Could it be that the polarity of PWM changed with 2.2 firmware? We are seeing opposite behaviour of PWM in the newest firmware. For example, a duty cycle of 1 used to mean full on and duty cycle of 0 was fully off. This now seems to be reversed.
I see there was a PWM glitch fixed in the changelog, does this have something to do with it?
I use generic 5V leds with the + connected to 5V and the ground is switched on/off by a mosfet with it’s gate connected to the PWM pin.
There is 2 PWM helper classes in there, one works as expected, the other works weird. If you run the code I provided, you will see two leds with a dimming effect that work opposite of each other. When one is fully off, the other is fully on. Both of the pwm led helpers are fed the same value.
Please try around with the enable/disable combined with the brightness features in this code. You will see that both classes operate very differently.