Will this schematic work for powering microcontroller with solar?

Projects General Guidance
21

Also, a seperated small regulated supply for the 8266 is better, please check its max. allowable supply voltage. Don't over volt you microcontroller and at the same time to charge the batteries.

Better seperate the two system supplies.

23

My suggestion (for you reference only):

  1. A simple method to check the full charge is very easy..... set each battery final voltage to 1.4V, thus
    total voltage of buck regulator = 4.2V + diode voltage drop too.

  2. Take out your 8266. Take out the MOSFET, LED

  3. Ensure the final total battery voltage <=4.2V by checking every 30 min. for voltage and temperature.

(In your design state, to ensure low changing current, better make sure each battery voltage has at least 1.2V, otherwise charge it by formal charger ..... because you still have no current limit circuit)

  1. In this way, the battery should be charged to full in 6 hours. I suppose.

  2. Discharge the batteries to see it is charged, say, by a flashlight

Cheer.

24

As Ray25 says, it would be useful to know what happens when it's just the panel, the buck converter and the batteries. No processor or LED. Do the batteries charge then? What is the voltage on the input and the output of the buck converter when that charging is going on? How high does the battery voltage get?

But basically, if the panel and buck converter can't provide enough power for the processor and LED, then the battery will be in continuous discharge mode to make up the current deficit, and will never charge.

Edit: I question the use of a regular buck converter with a solar panel. If you add the current requirements of the processor, the LED and the battery charging, the buck converter may drag the panel voltage down to a very inefficient point, and the panel will not be producing much power. It seems something like an MPPT converter would be needed, which tries to keep the panel at its maximum power point for the current level of illumination. Maybe even a linear regulator would do better, and you might not need the diode.

25

UPDATE: The batteries were old and couldn't deliver current properly. Thank you all for your suggestions.

26

NiMH is very robust and can be trickle charged indefinitely. This in contrast to Li-ion batteries which will start doing unwanted things, such as bursting into flames.

A simple current limiting resistor is enough for NiMH charging, tough for faster charging a proper charge circuit is needed.

27

Joule Thief fed by small solar cell trickle?

28

The YX8018 or one of its many varieties works great for this - these chips you find in all those cheap Chinese garden lights. They're also readily available on sites like Taobao or Aliexpress. All they need for external components is an inductor (smaller inductor for larger output current). It's basically an advanced joule thief, with battery protection and darkness detection.

However, circuits like a joule thief or the YX8018 are a current source, not a voltage source. That makes it great for running LEDs, not so much for running an MCU as they will push the voltage easily to 20V or so - as an experiment I've operated a strings of several LEDs in series on a single NiMH battery and a YX8018 based circuit. Trying to regulate the voltage means you're going to have to dump all excess current. A waste of energy.

2 Likes
29

I found that 4 diodes can tame and rectify piezo spikes way over 20V. I've used a cap to hold piezo tap power to keep a led lit. That's about a trickle.
I have small boost converters, 0.9V min to 5V.

1 Like
30

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.