Powering DC Motor from Solar panel

[mensbufo]

Hello, everyone! Sorry if I say something silly, I am just a beginner :'(
I have a school project where we need to build solar-powered vehicles (a boat in my case). The rules are as follows: No supercapacitors or batteries; the boat can only charge for the first 30 seconds of the match. From what I could find at home, I gathered 4 solar panels, each rated for around 1-2W, with a voltage ranges around 4-5V. What I need to do now is select a DC motor, but the bigger question is whether it is worth overengineering something like a simple MPPT charge controller to charge some large capacitors, or something similar that will stabilize and control the motor's power from the solar panel? If possible, can you recommend some simple solutions or ideas of what I should add on top of just the bare motor/solar panel setup? And maybe even add some super simple radio control to turn left/right, on/off motor, in case if the boat will try to change its course, but this is additional and will depend on how well the boat will float.
Thank you for your answer!

[BeBuLamar]

Do you know the lighting condition at the race?

[mensbufo]

Yes, solar panels will gain power from sun. Produced power will depend on weather, but of course, in case of bad weather, the race will be rescheduled to a sunnier day.

[Kim Christensen]

A regular capacitor in parallel with the solar panels will help a lot. It'll provide the motor start current which will be much higher than it's running current.

[Algoma]

Effeciency of the panels are the most critical to what you can do.

Is it a distance or endurance race? If you're racing for distance, Overvolt that motor for more RPM. Most DC motors will run at twice their rated voltage for a shorter life, provided its kept from overheating too quickly. In a race that motor is disposable, it only needs to last the race. A nice compact DC-DC converter will run any other electronics if needed. Just need a couple of caps to absorb the noise form the motor.

Otherwise its a balance of high torque @ low speed using a large prop ... or High speed motor using a smaller prop to move more water. It depends on the number of poles in the motor.  If its fresh water. perhaps put that motor right in the drink for cooling while you vastly overpower it. a DC motor can run wet if you dont care for it to survive long. Though the fluid resistance will limit RPM of an flooded open motor.

Given your 4 panels, I would say 2 Series Groups, wired in Parallel to try to put 8-10V through a 6V motor.. If you're motor is too big it will drop the voltage of the whole system. Monitor the voltage across the motor under load, if its dropping too low, Series the whole 4 panels together to keep that voltage as high as possible in more lighting conditions. Wet solar panels due to water drops can loose a lot of effeiency.

[mensbufo]

But around how big capacitance I need to connect in parallel with motor? Also my idea was for capacitors to gather energy in the first 30 seconds of the match, when it is allowed, and then power motor from them, because then I can power up motor with stable current. Is this idea garbage or not?

[Algoma]

Its good.. yes, As many capacitors as you're allowed within a reasonable weight limit. then use a mosfet of reasonable size to turn on the motor from your controller. They can really only enough to provide the current to kickstart the motor at the full available voltage of the panels and get things moving. Capacitors generally cant store much energy compared to what a motor will use without adding weight. They really only help with getting the inital intertia of the motor, water and boat moving. After that its the weight of the boat and fluid resistance. Once the boat is moving the capacitors could burdon on the system, unless the panels are producing more energy than the motor uses in running operation.

Like a single phase AC motor that use a centrifugal switch to disconnect the Start Capacitor once its up to speed. And then only maintains a smaller run capacitor connected for effiency. You'll need to test variations to find the right balance.

[Algoma]

Being a race, perhaps a small relay with the motor on the Normally Closed contacts would work better than a mosfet. With the controller holding the relay open until the start signal is given. That way if the controller looses power or reboots mid race, the motor can keep running until the controller can hopefully wake up again and resume control.

[Ian.M]

The ban on batteries and supercaps means you cant gather a useful amount of energy economically in the 30s charging time.  e.g. if you buy the largest capacitance electrolytic capacitor Digikey stock: https://www.digikey.lv/en/products/detail/nichicon/LNR1A225MSE/3930044
and charge it to 10V you'll only store 110 joules of energy, about 12% of the energy in a single lithium 2016 coin cell.  However that capacitor will cost you over €240, is about the size of a drink can, and, with a suitable high efficiency charging circuit will add over 1/2 kg weight to your boat, increasing its displacement and thus drag.   If you needed lots of power for only a couple of  seconds it might be helpful, but over several minutes or longer the stored power wont make up for the drag.

Algoma suggests that a capacitor bank may be useful to provide the starting surge for your motor to allow it to accelerate  to full shaft speed very quickly.  That makes me ask if the capacitor bank has to be on the boat?  If the capacitors can stay on the shore, and there is a low or zero force disconnect for the wires between them and the boat as the boat pulls away from the shore, the weight of the capacitor bank becomes irrelevant.  You can design to over-voltage the motor considerably using the 30s accumulated energy, for a few seconds to accelerate the boat, then leave the panels driving the motor direct with no heavy energy storage or power conversion electronics onboard!

[Algoma]

not sure the limits of the rules, but at light weight pumped water / Compressed air storage system.. Pump water into a small, sealed, air filled plastic bottle for the 30 seconds, then use the triggered expulsion of water through a small pipe intially drive forward.. again its a (Added water weight / drag trade off) unless its allowed to eject accessories. .. At that rate, A simple spring loaded inertal weight is drawn in the first 30 seconds and then forcefully ejected from the craft rearwards at the start.

keep it simple, focus on lower weight, more power to the motor. Tortise vs rabbit type race here. A lighter boat, lower drag with more balanced motor for the task will always win. it would be better to engineer simple Hydroplanes to reduce drag, than add a bunch of other dead weight that can fail.

[Algoma]

Another idea for pre-race energy storage, Flywheel / paddle wheel for propultion. Spin up out of the water pre-race and drop the wheel into the water at the start. DC motors like flywheels, and 3D printed pulleys are quick to print out various speed / torque adjustments using an elastic band as a belt.

[Shell Albert]

I have experience of energy harvesting from solar panel.
below is my schematic, wish it could be helpful.
the first stage is a booster, charging the farad capacitor to a desired voltage, in my design, it's 4.5v.
for my low power consumption application, 1F is adequate. But for driving a motor, big capacitance will be needed in my opinion.

[beanflying]

Have a good trawl through the links and the rest of site here https://www.modelsolar.org.au/tech-support-old/design-guide-student-design Way back when my Dad and a Uni Lecturer set this up as a Statewide program.

The MPPT has become a bit of a must to be competitive as pulling the Solar array to a low voltage costs way more potential Watts in than the MPPT consumes. There are some curquit diagrams out there online but most competitors went the easy option and one of these https://www.scorpiotechnology.com.au/starter-solar-challenge-kits

Typically Faulhaber or similar quality motors get used by the winning cars.

[Siwastaja]

I don't think any sensible amount of electrolytic capacitors would help at all. In 30-second timescale, supercapacitors and batteries are the only sensible energy storage options, and if they are banned, then the 30-second "charge" time is a red herring. This leaves you running the motor in realtime by whatever amount of light is produced at that very moment.

You definitely could do maximum power point control directly to the motor, then only modest amount of capacitance in parallel with the PV panel is needed, basically to satisfy the ripple current need of the buck input (which can run at pretty high frequency, say 20-50 kHz). When I say buck, I just mean a half-bridge driving the motor; the motor itself is the inductor, and the motor RPM (back-EMF voltage) is equivalent to output "capacitor".

It doesn't need to be an actual maximum power point tracking algorithm, even just constant input voltage control would be an improvement over direct connection of the motor. Basically, something which understands if the panel voltage is sagging too low and reduce motor duty cycle as a result.

You see, if you directly connect the motor, then maximum motor current at zero RPM will be just the short circuit current of the panel, slightly more than I_mpp. On the other hand, if you run the panel at I_mpp at V_mpp, and buck down for the motor (again, motor winding itself does the job of the inductor), then motor current can be many times higher than I_mpp. Since power is conserved in a switching converter (which motor controller is), e.g. 10V * 1A input can provide 1V * 10A output for the motor - minus losses of course, so maybe 1V * 9A in reality.

Does this make a lot of difference, I don't know. Maybe in a boat, even with unoptimal low-speed torque, the motor pretty much accelerates near maximum speed in split second anyway, because there is little inertia, and the work done by the rotor kinda quadratic to RPM, so there is not much mechanical load at low speed anyway. In a car things would be pretty different; higher motor current during the first few seconds would give you better acceleration.

[tszaboo]

I don't think MPPT makes sense. You either have an undersized motor, or you have an undersized solar panel.
So if you have more energy available than your motors nominal capacity, what do you do with it? Drive the motor faster. Your output voltage changes with the input voltage.
If you just select a solar panel that has it's MPPT voltage at the nominal voltage of your motor, the system will be quite optimal. So 6V solar panel and 5V motor. Or 14-15V solar panel and 12V motor.

[beanflying]

I don't think MPPT makes sense. You either have an undersized motor, or you have an undersized solar panel.
So if you have more energy available than your motors nominal capacity, what do you do with it? Drive the motor faster.
If you just select a solar panel that has it's MPPT voltage at the nominal voltage of your motor, the system will be quite optimal. So 6V solar panel and 5V motor. Or 14-15V solar panel and 12V motor.

The issue is the load is not constant so it collapses the voltage produced by the Solar Panel and the car stalls in particular it costs a heap of time on getting going if at all in low light conditions. The program is now 30 odd years old in this state alone and the reasons to run some form of MPPT/Switchmode controlled supply have actual data and race results behind it.

The last local School program I was involved with helping 6 or 7 years ago I was teaching the 11 and 12 year olds basic soldering to install the controller built by a local HAM using an online source circuit. (If I find the diagram I will link it or post it here) We ran basic tests with and without to show the kids WHY it was being used and it was easy to measure performance changes without datalogging anything but just by looking or using a stopwatch.

Another Aussie Solar Car controller (not model) https://www.aerl.com.au/solar-car-controller/

[Siwastaja]

The issue is the load is not constant so it collapses the voltage produced by the Solar Panel and the car stalls

Well said, this is exactly what happens. Solar panel roughly provides constant current, therefore approximately constant torque. Constant torque means linearly decreasing power, as speed drops. But the panel could produce constant power, if used at optimum voltage, not at the motor's voltage.

By adding a motor controller, we can break free from this relationship, i.e., step down voltage, and thus provide higher current at lower motor voltages, and thus, much more torque at low speeds. High speed torque would still be limited by the panel power, of course.

Improvement in top speed would be pretty limited compared to a hand-picked combination of panel+motor, but acceleration from zero speed would be drastically faster.

But as I said, with a boat, I don't know. There is little mechanical load at zero RPM anyway because of fluid dynamics and how propellers work. Even with starting torque limited to the same as maximum speed torque, it would speed up quickly anyway. Adding MPPT of course helps, but I don't know if it helps enough to be worth the trouble.

[Ian.M]

Is it so much trouble vs sourcing a motor that matches the panel (assuming the panel is standardised for the competition)? 

The boat doesn't need reverse so the effective minimal buck converter circuit is the motor, a low-side MOSFET, a free-wheeling Schottky diode across the motor, and just enough decoupling on the supply to keep the ripple reasonable.  Add a low pin count 8 bit MCU to generate the PWM, monitor the panel voltage and attempt to stabilise it at the pre-computed max power point.  If you want to gold-plate it, give the MCU a photo-sensor so it can monitor illumination and adjust the target voltage accordingly, or if you are feeling brave, implement MPPT.   

You almost certainly need the MOSFET + a control circuit just to start the motor at the GO signal, so the extra weight is minimal.   If you have a negative aptitude for coding, even a CMOS 555 PWMing the motor to match it to the panel would be better than a mismatched motor.

[Seekonk]

Any buck converter can easily be turned into a linear current booster by monitoring panel voltage and faking the sense pin into thinking there is over voltage on the output.  No need for MPPT.  Having a fixed voltage on the panels gets you 95% of the way there. I've been doing this for years to make cheap charge controllers.

[tszaboo]

I don't think MPPT makes sense. You either have an undersized motor, or you have an undersized solar panel.
So if you have more energy available than your motors nominal capacity, what do you do with it? Drive the motor faster.
If you just select a solar panel that has it's MPPT voltage at the nominal voltage of your motor, the system will be quite optimal. So 6V solar panel and 5V motor. Or 14-15V solar panel and 12V motor.

The issue is the load is not constant so it collapses the voltage produced by the Solar Panel and the car stalls in particular it costs a heap of time on getting going if at all in low light conditions. The program is now 30 odd years old in this state alone and the reasons to run some form of MPPT/Switchmode controlled supply have actual data and race results behind it.

The last local School program I was involved with helping 6 or 7 years ago I was teaching the 11 and 12 year olds basic soldering to install the controller built by a local HAM using an online source circuit. (If I find the diagram I will link it or post it here) We ran basic tests with and without to show the kids WHY it was being used and it was easy to measure performance changes without datalogging anything but just by looking or using a stopwatch.

Another Aussie Solar Car controller (not model) https://www.aerl.com.au/solar-car-controller/

I don't think a school project expects you to do MPPT and all kinds of magic.
If you don't have a sunny day, you postpone the race.
You would need to do PCB design for starters, because you wouldn't get a MPPT controllers for motors. Or maybe you can wire it up with a FET, use the capacitor as a CV source, PWM the motor, use it's inductance. And deplete the capacitor by the end of the race. Do you expect a high schooler to do all that?

[Doctorandus_P]

The voltage over a capacitor of 1F will change with 1V/s when a current of 1A is drawn. (In metric countries, I do not know about bananas). Even if you have a 10.000uF capacitor (= 10mF = 0.01F) it will be empty quickly when it's only charged to a low voltage (solar panel).

I guess your best option is to optimize for energy efficiency. From hull and propeller geometry to the motor type (BLDC is in general a lot better then brushed motors) and the electronics.

You can also think of other ways to store energy. For example when pumping water into a soda bottle (without letting the air out).

[beanflying]

I don't think MPPT makes sense. You either have an undersized motor, or you have an undersized solar panel.
So if you have more energy available than your motors nominal capacity, what do you do with it? Drive the motor faster.
If you just select a solar panel that has it's MPPT voltage at the nominal voltage of your motor, the system will be quite optimal. So 6V solar panel and 5V motor. Or 14-15V solar panel and 12V motor.

The issue is the load is not constant so it collapses the voltage produced by the Solar Panel and the car stalls in particular it costs a heap of time on getting going if at all in low light conditions. The program is now 30 odd years old in this state alone and the reasons to run some form of MPPT/Switchmode controlled supply have actual data and race results behind it.

The last local School program I was involved with helping 6 or 7 years ago I was teaching the 11 and 12 year olds basic soldering to install the controller built by a local HAM using an online source circuit. (If I find the diagram I will link it or post it here) We ran basic tests with and without to show the kids WHY it was being used and it was easy to measure performance changes without datalogging anything but just by looking or using a stopwatch.

Another Aussie Solar Car controller (not model) https://www.aerl.com.au/solar-car-controller/

I don't think a school project expects you to do MPPT and all kinds of magic.
If you don't have a sunny day, you postpone the race.
You would need to do PCB design for starters, because you wouldn't get a MPPT controllers for motors. Or maybe you can wire it up with a FET, use the capacitor as a CV source, PWM the motor, use it's inductance. And deplete the capacitor by the end of the race. Do you expect a high schooler to do all that?

Follow the links in my earlier posts the MPPT's for this sort of model Solar Car are a retail product. Also look at the general scope of these Vehicles, limited Solar Panel Wattage to level the playing field and avoid an arms race on efficiency etc.

The electrical optimization has worked for several decades and it is now far more a vehicle design contest and getting it all to work at it's peak.

[Ian.M]

I'd love to be wrong, but without supercapacitor energy storage, I doubt a small solar powered electric hydrofoil will have enough thrust to get the hull out of the water and planing.  A conventional planing hull will have higher drag at transition than a hydrofoil.  As a conventional displacement/semi-planing hull's max. efficient speed is limited by the square root of the waterline length,  and a lightweight large area platform is needed to support the solar panels, the optimum hull type is probably a catamaran, with long narrow hulls (possibly rigid plastic tubes bent up at the stern and bent up some, heat-softened and flattened to a knife edge cutwater at the bow).  A design displacement with the hull tubes half immersed would minimise wetted area and thus drag.   The biggest efficiency gains are likely to be in the propeller and drive train.  Minimising the shaft angle is likely to be desirable, so that favors a gearbox on the forward cross-beam, with the shaft running back from it between the hulls through a P bracket on the rear cross-beam, with the motor mounted near centrally, positioned for optimum fore & aft trim.   As beanflying points out the controllers are an off the shelf product,  its probably worth concentrating on iteratively optimising the hull design by tow testing to measure the drag at various speeds, when loaded with the known weight of the panels and a best guess at the weight of the propulsion system, then using that data to select a controller, motor and gearbox.  Then the propeller need to be optimised.   Best efficiency will be a slow turning, large pitch large diameter propeller, but the tradeoff is steeper shaft angle and greater gearbox weight, so a faster turning smaller pitch and diameter propeller may come out ahead in practice.