1. How long to charge the
1. How long to charge the cap, under what light conditions?
2. How long does the motor run? While on a bread board versus in a bot on a surface?
Well, time to charge and discharge depends on software parameters
I can choose to wait to refuel a lot of energy then discharge down to minimum energy obtaining long but infrequent motor activity or i can chose to burst motors frequently…
But if the question is: “is it efficent to convert light in motion?” The answer is Yes!
Clearly, I have made my tests with motor mounted under a piece of plastic of the right weight to simulate realistic bot motion.
I can report only a very rough example:
Under a 60W light bulb placed at 30cm it charge for 3 seconds then produce a 50mS motor burst on both motors . This is enought to make the bot move for 1cm (all numer are only ad approximation)
anyway i have got 2 software parameter which regulate the “solarengine”
- the voltage threshold (3.1V in the example)
- the quantum time of activity (50mS in the example)
aka: when capacitors are at 3.1v i am sure that I can operate at full power for 50mS remaining safe from brown-out
this level of control is the reason why I prefer a zener & adc reading solution istead of a fixed ic trigger.
3. What happens after the motor runs?
It go in lo power mode… it start a lazy poling loop untill voltage reach the threshold again
4. What are the advantages over a “Photopopper” circuit, that uses much cheaper components.
I have played a lot with beams and it is not so easy to obtain efficiency in beam robots…
when you have a bunch of npn and pnp junction tied together and you try to undestrand why your bot is not efficent or have some strange behaviour in a particular light conditions… you may go crazy…
I thought beams are cool too, but to discover a micro wich cost 5$ and consume uA is extra-cool 
