Skip to content

H-Bridge MOSFET Selection

Calculate H-bridge MOSFET requirements including peak current, conduction losses, and minimum current rating for DC motor drivers.

Two FETs in I²R conduction · one hard-switched, V·I·t_sw·f at 50 ns · 1.5× peak marginAPI slug h-bridge-selection

Sign in to saveSave

Operating point

Worst case evaluated across the input rangeExample
×

1 to 15 ×

100 Hz or more

The share of each period the supply drives the motor: the supply delivers V·I·D (sign-magnitude PWM, TI SLVA504A).
%

0 to 100 %

Loss modelRDS 10 mΩtr+tf 100 nstd 100 nsVF 1 VQg 30 nCVg 10 V

EfficiencyPASS
99.3%
124 mW lost of 17.88 W out
Min. MOSFET Current Rating
18A
18 A at Irated 2 A
Peak Current
12A
12 A at Irated 2 A
Total Bridge Loss (est.)
104mW
104 mW at Vm 12 V, Irated 2 A
Efficiency, Worst Case
99.3%
at Vm 12 V, Irated 2 A
MOSFET Drain-Source Rating
20V
at Vm 12 V

Parts and stress

PartValueMust handle
Q1–Q4Bridge10mΩVDS ≥ 20 V at Vm 12 VID ≥ 18 A at Irated 2 AQg ≤ 30 nC

Where the 124 mW goes

Conduction Loss (two FETs)
80 mW
Switching Loss (per FET)
24 mW
Gate-Drive Loss
12 mW
Dead-Time Loss
8 mW

Biggest levers: lower-R_DS(on) FETs and faster switching edges.

Waveforms at the nominal operating point

Ideal switching, 2.5 periods
VAB
ton 37.5 µs0246810120 s20 µs40 µs60 µs80 µs100 µs120 µsV
IM
1.81.91.91.92.02.02.12.12.20 s20 µs40 µs60 µs80 µs100 µs120 µsA
Conduction Loss (per FET)Pcond
40mW

Formula

TI SLVA504A, Calculating Power Dissipation for a H-Bridge or Half Bridge Driver (A. Gopalan, 2012, revised 2021), Table 1-1 and eqs. 2-3; International Rectifier AN-1070 (J. Cerezo), P_GATE
Ipeak=Irated×k,Pcond=I2×RDS(on),Ptotal=2⋅Pcond+PswI_{peak} = I_{rated} × k,\quad P_{cond} = I²× R_{DS}(on),\quad P_{total} = 2·P_{cond} + P_{sw}
SymbolQuantityYour value
kInrush multiplier (×)6 ×
RDSMOSFET on-resistance (Ω)10 mΩ

Assumptions

5 items
  1. 01An H-bridge under sign-magnitude PWM with synchronous recirculation (TI SLVA504A, Table 1-1): two FETs carry the motor current at any instant, 2·I²·R_DS(on) at any duty cycle, and one is hard-switched each period, ½·V·I·(t_r + t_f)·f_pwm (eq. 2).
  2. 02The peak current is the inrush multiplier times the motor current and the minimum FET current rating 1.5 times that peak; the drain-source rating is the voltage sold at or above 1.5 times the highest supply, with no further allowance for inductive spikes.
  3. 03The supply and the motor current are ranges, evaluated at every pair of their corners: the peak current and the current rating at the highest current, the drain-source rating at the highest supply, and the bridge loss and the efficiency each at the corner where they are worst.
  4. 04The loss model, at the nominal supply and current, adds the body diode through both dead times, V_F·I·2·t_d·f (eq. 3), and the gate drive 2·Q_g·V_g·f of the two FETs that switch, dissipated in the driver, against the supply’s V·I·D; the recirculating FET’s own edges (eq. 4) and reverse recovery are left out, as SLVA504A leaves them.
  5. 05No motor inductance is modelled, so the motor current is its average, flat and without ripple, and R_DS(on) is taken as entered, without its rise with temperature.

Worked example

Automotive 9–16 V motor, 1–5 A

An automotive motor draws 3 A running (1 to 5 A) from a supply of 13.5 V nominal, 9 to 16 V, and 6 times its current at start. The bridge’s FETs have 8 mΩ and 200 ns of edges and switch at 20 kHz and 75 % duty. Each edge has 200 ns of dead time, when 0.8 V body diodes carry the current, and the FETs’ 40 nC of gate charge is driven at 10 V. What must the FETs be rated for, and where are the bridge’s loss and efficiency worst?

StepExpressionResult
Peak current, at the highest motor current6 × 5 A30 A
Minimum FET current rating1.5 × 30 A45 A
Derated drain-source voltage, at the highest supply1.5 × 16 V24 V
Drain-source rating, the next standard value≥ 24 V25 V
Bridge loss at 13.5 V and 3 A2 × 3² × 8 mΩ + ½ × 13.5 V × 3 A × 200 ns × 20 kHz225 mW
Bridge loss, worst at 16 V and 5 A2 × 5² × 8 mΩ + ½ × 16 V × 5 A × 200 ns × 20 kHz560 mW
Conduction and switching loss at 9 V and 5 A2 × 5² × 8 mΩ + ½ × 9 V × 5 A × 200 ns × 20 kHz490 mW
Dead-time loss at 5 A0.8 V × 5 A × 2 × 200 ns × 20 kHz32 mW
Gate-drive loss2 × 40 nC × 10 V × 20 kHz16 mW
Power from the supply at 9 V and 5 A9 V × 5 A × 75 %33.75 W
Efficiency, worst at 9 V and 5 A1 − (0.49 + 0.032 + 0.016) / 33.7598.4 %

How it works

This calculator determines MOSFET ratings and gate driver requirements for H-bridge motor control circuits. Power electronics engineers, robotics designers, and EV builders use it to select components that handle motor current with adequate voltage and thermal margins. Proper H-bridge design prevents catastrophic shoot-through failures and ensures reliable bidirectional motor control.

Per power electronics fundamentals (Mohan, 'Power Electronics', 3rd ed.), an H-bridge uses four switches to enable forward, reverse, and braking by controlling current direction through the motor. Key selection parameters, as this calculator takes them: V_DS rating at or above 1.5× the highest supply voltage (many designers allow 2× for inductive voltage spikes), I_D continuous ≥ 1.5× the peak current, the inrush multiplier times the motor's rated current, and R_DS(on) low enough to limit conduction losses to acceptable thermal budget.

MOSFET voltage derating is critical: per JEDEC reliability guidelines, voltage stress should be <80% of V_DS rating for long-term reliability. A 24V motor system with inductive spikes reaching 1.5× supply requires MOSFETs rated ≥ 60V (24×1.5÷0.80 = 45V minimum, use next standard rating), stricter than this calculator's 1.5 × 24V = 36V, a 40V part. Conduction losses scale with R_DS(on): modern power MOSFETs achieve 1-10 mΩ at 30-60V ratings, enabling 10A continuous with only 0.1-1W loss per FET. Total H-bridge losses at 10A: 0.2-2W for quality MOSFETs versus 60-100W for integrated drivers with 300-500 mΩ internal switches.

Practical tips

  • Per Texas Instruments application notes, use integrated H-bridge driver ICs (DRV8876, DRV8874) for motors <5A where convenience outweighs efficiency; discrete MOSFETs with <10 mΩ R_DS(on) for >5A where 5-10% efficiency gain matters
  • Place 100 nF ceramic capacitors within 10mm of each MOSFET drain-source per EMC guidelines to suppress 10-100 MHz switching transients; add 100-470 µF electrolytic across DC bus for inrush energy
  • Implement dead-time of 50-200 ns between high-side off and low-side on per MOSFET turn-off specifications—IR2104 and similar drivers include automatic dead-time insertion

Common mistakes

  • Selecting MOSFETs at exactly supply voltage: Per JEDEC, inductive kickback reaches 1.5-2× supply voltage; a 24V system needs at least a 40V part (1.5 × 24V = 36V, this calculator's rule) and a 60V part where the kickback reaches 2×
  • Omitting freewheeling diodes on discrete builds: MOSFET body diodes conduct during dead-time but have slow reverse recovery (50-200 ns); add external Schottky diodes for currents >10A to reduce switching losses by 20-40%
  • Using single gate resistor for all four MOSFETs: Each gate needs individual resistor (10-47Ω typical) to prevent parasitic oscillation and allow independent tuning per Infineon gate driver guidelines

Frequently asked questions

3 questions
What is shoot-through and how do I prevent it?

Per power electronics safety guidelines: Shoot-through occurs when both high-side and low-side MOSFETs in one leg conduct simultaneously, creating a near-short from supply to ground. Current spikes to hundreds of amperes in nanoseconds, destroying MOSFETs. Prevention requires dead-time (50-200 ns) between turning one FET off and the other on. Gate driver ICs (IR2104, DRV8876) implement this automatically. For discrete designs, add RC delay or use a dedicated dead-time controller IC.

When should I use an integrated H-bridge IC vs. discrete MOSFETs?

Per cost/performance analysis: Integrated ICs (DRV8833, TB6612FNG, L298N) suit currents <3-5A where USD 2-5 component cost and simple design outweigh 3-10% efficiency loss from higher R_DS(on). Discrete MOSFETs suit currents >5A where 2-20 mΩ R_DS(on) (vs. 200-500 mΩ integrated) improves efficiency from 85% to 95%+, saving significant power in battery applications. Discrete designs require separate gate drivers, adding USD 3-5 and PCB complexity.

Why does my H-bridge heat up even when the motor is not running?

Three causes per troubleshooting guides: (1) PWM switching loses energy at every edge whatever the duty cycle—use 0% or 100% duty for idle, where nothing switches; (2) Body diode conduction during dead-time—each PWM cycle conducts through slow body diode for 100-500 ns, dissipating power independent of motor current; (3) Gate driver quiescent current (5-20 mA) and bootstrap capacitor charging losses are fixed overhead. Reduce heating by lowering PWM frequency during low-speed operation and using brake mode (both low-side on) instead of coast mode at zero speed.

Shop Components

As an Amazon Associate we earn from qualifying purchases.

  • NEMA 17 Stepper MotorNEMA 17 bipolar stepper motors for precision motion controlAmazon →
  • Stepper Motor Driver (A4988)A4988 stepper driver modules for microstepping controlAmazon →
  • DC Motor with Encoder12 V DC motors with encoders for closed-loop drive applicationsAmazon →