1
\$\begingroup\$

I am designing an ESC for BLDC motors for learning purposes. For the dimensioning of the MOSFETs, I want to calculate the RMS phase current so I can calculate the switching losses of each FET. The issue is, I don't really understand how to determine the phase currents. Let's take the following assumptions:

  • Battery supply 36V
  • Motor electrical power (ignoring mechanical power and efficiency) 500W

The above assumptions give a current of about 13.9A, which in my understanding is called the rated current of the motor. How is then the RMS and peak current calculated per phase? Does the control type matter? What is the current difference between 6 step commutation and FOC?

I will definitely measure this, but before then I would be interested to understand the math behind it.

\$\endgroup\$
0

3 Answers 3

3
\$\begingroup\$

Your controller will be attempting to generate a near-sinusoidal waveform to create a field to match that generated by the rotor as it rotated inside the stator. The motor has inductance, which will define how quickly the current can change when voltage is applied.

How you simulate the sinusoidal waveform in your controller depends on many issues: your commutation rate as compared to your maximum motor speed being of primary importance. The relationship between any three-phase peak and rms values hold for your motor. However, you are using PWM to generate your sinusoid, so your instantaneous peak current must have the duty cycle applied to it to give you your answer.

In the first sample below, one of my motor designs is providing an RMS phase current of 5 amperes while not running full speed. The supply current is less than the RMS phase current of a single phase, but the PWM means that the RMS phase voltage is only 13.9 volts.

The good news: the math is straightforward for brushless permanent magnet motors: torque is directly proportional to current and speed is directly proportional to voltage.

So design your controller to handle peak currents 25-50% higher than the motor rating, and provide a voltage higher than the motor voltage rating to give you some headroom and keep you from stalling. You need this because when you command a speed increase under a full torque load, your motor must handle the additional torque from inertia, and if you are running at full speed and put on a load, the controller will want to increase voltage temporarily to get back up to speed. This headroom will keep your PID loop working.

You can figure out the current from the motor's torque constant, usually expressed in torque/current (Nm/A). There is also a back emf constant, expressed in a number of different ways: Voltage/Speed (sometimes V/kRPM). If you play with radians, volts and amperes you can prove that these two constants are actually the same variable, with units swapped around, but often both are presented.

Any permanent magnet brushless motor can be run either six-step or sinusoidal. The second example shows the same motor running six-step. Motors are wound differently for different applications. Motors labelled "PMSM motors" are generally designed to be run sinusoidal; the brushless motor shown below was wound for either operation. The price for running a sinusoidal motor in six-step mode is torque and current ripple.

Good luck! sinusoidal

six-step

\$\endgroup\$
2
  • \$\begingroup\$ Thanks a lot for the detailed answer. The plots are very insightful too! Basically, as a rule of thumb, the MOSFETs should have a 25-50% headroom for the maximum current. However, here I'm still interested in the math, especially for the sinusoidal control. In 6 step, it's pretty straightforward. The current is regulated by a fixed PWM and the value can be calculated using the phase resistance and inductance. \$\endgroup\$ Commented Mar 16 at 8:22
  • \$\begingroup\$ Be careful using MOSFET ratings - you must calculate your switching losses in the MOSFET, apply the theta(JC) of theta(JA) to determine the junction temperature. If you are a math person, you will have fun with googling articles on FOC motor control, in which the math is explained. \$\endgroup\$ Commented Mar 17 at 15:15
0
\$\begingroup\$

What control method are you using?, assuming your implementing 6step commutation( trapezoidal control method) only 2 phases conduct at any instance while the third phase is floating, the rms phase current can be calculated by rms_phase = 13.9/sqrt(3) = 8.02A, peak current can also be calculated using I_peak =2*13.9/sqrt(3) =16.04A. Assuming field oriented control i_phase = 13.9/2 = 6.95A and I_peak = sqrt(2)*6.95 =9.38A

\$\endgroup\$
0
\$\begingroup\$

Yes control method matters, assuming your implementing 6step commutation( aka trapezoidal control method) only 2 phases conduct at any instance while the third phase is floating, the rms phase current can be calculated by rms_phase = 13.9/sqrt(3) = 8.02A, peak current can also be calculated using I_peak =2*13.9/sqrt(3) =16.04A. Assuming field oriented control i_phase = 13.9/2 = 6.95A and I_peak = sqrt(2)*6.95 =9.38A. You can see that 6 step commutation has highr phase currents producing higher torque

\$\endgroup\$

Your Answer

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

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.