Small Devices - Solar to Supercapacitor charger - Design check

Hi all,

I want to build a small photovoltaic system to supply power to all devices that typically use USB 5V charging.

Since the main power draw will be from mobile Phones, the total energy stored doesn't need to be large but has to withstand many charge/discharge cycles. Additionally the system should cope with many devices connected simultaniously, so decent output power is required as well. That's why I opted for supercapacitors which seem to fit the bill.

So I've got two major challenges, 1. getting the power from the solar panel into the supercaps and 2. getting the power from the supercaps into the devices.

The Charge Controller:
First things first, I will need a charge controller for my supercaps, what I imagined was using an arduino and PWM.
The solar panel connects to a smaller capacitor (as buffer) in parallel after which a mosfet is used as a switch which is in turn controlled by the arduinos PWM.
A voltage divider before the mosfet tells the arduino the voltage of the panel. And since we know the IV-curve of the panel we can guesstimate a good voltage for good power output. Switchig the mosfet with the arduino we can vary the load. (I think this is the closest you can get before famcy mppt, and i want to start simple).

Solar Panel and Supercapacitor selection:
Super capacitors usually come in the 2.7V variety and I can hook them up in parallel or series configuration. Since, as I understand, my charge controller handles the voltage, it shouldn't matter what maximum voltage the supercap array has - as long as we can reach the maximum voltage within the comfort zone of the solar panel and we can shut off the power to the supercaps as soon as they reach their maximum voltage. Balancing and overcharge protection is provided for each supercap individually via tl431 voltage reference and dumping excess power into resistors as is done on the "common" protector boards.
The solar panel in turn should have an open circuit voltage (Voc) of at least 1.5times the maximum supercapacitor voltage to account for temperature and the ~80% Voc maximum power point.

Considering discharge:
To get to the desired output voltage of 5V we will need some kind of boost or buck converter depending on what the voltage on the capacitor array will be.
It seems attractive to just use all capacitors in parallel in order to only use a boost converter which can work to really really low voltages in order to get as much energy out of the capacitors as possible.
If the capacitor array will have a higher voltage than 5v, we need to use a buck conerter which only works at best down unto 5V, potentially leaving a lot of the power unused. I took a look at Buck-boost converters but they seem to need a relatively high minimum voltage, which again is unfortunate.

Just learning, so any comments or suggestions welcome!

You might find this tutorial on a supercap/solar powered Arduino helpful.

You don't need a charge controller with a supercap.

Thanks for the link, a good read as always!

I'm aiming for 3000F to 10000F so quite a bit different scale, using a zener diode to bleed off excess voltage is a bit to dangerous for my scale I think. Also, when the capacitor array is at, let's say 1V I would force the solar panel into the very left of it's IV curve and lose a lot of potential power.

jazzar:
I'm aiming for 3000F to 10000F so quite a bit different scale, using a zener diode to bleed off excess voltage is a bit to dangerous for my scale I think.

Zener has nothing to do with the size of the cap, only with Isc of the solar panel.
Leo..

Wawa:
I'm aiming for 3000F to 10000F...

100,000 Farads? Do you have any clue how much that would cost? You would be money ahead if you buy a Tesla Powerwall and a whole house solar array.

10,000F.

Only $65 at Alibaba, if you order 24 of them. Which is, ummm, $1560.

Wawa:
Zener has nothing to do with the size of the cap, only with Isc of the solar panel.
Leo..

A Zener diode rated ~2.7V will still allow voltages above 2.7V as the cutoff isn't that abrupt. And going below 2.7V to ensure the capacitors never exceed 2.7V will waste some capacitance. Both things I want to avoid which is done by a voltage regulator.

SteveMann:
100,000 Farads? Do you have any clue how much that would cost? You would be money ahead if you buy a Tesla Powerwall and a whole house solar array.

jremington:
10,000F.

Only $65 at Alibaba, if you order 24 of them. Which is, ummm, $1560.

3kF to 10kF, not 100kF :wink: Which puts me out at around 3-5€ per 500F@2.7V capacitor which is a minimum of 18€ for 3000F which I think is reasonable.