Buck Boost converter works great w/ Solar + Super Capacitor

[kni]

Thanks to the excellent info in this forum, I've put together a few moteino + solar + super capacitor projects. They are all running, indefinitely, on a 7.5f super cap.  It all started with Felix's super cap blog. Thanks Felix!

I've noticed a common problem, however. Properly managing the voltage from the solar panel w/o over driving the capacitor required some experimentation.

Previously, I'd start with a single schottky diode, then I would monitor the voltage in full sun. If the voltage went too high, I would add one or more diodes to get the voltage down around the max voltage rating of the capacitor. Not only is this time consuming, but the result would only charge on the sunniest days.

I don't know why I didn't think of this sooner, but I had a couple of these buck boost converters lying around:
https://www.ebay.com/itm/DD2412SA-2-24V-to-5V-6V-9V-12V-15V-DC-DC-Boost-Buck-Converter-Power-Supply/183396276937?ssPageName=STRK%3AMEBIDX%3AIT&var=690705671250&_trksid=p2057872.m2749.l2649

Wow, did these do the trick! Now my super capacitor projects charge back up to the target voltage at sunrise, even on cloudy days.

[Uncle Buzz]

What about the reverse current consumption when the sun is out (at least at night) ? Do you still use a diode ?

[kni]

Yes, good question. I did not use a diode, just to see what would happen. It appears that this particular converter does not allow current to leak back into the solar panel over night. See screenshot below.

The circuit only lost 0.43 volts overnight, which can be accounted for normal operation. The circuit in question is a moteino + 7.5F super cap + 1M/1M voltage divider + HC-SR04p 3.3v ultrasonic sensor. I suppose if I want to be absolutely sure, I should unplug the solar panel overnight and observe the voltage drop the next morning.

[Uncle Buzz]

I don't talk only about the solar panel reverse current consumption, but also the electronic of the converter who could waste some current when the solar panel won't power it, because it's probably not designed for such case.

Like the MCP1702 who drains some µA from the power supply, but much more when the power supply is lower than its output if you supply the motenio directly with 3.3V on 3.3V pin

From your picture, 5.28V is quite high for a  5V regulator

0.43V in 10h with 7.5F => 1F = 1A/V/s, 0.48V * 7.5F / (10 * 3600s) = 89µA if I'm not wrong, its quite high, but it seems to be not too much for your configuration.

[kni]

Thanks. This is exactly the kind of feedback I was looking for.

Yeah, 5.28v certainly is high, and I would agree it does speak to the overall quality of the device. Fortunately, the circuit is tolerant up to 5.4v (rating of the super capacitor).

The MCP1702 is certainly an efficient device, but according to page 3 note 1 of the spec, the minimum input voltage (Vout + Vdropout) works out to 5.33v, which is too high for my application.

Referring back to the graph, the voltage starts climbing up at ~6:20am, which is right when the sun comes up this time of year. Note, however, the panel itself is still in the shade during this time.

I'm looking for something that will boost the voltage up when the panel is in the shade, as well as regulate the voltage down when the panel is in full sun.

So far, the little buck boost module sold on E-Bay seems to be doing that, but If you know of something more efficient by all means let me know. Unless I know exactly what I am looking for, I tend to get information overload when I search the Mouser website.

[Uncle Buzz]

My example with MCP1702 was not to tell you to replace your buck-boost with it, but to show how an efficient product can be a bad choice if not used as it should.

The problem with buck-boost, if they can be very efficient for some mA and more, they waste some energy to work. With some low power systems like moteino, the wasted energy is often much higher than the needed energy for the whole system!

Anyway, you have some of them, and it works, you don't need to have the most efficient system, you just need to have a working system. If it works, no matter if you waste some free energy!

The solar panel with a diode is not efficient, but it's the simplest way to make things working, you just have to use a small solar panel instead of a very tiny solar panel and a bunch of electronics to make things very efficient... Who cares?