The innovations in this heliostat design are intended to bring the manufacturing difficulty and cost low enough that direct solar heating can be applied pragmatically by individuals and small scale industry.
The focus of innovation is on the following:
Reducing the number of trace elements required to function. I settled on using switched reluctance motoring to remove the need for rare earth magnets in other motor designs. I plan to use relays for motor stepping rather than mosfets.
Minimizing the mechanical complexity. My design has only 1 point of slowly moving contact to wear, with no belts or gears to replace.
Using inexpensive production equipment to increase replicability.
How does this help solve a problem that matters to everyone? The most cost effective way to harness solar power is to directly provide heating, lighting, cooking power, or metal melting power. This design will make those applications more accessible to the people who will apply them the best.
The alignment issues are too much of a headache with the radial stepping design, and the precision is quite poor with that configuration. Inspired by a YouTube video, I've switched directions to pursue a worm drive transmission.
A bearing ball mounted on a shaft, and a piece of hdpe with a drill hole to centre things. There will be an opposing force holding the hdpe (and mirror) down caused by two more bearing surfaces, one on each of the axis rotors.
To prove the concept of integrating the motors into the structure, I need to test that I can actually move a single axis. This video shows that there is enough force generated to move things. As the structure gets heavier I may have to increase the number of turns of wire (and the voltage).
I tried to operate this motor with three coils to allow continuous travel, but I could not align them in such a way as to allow the rotor to hang free without binding. I think that if I can just get the axle of the rotor to be more rigid without drastically increasing the friction, the alignment will be less of an issue.
Access to 3D printed parts would greatly accelerate this project. I would use the printing time to produce the mirror mounting brackets, radial rotors and the fulcrum tower. I would be able to iterate on my initial design right away!
I have a simpler motion control for 3 axis motion if your interested: its 3 ball joints in a triangle, that stands on 3 legs, two of the legs can be extended and retracted via a threaded rod turning in a nut welded to a tube (basic linear actuator) the other is fixed in length. By extending or retracting the legs a relatively short distance a substantial angle can be achieved with fine threads for the extend/retract motion, the position is very solid when at rest. Motion is slow, but suitable for a sun tracker. I'm building a prototype in a couple months, busy with a bigger projects at present, I have some old 3D models somewhere, the first prototype uses two small 24v DC motors with integrated encoders from an ID card printer... I don't check hackaday much, so if you like, feedback to cubespawn at gmail, Cheers!
Ah, now I see it. Because the Gimbal is offset from the two axes, a pitch and a roll movement would require some part of the frame to deflect to accommodate.
Yes, with only one point of contact the initial sketch can't actually work. I've revised my ideas and now have two additional bearing surfaces that keep the rotor structures from binding against the central post as well as keeping the top gimbal from separating.
I have a simpler motion control for 3 axis motion if your interested: its 3 ball joints in a triangle, that stands on 3 legs, two of the legs can be extended and retracted via a threaded rod turning in a nut welded to a tube (basic linear actuator) the other is fixed in length. By extending or retracting the legs a relatively short distance a substantial angle can be achieved with fine threads for the extend/retract motion, the position is very solid when at rest. Motion is slow, but suitable for a sun tracker. I'm building a prototype in a couple months, busy with a bigger projects at present, I have some old 3D models somewhere, the first prototype uses two small 24v DC motors with integrated encoders from an ID card printer... I don't check hackaday much, so if you like, feedback to cubespawn at gmail, Cheers!