If I want to charge a battery from a solar panel without using a commercial charger, then the essential problem is to create two voltage-triggered switches. One switch turns current on when the voltage in the battery drops below a certain threshold and the other disconnects the panel when the voltage rises to the "full"/higher threshold.

One advantage of having a system like this is that I can select the threshold voltages, instead of the charger guessing what might be optimal.

What would be the right kind of circuit to solve this problem? Ideally, I would like a system that requires no power to operate, at least when it is charging, because I want the energy to be used to charge the battery, not run the circuit.

  • \$\begingroup\$ The term you're looking for is "energy harvesting". \$\endgroup\$ Commented Sep 26, 2016 at 0:39
  • \$\begingroup\$ Maximum power point tracking \$\endgroup\$
    – Matt Young
    Commented Sep 26, 2016 at 1:09
  • 1
    \$\begingroup\$ What kind of batteries are you charging? \$\endgroup\$
    – user57037
    Commented Sep 26, 2016 at 2:19
  • \$\begingroup\$ It seems to me that the problem is a bit more complicated than you appreciate. Also, good chargers don't "guess" what the optimal charge voltage should be. Good chargers are optimized to the specific battery chemistry and temperature to charge them quickly and without damaging the batteries. \$\endgroup\$
    – user57037
    Commented Sep 26, 2016 at 2:27
  • 1
    \$\begingroup\$ After your edit it now seems clear that you actually only want one switch, which turns charging current on at the lower threshold and turns it off at the higher threshold - right? \$\endgroup\$ Commented Sep 26, 2016 at 3:03

1 Answer 1


This disadvantage of this method is the loss of energy if the PV ope circuit is more than 18V due to impedance mismatch or lack of Maximum Power Transfer or PMT which runs approx 80% of Voc.

But this is basically how a car alternator works with lots of power and inefficiency behind it.

YOu can choose a high side Logic switch for automotive high switched currents and logic level input from a comparator with a precision reference voltage to scale 14.2 down to the reference voltage, assuming lead acid battery.

If that is of interest, I or someone can show you this inefficient method.

Ideally you want the battery voltage to be 80% of the open circuit voltage or Voc.

This will explain why a 42% improvement is possible with PMMT. enter image description here

ref www.roguepowertech.com/documents/MPT3048Rev2-Manual.pdf

  • \$\begingroup\$ You have to have a lot of hysteresis or a time delay. Otherwise the FET may turn on and off rapidly during some conditions, and cause the FET to essentially go into the linear region. This could lead to excessive dissipation. This is one of the pitfalls. If the panel is well matched to the battery (which the OP still has not identified) it can be a decent way to go. The inefficiency is not that bad. \$\endgroup\$
    – user57037
    Commented Sep 26, 2016 at 3:15
  • \$\begingroup\$ good point, you can add 1~3% positive feedback with a delay cap and you need a Schottky diode to prevent reverse drain \$\endgroup\$ Commented Sep 26, 2016 at 3:18
  • 1
    \$\begingroup\$ then one can add an Imax current rated choke and call it a DC-DC converter with 10kF battery and small plastic cap added for improved efficiency.. \$\endgroup\$ Commented Sep 26, 2016 at 3:33

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.