RE: Blog 5 - Magic Light Box: Building The Prototype

jc2048 — Mon, 05 Feb 2024 12:09:26 GMT

In the schematic in your blog, the load is an LED driver (buck switcher) that is controlling a string of LEDs. Normally, that switcher would be trying to regulate the current through the LEDs. Your control loop is finding supply voltages for it where it doesn't function properly but has the switch on enabling current flow through the LEDs (suggests it's probably a self-oscillating hysteretic part). Whilst that is working for you, it is messy and it confused me as to what you were trying to show me. Personally, I'd have started with a resistive load (resistance similar to the dynamic resistance of the LEDs at the operating current) and then moved to a simple string of LEDs when I knew that the control loop was functioning and controlling the current properly.

If anyone tries to replicate what you've done, there's also a "gotcha" built in. Consider what happens if your control loop is set to a higher current than the buck converter works to. Your loop will raise the voltage on C13 in the expectation that the current will increase, but it won't because the buck converter will reduce its duty cycle to compensate. Your loop will again up the voltage, and so on. With the LED controller functioning, there is no 'right' voltage for C13. So, make sure you have some mechanism to limit the voltage before something gets damaged.

Ideally, I'd have liked to have seen more on the loop control and how it responded dynamically, but I can see that you ran out of time after an extensive and demanding design challenge, so not to worry.