existing Scorpion 4015-1070KV. Because there isn't too much room under the canopy of the 480, I decided to cut the wires back a little bit, then I noticed that the 2 × 12 0.42mm wires are soldered and cleaned up as much as it's required originally. After shortening the wires I could not put new connectors on so decided to rewind the motor.
This gave me an idea, can the winding make the same KV with a bit higher torque? The double Y connection allows this in theory. I have read many times that it's not a good idea to use more than 18mm2 copper wire on a 40mm long armature, because it will be hard to make it with single strand, and the internal resistance is going to be too low, which results too high current on heavy loads. This saturates the core which blows the battery up.
Iron cannot be magnetised endlessly, at a certain point (saturation) no matter how much more current is drawn by the motor, the magnetic field is not going to be stronger.
12 x 0.42mm threads are collected together in parallel in the original winding, and this is the diameter of one turn. There are 4 or 5 turns on two arms next to each other, this means 4.5 loops on one stator arm in average, and this is specified on the data sheet (4.5T). Original winding is delta, in other words: the three phases are in a delta shape, where the three corners of the "triangle" are connected to the ESC.
The RPM of a delta connection (in case of equal turns) is the square root of 3 (1.73), the double Y connection gives 2, therefore when I want to use YY on same KV I must put as many turns as there are in the original one multiplied by (2 / √3) on each stator arm, which is 1.15 × 4.5T which is equal to 5.2T, that's why each arm has got 5 loops of 1.5mm wire. I used 1.5mm wire because 10 wires in each groove gives 17.7 mm2 which is a tiny bit less than 18mm2 recommended.
It is important, few thick wires make less unfilled space than many thin wires. As I used 1.5mm wire, I put about 20 per cent more copper in the motor, which generates lower internal resistance (copper loss), about 10-15mΩ. This matters when the motor takes higher current and the motor is under higher load. The copper loss is I2 × R. When resistance is lower, loss is lower. High current generates high loss with high resistance. During the idle running with no load, this is less important, because the loss is higher on the core, but the current is low. This is 6-7A (with 6S battery) in my case, what is not too low, but this current builds the magnetic field up to start the motor spinning and to overcome the mechanical resistance produced by bearings for example. And this current needs to overcome internal resistance of copper as well, which is not too high on idle run: 62 × 0.01Ω equals to 0.36W. But with 100A peaks, the current loss goes up to 100W.
Each loop is a challenge during the winding with this thick wire. I had to prevent damaging the double enamel insulation, it could not be loose, otherwise the wire on other stator arms could not fit in and so on.
But the end result? Well, I can’t complain, it is quite satisfying: 1100KV and a very powerful motor from a good XLPower base.
Thanks to Alan Matuszczyk-Robinson for his English proofreading.