Brushless DC (BLDC) Motor Design: A New Starting Point
                    
                    
                    Introduction
 
Brushless DC motor (BLDC) design  is complicated. It doesn’t help that customers begin their electronics  selection (old starting point) navigating through a maze of enormous; MOSFET, IGBT and Gate driver product portfolios.
 
onsemi is here to help, delivering a “new 1st  order approximation starting point”, matching gate drive to switch  (N-FET, or IGBT), much closer to the customer’s final decision, and  leapfrogging the “old starting point” of seemingly endless product  portfolio grazing. This includes 5 comprehensive tables incorporating  motor voltages of; 12V, 24V, 48V, 60V, 120V, 200V, 300V, 400V and 650V,  up to 6kW.
 
 
 
Figure 1 
 
Brushless DC Motors (BLDC)
 
Brushless DC (BLDC) motors  offer many advantages over brushed permanent magnet DC (PMDC) motors,  specifically higher reliability, little to no maintenance, lower  electrical and acoustic noise, better thermal performance, higher speed  range, and higher power density. A typical BLDC motor uses permanent  magnets on the rotor, and three armature windings (U, V, W) on the  stator. A microcontroller (MCU) implements one of a variety of control  and modulation schemes (trapezoidal, sinusoidal, FOC with SVM, DTC,  etc.) to strategically energize the motor windings.  This creates  electromagnetic fields, causing an interactive force between the rotor  magnets and the stator windings.  When done properly, this interaction  allows for precise control of the motor’s speed, torque, or power in the  desired direction.
 
Figure 2 illustrates a typical block diagram of a 3-phase BLDC motor.  The MCU executes the control and modulation scheme firmware, where it  commands its PWM peripherals to output six coordinated duty cycles to  three half-bridge gate drivers.  These three drivers act as power  steering for the six power MOSFETs in the output bridge, energizing the  low-side (LS) and high-side (HS) U, V, and W MOSFETs.  These are  typically N-Channel MOSFETS rated for 1.5~2.0x the motor voltage, up to ~300V. Above 300V, N-Channel MOSFETs are typically replaced by IGBTs for their higher power performance.  
 
The MCU can measure the current through each winding via the FAN4852, CMOS Op-Amp  (9MHz typical BW), and optionally, can assess the rotor’s angular  position with Hall Effect sensor feedback. Alternatively, a sensor-less  architecture can be implemented with more processing overhead. The RSL10 BLE can be used for asset tracking, Firmware over the Air Updates (FOTA), feature selection/tuning, and telemetry data gathering.
 
 
 

 
Figure 2
 
 
 
BLDC Table #1: 12V & 24V (N-FETs) up to 1.1kW
 
Table 1 below lists a “new 1st order approximation starting point” for matching BLDC gate drive to N-Channel MOSFETs for 12V from 93W to 372W, and 24V from 186W to 1.1kW.
 
 
 
Table 1
 
 
 
BLDC Table #2: 48V & 60V (N-FETs) up to 1.5kW
 
Table 2 below lists a “new 1st order approximation starting point” for matching BLDC gate drive to N-Channel MOSFETs for 48V from 186W to 1.5kW, and 60V from 186W to 1.5kW.
 
 
 
 
Table 2
 
 
 
BLDC Table #3: 48V & 60V (N-FETs) up to 3kW
 
Table 3 below lists a “new 1st order approximation starting point” for matching BLDC gate drive to N-Channel MOSFETs for 120V from 186W to 1.8kW, and 200V from 186W to 3kW.
 

 
Table 3
 
 
 
BLDC Table #4: 300V & 400V (IGBT’s) up to 6kW
 
Table 4 below lists a “new 1st order approximation starting point” for matching BLDC gate drive to IGBTs for 300V from 372W to 4.5kW, and 400V from 372W to 6kW.
 
 
 
 Table 4
 
 
 
BLDC Table #5: 300V, 400V & 650V (IPMs) up to 6kW
 
Table 5 below lists a “new 1st order approximation starting point” for Integrated Power Modules (IPMs), where the gate drive and IGBTs are integrated into an easy to use module, for 300V from 372W to 4.5kW, 400V from 372W to 6kW, and 650V from 372W to 6kW.
 
 
 
 Table 5
 
 
 
onsemi offers a great online tool for building BLDC with IPMs (Integrated Power Modules).  The user enters 15 operating conditions and the tool produces multiple  detailed Analysis Tables plus twelve graphs capturing critical thermal  and power performance (Figure 3).
 
 
 
 Figure 3
 
 
 
BLDC Tables #1 - #5
 
BLDC  is complicated with hundreds of decisions to be made start to finish.  For example, If you took 3 different customers; a, b and c (Figure 1),  beginning at the same “starting point” (24V, 1 1/4hp motor), by the time  all 3 customers navigated their respective decision trees, their final  designs will be completely different. This is because each customer has  their own threshold for cost, efficiency, power density, form factor,  maintenance, lifetime etc. So, it is impossible to build gate drive to  switch (MOSFET/IGBT)  matching tables, and have it be correct for every customer. If we  tried, we might be correct for one customer, and incorrect for another  999 customers. However, we can make some reasonable assumptions, based  on intelligent engineering considerations, and produce a “1st  order approximation” that lies somewhere between handing the customer  Switch and gate drivers portfolios (old starting point: you are on your  own), and the customer’s final decision.
 
 
 
1st Order Approximation Engineering Considerations
 
1)     Cost: We strived to filter the lowest cost while meeting the following considerations.
 
2)     Topologies: The Trapezoidal (a.k.a. 6-Step  Control) Commutation was selected because it is relatively simple to  control and yields high efficiency and high Peak Torque. Since only two  power switches are on at any one time, each switch experiences a 33% “on  time” duty cycle.
 
3)     PWM Duty Cycle: PWM Frequency of 15 kHz. This is typical for most BLDC under 6kW. 
 
4)     Gate Driver: Junction Isolated Gate drivers. These tables do not include Galvanic Isolation.
 
5)     Temperature: Ambient temp of 850C.
 
6)     Gate Drive Calculation: The rated Gate drive is calculated by dividing the QG(TOT) (nC) by the ON/Off time (ns). We chose 50ns ON/Off for N-FETs, and 200ns for IGBTs.
 
7)     N-FET Junction Temperature: (TJ) for surface mount packaging (no heat sink) is calculated by TJ = PDISS x RθJA + Ambient, leaving a minimum of 250C headroom below the maximum rated TJ.
 
    - Where:
 
          i.      RθJA = Junction to Ambient thermal resistance
 
8)     IGBT Junction Temperature: IGBT Junction Temperature (TJ) for through hole packaging with a heat sink is calculated by TJ = PDISS x (RθJC + RθCS + RθSA)+ Ambient, leaving a minimum of 500C headroom below the maximum rated TJ.
 
    - Where:
 
          i.      RθJC = Junction to Case thermal resistance
 
          ii.      RθCS = Case to Heat Sink thermal resistance
 
          iii.      RθSA = Heat Sink to Ambient thermal resistance
 
9)     N-FETs Power Dissipation: IPHASE2 (A) x RDSON (ohms).
 
10)   IGBT Power Dissipation: Switching Loss + Conduction Loss + Diode Loss
 
    - Where:
 
          i.      Switching Loss  = Ets (J) x PWM Frequency (Hz)
 
          ii.      Conduction Loss = IPHASE (A) x VCE(SAT) (V)
 
          iii.      Diode Loss = (switching loss + conduction loss) x 0.25
 
11) Switch Voltage Rating: N-FET V(BR)DSS and IGBT VCES = 2-3x Motor Voltage
 
12) Switch Current Rating: N-FET ID and IGBT IC = 3 x IPHASE.
 
13) Motor Phase Current:  IPHASE = 1.23 x POUT /  VBUS
 
    - Where:
 
           i.      IPHASE = Motor Phase Current, Amps
 
           ii.      POUT = Electrical Power Output of Inverter to Motor
 
           iii.      PF = Motor Power Factor, 0.0 – 1.0, 1.0 is ideal (we assume 0.85)
 
           iv.      VBUS = Motor Bus Voltage, VDC, or 24V
 
           v.      MI = Modulation Index, 0.0 – 1.0, typical is 0.9 (we assume 0.9)