Looks like loneoceans has designed an 8A boost on a 17mm board but I’m not sure if anyone’s building it. I currently have no real use for this but it’s an interesting challenge. Low priority at present.
Thermal regulation would be implemented, as currently planned, by a SC70-5 temp sensor mounted on the MCPCB close to the LED and relaying directly to an ADC line on the microcontroller. The micro can use its internal temp indicator to trigger a shutdown based on driver PCB temperature, but current regulation would be done by limiting the max temp of the LED board, ramping down to maintain a maximum temperature (value TBD).
One thing to consider with the control package will be current sampling. I can either measure it directly from the low-side resistor and then use a resistor divider to scale my PWM output accordingly, or use a current-sense amplifier (ie INA199A1) to measure from a low- or high-side resistor. The former approach reduces parts and complexity but also reduces noise immunity and increases measurement error. The latter approach should result in a more precise setpoint at the cost of more parts, both passive and active, but should also allow for a smaller current-sense resistor and therefore a better efficiency. Amplifier sampling with kelvin connections will minimize ground-current errors also, and the higher voltages in use will reduce input offset voltage error. I assume the increase in precision (and marginal efficiency gain) is worth the extra effort?