MOSFET Overheating at High RPM in a Three-Phase BLDC Motor Driver

[Third Harmonics]

Hi everyone,

I am working on a three-phase BLDC motor control application using the following components:

• MOSFET: STL90N6F7
• Gate Driver IC: DRV8334
• Microcontroller: STM32G431RBT3
• DC Bus Voltage: 24 VDC
• Motor Speed: 60,000–70,000 RPM
• Measured Peak load Current: Approximately 5 A

The motor is operating normally; however, when running at higher speeds (approximately 60,000–70,000 RPM), the MOSFET package surface temperature rises to around 60°C.

Considering that the DC bus voltage is 24 V and the measured peak input current is approximately 5 A, I would like to understand why the MOSFETs are experiencing this level of heating at high RPM.

Could the heating be caused by switching losses, gate-drive voltage, dead-time settings, PWM frequency, reverse-recovery losses, PCB layout, or other factors associated with high-speed motor operation?

I would appreciate any suggestions on how to identify the root cause and reduce the MOSFET temperature. If anyone has encountered a similar issue with the DRV8334 or STL90N6F7, please share your experience.

 

[rich.g.williams]

MOSFETS can heat it for three basic reasons, On resistance, Avalanche energy (inductive loads generate reverse peaks), Switching too slow.

Would need to see your schematic and know about your driver and MOSFET type.

 

[Third Harmonics]

Additional Details: MOSFET Driver Schematic and Component Datasheets

I have attached the schematic of the MOSFET driving circuit implemented using the DRV8334, along with the datasheets for the DRV8334 gate driver IC and the STL90N6F7 MOSFETs.

The PWM switching frequency is 20 kHz.

Additionally, inductors are connected in series between the three motor phase outputs (MOT A, MOT B, and MOT C) and the output connector J13. I have also attached the datasheet for these inductors.

Please refer to the attached schematic and datasheets when analysing the issue.

Based on these circuit details and component specifications, could you help identify the possible causes of MOSFET overheating at high motor speeds (60,000–70,000 RPM), even though the measured peak input current is approximately 5 A at a 24 VDC supply?

I would appreciate any suggestions for further measurements, potential design issues, or modifications that could help identify the root cause and reduce the MOSFET temperature.

[rich.g.williams]

When Q6 turns off MOTC will rise very quickly (inductive flyback) current wants to keep flowing so it will flow through body diode of Q3 back into the +24V line, or even cause some avalanche in Q6. 

Same will happen when Q3 turns off.  And same for other phases.

Schottky diodes in parallel with body diodes might help. Also C18 may need to be two capacitors, a good low ESR (film maybe) 100n or 1u and a second more like 100u. Same for other phases.

Layout to minimise ground loop type currents is important.

It's hard to see the inductive flyback on the scope because it's clamped.

Another way is putting Schottky diodes in series between say Drain of Q6 to MOTC and Source of Q3 to MOTC.

Also switching waveforms, It's not good enough to turn Q6 on then as it goes off to immediately turn Q3 on, you will get cross commutation. I'm sure you are already doing that as you mentioned dead time.

The inductors you mention make the problem worse they act to prevent inrush current as say Q6 turns on but also store energy that returns when Q6 turns off etc.

Some ideas for you to help solve the problem.