Comment Re:The root course is bad testing (Score 1) 25
I don't know if the issue was explained to me but I suspect that the LED with voltage supplied it would light up full brightness and then go dark immediately once voltage was removed, which is unlike an incandescent indicator or even many other LED types. Our attempts to modify the PWM code to dim it only resulted in blinking brightly.
I can't remember ever having come across an LED that wouldn't respond well to PWM that was at a fast enough pulse rate and sufficient resolution in the duty cycle, but I 100% agree that real hardware doesn't always respond like the simulations did, particularly when it's integrated into a larger system. A few months back at work, we shipped an avionics update, and we're kind of amazed that it's worked in several types/blocks of the real aircraft exactly as it did in the simulations - the customer hasn't reported a single problem in five months of ground and flight testing, and is releasing it organization-wide very soon. Obviously we screwed something up.