Drone Computer Vision System

Introduction

For my work for a research project at the University of British Columbia, I designed a carrier board for a CM5 computer vision system. The board is designed to be compact and lightweight, while still providing the necessary power and connectivity for the CM5.

Design

The design supports a Raspberry Pi CM5, magless ethernet, 2 UART ports, a 6S LiPo battery input, and USB-C connectivity. As well it has a slot for a PCI-e card for a Halo AI accelerator.

Impedance Matching and Signal Integrity

This was the first board I designed that required impedance matching and signal integrity analysis. I used the new impedance and signal integrity tools in KiCAD 10 to analyze the board and ensure that the signals were within the required specifications. I found out quickly that to meet the impedance requirements my traces would be much wider than possible with the current default JLCPCB stackup. I switched to using a custom impedance controlled stackup that I designed to meet the requirements. This was a great learning experience and I learned a lot about impedance matching and signal integrity.

Conclusion

This design allowed me to work with digital signals further improving my range of skills.