Solar-powered model planes

[PaulJ]

'Terrific achievement! Well done!

Paul

[Ralkbirdy]

I hope you all enjoy the building and flying of these planes. My satisfaction is always with simple planes, these are fun just being able to launch and fly when there is good sunshine.

It is however right back to the early days of electric planes and being careful about weight.

My 8 cell plane Is 228g, 28.5g/cell. It now has a 2205 2300kv motor (28g), it has an integrated nut saving a few more grams. 6x4 (4g)prop. This plane has no Pololu regulator. It is flying better with this motor, having lost a few grams of weight and a good prop /motor combination.

12 cell plane is 408g, 34g/cell. It now has 1400kv 2212 52g going to try a 2212 100kv motor and a larger prop.

For 10 cells Martin is using a 28g 2205 2300kv motor with a tiny (5x4?) prop, I would try a 2208 1800kv motor on a 6x4 prop.

For the 8 and 10 cell planes I used a balsa spar, but went for a heavier spruce spar in the 12c plane.

The cells are about 6g, so 8c is 48g + 32g = 80g about 33%. 12c 72g + 16g + 52g = 120g, about 30%.

Fixing the cells is a challenge and this glasscloth method on Depron has worked well. If you have no experience of solar cells, be aware that shadowing of a single cell, will virtually cut of all power to the motor, so avoid shadows.

Mike

[Ralkbirdy]

My 8c plane is more responsive, the 12c with the larger aspect ratio has much more sluggish aileron response, Martin's 10c is good. We talked about a delta with possibly 8c along the back and 4c in front, giving a more responsive plane. With a shorter wingspan I can go back to a 6mm balsa spar.
As there is no moveable weight getting the motor position will be important. This project may start in a couple of weeks time.

[Ralkbirdy]

I am using Frsky Rx, old V8 4 channel, a new 4 channel clone ( see other thread). Servos need to work ok at 3.3v, so far Dymond D47, Graupner 141 and Hitec 45 are all good.

On the 8c model the solar cells are at 4.2v, and everything works fine just connecting direct to the solar cells at the ESC.

For 10 /12c, we have used a Pololu 3.3v regulator which is connected to the ESC power supply and then the 3.3v powers the Arduino and Rx, the Rx is supplied through the normal ESC lead, signal, +ve and -ve.

Supercaps are not needed, the solar cells power the Rx fine in shade, and in Sun the motor is happy.

Even with a brushed motor some form of ESC is needed to switch it on and off, generally they are heavier for similar power, and brushless motor, ESC is easy to obtain so can see no benefit.

The cells benefit from some protection against damage, best to avoid being inside a confined space as getting them too hot is best avoided, performance decreases with temperature..

[Ralkbirdy]

Martin will post the arduino sketch, but here is his basic scheme for the arduino

[Martin]

Edit: I modified the sketch to use the Arduino's internal voltage reference (1.1V) instead of VCC for the Analog reference voltage. That means the recommended pair of resistors for the divider are now 68K and 10K. I'll leave the rest of this post below unedited, but please also refer to the post dated 30th July further down this thread for updates.

What's not shown on that last sketch is the other resistor of the pair that make the voltage divider (this second resistor was 10K and was soldered to the Arduino PCB from A3 to GND).

We use 3.3V 8MHz Arduino Pro Minis, so using the normal Arduino analogue inputs you need to drop the solar cell voltage down to below 3.3V so the Arduino can measure it. With a 12-solar-cell set-up you get about 7.2V maximum voltage (no load) so we used a 15K in line with the input and then 10K to ground. That gives the Arduino a measuring range of 3.3V x (10 + 15) / 10 = 8.25V The 15K / 10K divider is fine for any number of cells up to 13. If you go above 13 cells you'll need to increase the 15K.

The minimum number of cells is set by the requirement to keep 85% of the no-load voltage above 3.5V (with a little bit of headroom for the voltage regulator on the Pololu). This limits the practical minimum number of cells to 8.

The exact ratio of resistors doesn't matter, as the Arduino sketch just measures whatever voltage it sees on A3, remembers the maximum value it's ever seen and then knows that it should do what is necessary with the throttle signal out to the ESC (on pin 10) to keep the voltage at or above 85% of that maximum.

As Mike said, with only 8 cells the maximum voltage you'll ever get is 4.8V, so there's no need to use a Pololu regulator - the Arduino is powered by connecting the solar cell input to V in (sometimes labelled RAW) and allowing it to use its own 3.3V regulator. The receiver and servos operate direct from the solar cells.

With 10 or 12 cells the no load voltage is 6V or 7.2V. Although most receivers and some servos would be fine working at these higher voltages, some receivers and especially servos could be damaged. Mike recommends using the Pololu D24V5F3 regulator. This is available from www.hobbytronics.co.uk Item #2842. This is a very efficient 3.3V regulator and by operating the receiver and servos through this, you keep the servos safe and with a consistent response even when the solar cell voltage varies. Also the servos draw less current when operating at only 3.3V. When the Pololu is present, it makes sense to run the Arduino from it too - so connect the Pololu output to the Arduino's VCC (NOT the V in or RAW).

[Shaun]

Hi Mike,

As Phil mentioned, we are joining the lipo-less free energy flying club. What in your opinion would be the optimum configuration for model wing span, cell count, motor and prop to fly in windier conditions and for aerobatics?

Cheers,

Shaun

[_AL_]

Outstanding project. Really nice.

Al

[Ralkbirdy]

This is my knowledge thread,

https://www.rcgroups.com/forums/showthr ... anes/page2

Acrobatic is probably misleading as they need to spend most of the time facing the sun to fly and even with banked turns there is noticeable power loss, but for being responsive I would go for the 10c flying wing.

If you want a delta there is more option for 12/13 cells in 2 rows.

The 7mm spruce weighs 40g/m, balsa is about 15g/m and is easier to have 6x12 reducing to 6x5 at the tips. Adding glass cloth may be better for wing rigidity on the lower surface than using the 7mm spruce spar, for less weight.

My 8c elevons were fluttering in the wind when I was holding it on the ground so I will add a layer of glass cloth to the underside and see if that seems to stiffen it up, but still getting a slim elevon, not a chunky 6mm one. I saw no flutter while flying.

[Shaun]

Hi Mike,

On my indoor aerobatic models I use carbon rod and thin strip especially on control surfaces to stiffen them up; works well and adds minimal weight. My models tend to have power to weight ratios in the order of 2 to 3 : 1 so I know it can stand the abuse and no flutter is evident.

I only had a hard balsa spar in my enlarged Pulstar. Ample strength and - no need for heavier spruce.

What Size (span) deltas have you in mind for say 12 Cells?

Cheers

Shaun