Flying on the power of the sun is definitely not a new idea, but it usually involves a battery between the solar panels and the propulsion system. [ukanduit] decided to lose the battery completely and control the speed of the motor with the output of the solar panels. This leads to some interesting flying characteristics, almost akin to sailing.
When a load tries to draw more current than a solar panel can provide, its output falls dramatically, so [ukanduit] had to take this into account. Using a ATTiny85, he built a MPPT (Maximum Power Point Tracker) unit that connects between the RC receiver and the motor speed controller. It monitors the output of the panels and modulates the speed of the motor accordingly, while ensuring that there is always enough power to run the servos and receiver. The airframe (named the Solar Bear) is a small lightweight flying wing, with a balsa and carbon fibre frame covered with clear film, with the solar cells housed inside the wing. Since the thrust of the motor is directly proportional to how much sunlight hits the top of wings, it requires the pilot to “tack” against the sun and use momentum to quickly get through turns before orienting into the sun again.
If you want to build your own controller, the schematics and software is up on RC Groups. Check out the Solar Bear in action, flown here by [AJWoods].
We’ve previously covered solar plane entered into the Hackaday prize as well as an actual seafaring solar powered unmanned sail boat that crossed the North Atlantic.
Beautiful! Would it add a significant amount of weight to add cells on the underside as well to allow climbing towards the sun too?
It would be interesting to see a (smaller) supercap in the line to add just a little bit of tolerance to the flight path.
from the link:
“This is a Maximum Power Point Tracker for solar powered aircraft.
It completely eliminates the need for a buffer battery or capacitor. This type of MPPT, along with Sunpower cells has finally made pure solar r/c practical.”
So why not put some solar panels on the bottom as well? I understand it would make it heavier, but it might be worthwhile to prevent stalls from lack of power. Maybe just like 2 – just enough to recover from an electric power stall.
There are flexible solar panels as well, perhaps using them as the wing surface. Guess flying only on cloudless days, if maneuvering in full sun conditions is challenging, probably fatal if the sun gets blocked briefly. Wonder if a small super capacitor would be good enough to smooth things over, when less than ideal electricity is produced. Seems like worth it, if it’ll make it easier to control, and less likely to crash.
The flex PV panels I know have much lower efficiency ( ~ 1/2) than the crystalline ones.
Maxeon C60 cells are thing mono-crystalline type with the top laminated in ITO. Its a good cell which is readily available from the usual etailer sites.
A typical bank angle would be unlikely to expose the solar cells to the sun. I cant see them giving any benefit.
Yup, IMO it’s a bit daft to design something that can’t be controlled in all attitudes, and if you don’t have power to your receiver if a gust flips it, it can’t be controlled.
I think it’s kinda an exercise in working within constraints. I mean… why use solar panels at all? why not use a gas engine like most rc planes? Because it’s fun to limit yourself and see what new and interesting things come out of it.
I think by now the vast majority of RC planes are electric, with the advent of cheap, dense energy storage in Li-po’s and cheap brushless motors & ESC’s.
Brush-less motors are the majority of power plants, as anything with a combustion or jet engine will require specific places to fly, but electric are considered “park flyers”, and can be flown anywhere.
It would be interesting to see if there is any way to make the plane self orient without power.
Does specify that the drive motor only gets power after the demands of the rx and servos.. So it will always be controllable – as those on such a light plane won’t need much power. I’d be surprised if full steering control was not maintainable on a pretty dull day with the cells away from the sun. Its impressive how much power the ambient reflected light can give and how little tiny servos and an rx will take. Just look at the huge RC gliders that fly off tiny cells practically forever (obviously there can come a point when the panels won’t give enough for the servo to hold the flaps position but on that design I’d be very surprised)
You could see (well hear mostly) how well the main motor was still spinning pointed away from the sun and that motor will have a much higher draw.
I think I would rather have a small power store onboard for control reasons though – only few seconds at full throttle worth. Does you no good to control the flaps if you can’t go fast enough for them to really work as needed – I’d probably have that as a toggle boost on the transmitter so you can choose to use only if/when needed thus ensuring it is there so you can avoid hitting that tree. If it was automatic and you got in trouble it could already have been drained.
That’s why all successful cars have wheels on top as well
If one’s goal is to build a plane that can fly reasonably well on solar alone, it would be daft to compromise the design by adding 2x the amount of solar panels for the same output. By adding solar panels to the bottoms of the wings, you’re doubling the weight of the power system, and whenever the bottom panels are providing power… the ones on the top are not, and vice versa. Regardless, in the RC Groups thread, the builder states that the plane can do mild aerobatics – loops and rolls, as is, so it seems there was nothing daft about the design.
With a bigger wing you would be able to soar when the power drops, rather like a solar eagle or vulture, but I guess bigger wing means bigger weight, it would be a fine balance. (Also with a soaring setup once you launch you can use the sun to gain altitude and if a cloud passes by just glide till the sun comes out.)
How about some sort of Super Cap that handles power for several seconds … enough to get it turned back into the sun?
I think even a small nicad like a CMOS backup size one has better power/weight storage though and would probably last a couple of mins to drive the receiver and several servo movements.
+1
RC Group. He’s in good hands with tons that are talented in both aerodynamics and electronics. This’ll be a good one to sit back and just watch.
that’s coooool!
Very nice! Maybe some ingenious optics (Glass fiber?) could make the light exposure better.
A gorgeous build!
https://www.youtube.com/user/rctestflight is at its fourth solar plane design revision. It worth a look.
Coming soon
v4 is up (has been for 6 days): https://www.youtube.com/watch?v=SwbJJoe09DI
very cool project, now put on some long range fpv gear and have fun :)
This is a fascinating project and post. I like the inconsistency of this power system- it’s a bit of a glider, but with a turbo that can provide propulsion *sometimes*.. It begs for a machine-learning piloting system that can optimize flight plans to maximize light exposure on the panels.
Reminds me of this other battery-less solar-powered flight vehicle covered on Hackaday just last year–
https://hackaday.com/2018/08/25/quadcopter-ditches-batteries-flies-on-solar-power-alone/
Great post, Danie!