Spacewrench over at Dorkbotpdx published a new build, a Power Playground project:
It’s a PMOS/NMOS H-Bridge with FETs that can handle 3 amps or so, plus a SPI current sensor, some switches & a rotary encoder (not stuffed yet), and a 7-segment display, all controlled by a Teensy-3.1 running FreeRTOS.
I made this because I’m always running into battery, power, inductor and transformer issues I don’t have any experience with. The idea is to use the H-bridge configuration and current sensors to experiment with moderate-current PWM, motor control, power-line synchronization, battery charging and discharging, etc.
H-bridges are great fun until (not if) for whatever reason the high and low side manages to turn on at the same time, however briefly…
it’s nice and cute, but i don’t recommend this hbridge architecture for an experimental project.
– using VBATT for mos gates is not a good idea : it blows up as soon as a transient exceeds 20V on VBATT
– using a Nmos only bridge is better, with a regulated gate supply (bootstrapped for the upper ones)
– using dead time control integrated in the mos drivers is good practice. Experimental software should not be relied on for that
– implementing an overcurrent cutoff in HW is a good idea for avoiding the smell of burnd mosfets, especially when dealing with experimental setups.
– don’t forget an overvoltage clamp that can handle a high power. It’s crucial when your motor goes into a generator quadrant, even more if your hbridge cannot handle an overvoltage.