← Darsh Patel

60 V BLDC motor controller

Personal · KiCad · routed, not fabricated AI-driven

On this page

I routed a six-layer, 55 mm BLDC motor-controller board in KiCad 10. It targets a 60 V maximum bus, 20 A continuous and 80 A peak, using the MAB Robotics MD80 mounting, connectors and encoder position. It has not been fabricated, and firmware has not started.

A TI DRV8353 gate driver runs six Infineon 100 V MOSFETs in 5 × 6 mm packages, with two 0.5 mΩ shunts on the top side between the FETs. An STM32G474 runs field-oriented control and talks CAN-FD, and an AS5047P reads the rotor angle on-axis. A 100 V buck, a 5 V buck and an LDO make the logic rails, and a 5.0SMDJ60A TVS holds a 60 V hot-plug below the parts' 100 V rating.

The outline, holes and connectors match MAB's STEP model to 0.000 mm, and the board has no unconnected nets. FastHenry estimates a 3 nH commutation loop; LTspice double-pulse runs predict a worst switching peak of 85.7 V. Current, loss and thermal limits need hardware measurements.

design files ↗

Top side in KiCad 3D: the six MOSFETs and two shunts along the bottom edge, the DRV8353 gate driver at centre left and the STM32G474 at centre right.
Top side in KiCad 3D: the six MOSFETs and two shunts along the bottom edge, the DRV8353 gate driver at centre left and the STM32G474 at centre right.
Bottom side: the row of 1206 DC-link capacitors across the middle, with the power supplies, CAN and RS-422 parts around them.
Bottom side: the row of 1206 DC-link capacitors across the middle, with the power supplies, CAN and RS-422 parts around them.
All six copper layers, top to bottom.
All six copper layers, top to bottom.
LTspice double-pulse runs at 60 V and 80 A with the 3 nH loop from FastHenry. At the chosen gate-drive strength the worst overshoot is 85.7 V, against the MOSFETs' 100 V rating.
LTspice double-pulse runs at 60 V and 80 A with the 3 nH loop from FastHenry. At the chosen gate-drive strength the worst overshoot is 85.7 V, against the MOSFETs' 100 V rating.