0
\$\begingroup\$

Suppose I have an ESC + BLDC motor setup and it's working fine. Now I want it to have some burst ability, that is to produce much more mechanical power in a few seconds. But the Kv is too low, so I changed this motor to another one with the same size and higher Kv.

The use case is the long-range fixed-wing drone. At cruise it requires very low power, but needs really high power at takeoff and climb.

I am worried that this could lower the efficiency at low power conditions. Since the duty cycle is low, the current of the input wire is much higher when PWM turns on and creates more voltage drop. Since the Kv increases, the inductance of the motor coil also drops and the ESC needs to increase PWM frequency to keep the triangle waveform of the motor current smooth.

Can the efficiency drop become significant? How to analyze it? How should I change ESC parameters to get higher efficiency at low power?

-----update-----

I've also considered using delta wye transformation. Use wye at cruise and delta at takeoff. But most relays have a contact resistance of more than 50mΩ which also makes a great efficiency drop. Is there a good delta-wye tramsformation implemention with very low contact resistance?

Maybe low-resistance relays can be quite heavy and don't worth the effort. We can take the same weight to make the motor larger: To make the triangle wave inside the motor smoother, we can change the inductance of the coil by increasing its number of turns. To keep the Kv constant, we use weaker magnets. To keep the winding resistance constant, we use a larger motor. But this also doesn't sound clever.

\$\endgroup\$
9
  • \$\begingroup\$ Sorry for the mistakes. I have corrected them. \$\endgroup\$
    – lotsof one
    Dec 16, 2022 at 8:51
  • \$\begingroup\$ (FWIW, no alternative promising substantially lower losses occurred to me.) \$\endgroup\$
    – greybeard
    Dec 16, 2022 at 11:57
  • \$\begingroup\$ What is the power source? 12s would open interesting possibilities. \$\endgroup\$
    – greybeard
    Dec 16, 2022 at 11:58
  • \$\begingroup\$ 12s doesn't sound good. Suppose I changed my battery setup from 6s to 12s(22.2V to 44.4V), the Kv will need to be reduced by half, and the number of turns of the coils will be 2x as much as before, then the winding resistance becomes 4x as much as before. Power doesn't change, current reduced by half, so coil heat loss will remain the same. As for the coil inductance, it becomes 4x, and voltage is 2x, so the current change rate will be half, but the required current is also half as before, so it doesn't reduce efficiency loss by ripple triangle wave current in the motor. \$\endgroup\$
    – lotsof one
    Dec 16, 2022 at 16:13
  • \$\begingroup\$ The only advantage is that it reduces the energy loss on the battery-to-ESC wire and the ESC-to-motor wire. I don't see many advantages to switching to 12s. \$\endgroup\$
    – lotsof one
    Dec 16, 2022 at 16:13

1 Answer 1

0
\$\begingroup\$

Yes, motor efficiency will decline at lower throttle level, due to the triangular current waveform causing more heating in the motor windings. In another question I showed test data for a very high Kv motor powered from 7 V at full throttle, then powered from 12 V with throttle lowered (to ~60%) get the same output power. As motor current became fully triangular at the PWM frequency of 12 kHz, the required power input increased by 15%.

My test motor had a Kv of 3750 rpm/V. A motor with lower Kv should have higher inductance, causing less ripple and achieving higher efficiency at part throttle.

The controller also has some loss. An ESC with 'synchronous rectification' or 'active freewheeling' has lower loss than one that just relies on the FET body diodes to recirculate back-emf current. This may reduce losses by a few percent, as well as reducing heating in the controller.

Some controllers have an option to set the PWM frequency. A higher frequency should reduce current ripple, but increases switching losses.

Calculating the total loss in the system accurately is tricky due to the large number of parameters involved and difficulty of deriving them from published data. The best way to analyze it may be to simply fly the drone and measure its performance in the air. Onboard telemetry that includes battery voltage and current, airspeed and altitude can be used to determine the cruise power and overall efficiency.

\$\endgroup\$
6
  • \$\begingroup\$ I think I really need to buy a scope if I want to optimize motor efficiency. Doing flight tests take really long time. \$\endgroup\$
    – lotsof one
    Dec 17, 2022 at 12:30
  • \$\begingroup\$ Is there a good way to measure switching losses of the ESC? \$\endgroup\$
    – lotsof one
    Dec 17, 2022 at 12:46
  • \$\begingroup\$ I found the datasheet of the MOSFET on my ESC: alltransistors.com/adv/pdfdatasheet_aosemi/aon6504.pdf I found that the sum of Turn-On Rise Time and Turn-Off Fall Time is 15.5ns. Suppose the PWM frequency is 1MHz, that's 1000ns every cycle. Can I assume the switching losses is less than 1.55%? \$\endgroup\$
    – lotsof one
    Dec 17, 2022 at 13:26
  • \$\begingroup\$ No, you can't assume switching losses from that info. However ESCs generally run at a much lower PWM frequency so switching losses are small. Best way to ensure low loss is to buy a high quality ESC with synchronous rectification, eg. Kontronik kontronik.com/en/products/speedcontroller/speedcontroller1/… \$\endgroup\$ Dec 17, 2022 at 20:56
  • \$\begingroup\$ I guess it's because MOSFETs can have much lower on-resistance or voltage drop than diodes that makes synchronous rectification efficient? \$\endgroup\$
    – lotsof one
    Dec 18, 2022 at 2:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.