Thanks Steve! Once I drew a timing diagram it all made sense to me - I would have never thought of that. I guess that's why my my EE friends say microprocessors make you lazy.
You are welcome. After I posted, I noticed Steve k had already explained it.
Even though I was a hardware eng (and therefore cannot be trusted to write object-oriented code
, I have done two microcontroller-controlled dynamo light circuits.
I went the software route because it cuts the number of components down - generally to one chip and a handful of other components. MCU timing does a great job of removing the need for monostables - BTW - that circuit above is a common one, it has the disadvantage (in some circumstances) that it also gives the high-speed output state at 0mph.
That said, if you go entirely MCU, you do have to write lines of code to cope with situations that a single capacitor would cure in the hardware realm!
I am very happy to be as helpful to you as I can.
For example, I did a nice one-transistor interface circuit to get safe timing pulses into the MCU from the nasty world of dynamo voltages, and a few other bits and pieces.
Sadly, I cannot post diagrams - I don't know how and don't want to spend time learning, but I can always email them to someone who has learned.
I could design hardware for you if you like - as could Steve K, who is probably a better bet for selecting inductors as he has more switching power experience than me.
The other Steve