1
\$\begingroup\$

I already did some testing. I have a 220mW solar panel that charges a 3V supercap (takes a long time) which eventually powers my wifi module that needs ~1.5W (1.5 J/s) for 1 second then ~90mW (0.09 J/s). I have a boost converter after the supercap to keep steady voltage for at least 30 seconds until module dies out, which is okay for now.

My main question is, is it possible instead of using the MPPT to charge a battery, can I use it to charge my supercap? I want to maximize the charging time. Right now with my solar panel, the efficeny is terrible. I can get 2.85V, but the current is around 0.03 amps. Which charges my capacitor super slow.

\$\endgroup\$
3
  • \$\begingroup\$ I want to maximize the charging time????? \$\endgroup\$
    – Andy aka
    Aug 20, 2020 at 12:02
  • \$\begingroup\$ Yes, it's possible. \$\endgroup\$
    – user253751
    Aug 20, 2020 at 12:05
  • \$\begingroup\$ Not easily as you have a current source going into 0 Ohms ESR well below Vmpt \$\endgroup\$ Aug 20, 2020 at 14:53

4 Answers 4

1
\$\begingroup\$

Right now with my solar panel, the efficeny is terrible. I can get 2.85V, but the current is around 0.03 amps. Which charges my capacitor super slow.

A 100% efficient MPPT buck converter can charge the capacitor without energy loss. This means the capacitor charges to some desired voltage and acquires an energy of ½ C V². And, because the MPPT buck is 100% efficient it only required an energy from the SP of ½ C V².

Now compare it with a resistor charging the capacitor; the energy lost in the resistor are equal to ½ C V² so, to charge a capacitor this way is 50% efficient compared to using a 100% efficient MPPT buck converter. Compare it with any conversion that doesn't use an inductor as an intermediary energy storage device and the efficiency is still 50%.

A practical MPPT buck converter might have an efficiency of 90% and therefore you could argue that using this you might need 40% less energy to charge the capacitor. That also translates to it charging faster but the gains might not be worth the complexity especially if you got a bigger SP.

\$\endgroup\$
3
  • \$\begingroup\$ From my circuit, I do have a resistor charging a capacitor from the solar panel. You're saying that it's only 50% efficient that way and I should look into a MPPT buck converter because it could get me almost 100% efficiency? I have a 3V solar panel which I need to be at 3V to charge the 3V rated capacitor, so I dont need to step down the voltage, correct? Also, does the MPPT maximize my power from the SP? will I be able to get 220mW of power in direct sunlight? \$\endgroup\$
    – Sherman
    Aug 20, 2020 at 15:31
  • \$\begingroup\$ I'm not recommending that at all. I'm making a comparison between using a resistor and using MPPT and concluding that it may not be worth it but, YOU have to decide whether the gains for using MPPT is worth the effort given that you could use a bigger SP. I assumed your SP could produce more than 3 volts hence why I thought you'd need a buck MPPT circuit. Of course MPPT maximizes power - that's what it stands for. I can't tell you what power you can get from your SP because I have no idea what SP you have (and that wasn't your question anyway). \$\endgroup\$
    – Andy aka
    Aug 20, 2020 at 17:10
  • \$\begingroup\$ If your panel is only 3 volts and your capacitor is to be charged to 3 volts, I doubt very much that you will get any serious benefit from MPPT. \$\endgroup\$
    – Andy aka
    Aug 20, 2020 at 17:12
1
\$\begingroup\$

you shouldn't need a resistor between the PV source and capacitor, having one there wastes power and it's not needed for a PV panel.

Be especially careful the maximum open circuit voltage of the panel doesn't exceed the maximum rated voltage of the capacitor and use a diode to prevent the cap from draining through the panel. A fuse is recommended in case the diode shorts out.

A zener diode rated at the maximum voltage rating of the cap can also limit the voltage if the PV panel can produce a higher voltage.

OK, that's the simple system but it's not as efficient as maintaining the PV panel at it's MPP.

I've been looking for a suitable MPP controller specifically for this application but have been unable to find one.

I've thought of how it would work.
A high efficiency DC-DC buck converter is needed to step the voltage down from the MPP voltage of the panel to whatever voltage is currently in the capacitor. Constant voltage will not work. But, this alone isn't enough because the capacitor has such a low internal resistance that it will pull enough current to drag the panel voltage down to near it's short circuit voltage... less than the MPP voltage. So, a voltage sensor measures the PV voltage and reduces the DC-DC converter output voltage if the PV voltage drops below it's MPP, and raises the DC-DC output if the PV voltage goes higher than it's MPP. That's it. Just doing that should maintain the panel at it's MPP and crank the maximum amount of current available (taking into account the ≈90% efficiency of the buck converter) into the capacitor.

Easier said than done, will likely have to build a custom controller... I've evaluated inexpensive buck converters and the way they operate would be difficult to adapt.

\$\endgroup\$
2
  • \$\begingroup\$ You are right about maximizing efficiency by maintaining the PV panel at its MPP. But the MPP moves around as the irradiance varies (clouds / seasons / time of day / flock of seagulls / etc ). As the sun gets brighter, the whole curve moves up, and the MPP will still be at the knee of the curve (but at a higher location because the curve moved up). There is a "perturb and observe" algorithm to look up that should do it. It basically slides up and down the curve a bit to see if the whole curve has shifted. I think a separate sensor is sometimes used to determine irradiance and help with this. \$\endgroup\$ Jan 7, 2023 at 20:13
  • \$\begingroup\$ partially true, I've found in practice though the mpp voltage doesn't change a lot with irradiance and clouds, changes a bit more with temperature. But we're talking about very small changes... using a constant assumed mpp voltage is going to work far better for this application than charging through a resistor or using a direct connection or other circuit that pulls panel voltage down significantly lower than it's mpp voltage \$\endgroup\$
    – user309900
    Jan 9, 2023 at 11:40
0
\$\begingroup\$

For a simple MPPT regulator use a converter to regulate the PV to 80% of Voc. Using a PD sun sensor or estimate with series FET voltage regulator with Gain feedback on voltage error. This may smaller e-Caps with low ESR to integrate then pump superCap like an hand air pump with suitable half bridge drivers.

Normally Vmpt varies from 82% to 72% for useful energy. Yet starting from xx low voltage requires a switched LC current storage with impedance matched to Voc/Isc=Zmpt.

Since boost regulators are efficient from voltage sources but very inneficient from current sources due to impedance mismatch. A carefully designed SEPIC converter might match Vmpt to Vcap if this equation is realized by design.

Examine and specify the V,I,Z during charging at each stage and include in the design question.

\$\endgroup\$
2
  • \$\begingroup\$ You have a lot information. I have a 2×2 inch 220mW Solar panel. How big are these inverters, converters and regulators? \$\endgroup\$
    – Sherman
    Aug 20, 2020 at 15:36
  • \$\begingroup\$ Only as big as they need to be to match source impedance and drive load \$\endgroup\$ Aug 20, 2020 at 17:10
0
\$\begingroup\$

You need a converter to match both the output of your solar cell at the MPPT point (which changes based upon how much light is hitting it), and also match the input to your supercapacitor.

In a sense it's a dual MPPT, because you want to maximize the power transferred out of the solar cell, but also maximize the power transferred into the supercapacitor.

Most MPPT converters target Lead-Acid batteries and are optimized to deliver into that kind of energy storage device, and are not prepared to handle the "battery" starting at zero volts and going to only 3 volts. But both devices are very low impedance, so do have some similarities that could be leveraged from, perhaps, an existing design.

To prove that it is possible, Please see the following simulation that I set up, a self-oscillating buck-boost (inverting boost) converter for a supercapacitor charger, and it achieved 75 percent efficiency, which is much better than a fancy resistor (an LDO or other linear regulator solution).

D1 is a 3 volt LED that sinks the power once the supercap is charged, and therefore prevents overcharging the supercap. Most of the power goes into charging the supercap at first, then switches over to lighting the LED once the supercapacitor has reached 3 volts. In real life, you probably want a more reliable solution to replace the LED, because of forward voltage variability of each LED within the same batch, and also significantly different forward voltage with temperature. You'll need a good voltage reference used to decide when a darlington or Sziklai-pair starts to turn on.

I have hidden actual values for protection, and your values will be different anyway, since the "supercap" was set to 0.02 Farad to keep simulation times reasonable.

self-oscillating buck-boost schematic with traces

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.