Just came across this novel time based and simple (non PID) way to control angle of brushed DC motor:
Thanks.
He writes that his system works load-independent, but I guess that’s only true for linear loads. (start inertia = stop inertia)
A robot arm that lifts a weight, will almost never have linear load conditions.
Then there are varying parameters, like stiction and friction he can’t control.
I’d name it “ballpark positioning” at best…
I dunno, sounds a bit ad-hoc to me… my guess is that after many iterations on this code the author will arrive back at PID control after all 
But let’s see, from the roadmap it sounds like they’re just starting and much is yet to be implemented.
1 Like
@o_lampe What you call ballpark is defined by him as Educated Guess.
Well, it’s a trained guess, since he uses a lookup table. If he’d extrapolate a function and implement real parameters it would be educated, IMHO.
I’ve made my own step/dir interface with a geared DC motor back in 2019. It kinda worked but the ATtiny44 was overwhelmed by interrupts from the encoder.
Today I’d use a MCU with hardware encoder.
1 Like