I wish to connect a couple of small photovoltaic panels and a small motor used as energy generator to a small battery. I found these components:


Rated power: 0.55W
Output Voltage:0.01V~5.5V
Output current: 0.01mA~100mA

2x solar panels (in series). Specs for each panel

Rated power: 0.3W
Output Voltage: 3V
Output current: 100 mA (computed by myself doing 0.3W/3V)

NiCD Battery:

Output Voltage: 4.8V

   Current Not sure.   
   I should read what is written on its charger 
   but I assume it should be something around 250 mA. 

What is the most appropriate way to connect them?

I sketched an idea but I am not sure if it will work.
Any constructive feedback is highly appreciated.

enter image description here

  • \$\begingroup\$ Depends what sort of battery. Definitely not if it's Li-ion. \$\endgroup\$
    – Finbarr
    Commented Aug 2, 2023 at 12:50
  • \$\begingroup\$ Ah ok, good to know. However, the battery I linked is a NiCD battery. \$\endgroup\$
    – Barzi2001
    Commented Aug 2, 2023 at 13:36
  • 1
    \$\begingroup\$ Schottky diode drops about 0.3V. PV will be OK with 3 x NimH but motor will depend critically on actual voltage generated. || You REALLY need to decide what sot of RPM it will achieve and test it. It MAY produce more V under low load and so load down OK to suit situation. PV panel is close to a constant current source so is "happy" if loaded to a lower voltage. \$\endgroup\$
    – Russell McMahon
    Commented Aug 3, 2023 at 9:25
  • 1
    \$\begingroup\$ @Barzi2001 The generator voltage WILL vary with wind speed. Voltage at various speed will probably matter given motor voltage ratings. What rotor do you intend to use? What diameter? What wind do you expect. || Expected power is 0.6 x V^3 x A x z || V m/s wind, A area m^2, Z = efficiency 0-1 Use 0.1 to 0.2 to start. || This facebook group should be helpful. (I'm a member). facebook.com/groups/windturbinemakers \$\endgroup\$
    – Russell McMahon
    Commented Aug 3, 2023 at 13:43
  • 1
    \$\begingroup\$ @Barzi2001 The 10 Ohm resistor is not needed IF it is intended to allow the TL431 to function as a clamp. It does however reduce the TL431 clamp current to blow the 100 mA max allowed. If you want the TL431 to sink more current than its max 100 mA you can use it to drive a PNP transistor or a P-Channel MOSFET. || Examples: see this SEEE answer by Spehro . Or using a MOSFET here \$\endgroup\$
    – Russell McMahon
    Commented Aug 10, 2023 at 12:00

1 Answer 1


This type of circuit will risk overloading your battery. NiCD is somewhat less sensitive then Li-ion, but this setup is not a good idea unless it's for a short demo. For a better solution, consider a battery charger chip like the MAX712 (see https://www.analog.com/media/en/technical-documentation/data-sheets/MAX712-MAX713.pdf)

Or even better, discard the NiCD batteries and use a Li-ion pack with a dedicated module such as the MH-CD42 (https://techfun.sk/en/product/charging-discharge-circuit-mh-cd42-for-batteries-with-5v-output/)

  • \$\begingroup\$ Good point! Thanks! Just for my understanding: you claimed that my scheme(s) is(are) not a good idea because the battery could reach 100% State-of-Charge (SoC) and the PV/motor could keep in pushing current into the battery? If so, what about estimating the SoC and then move the switch position on the shunt resistor depending on the current SoC (i.e. if the SoC > 90% move the switch to the shunt resistor, if SoC<35% move the switch to the battery)? Or there are other reasons by which my scheme is not good? \$\endgroup\$
    – Barzi2001
    Commented Aug 2, 2023 at 15:47
  • 1
    \$\begingroup\$ Yes, that would work, but if you have to do that manually then you really need to stay close and watch the circuit (which is why it's fine for a short demo), and if you want to automate the switch then measuring the actual SoC, under load, is fairly complicated. Both the chip and the module I mentioned do exactly that. \$\endgroup\$
    – Paul
    Commented Aug 2, 2023 at 16:56
  • \$\begingroup\$ This MH-CD42 looks awesome! My intention is to design a Kalman filter for SoC estimation and to control the switch (replaced with a potentiometer in the revised scheme) to track a desired SoC and keep it (say 40%) no matter of the load. The idea of using such variable shunt resistor is due to that if the SoC reaches its target (say 40%), then I reduce the electrical power at the input of the MH-CD42 to keep 40% os SoC while feeding the load. Not sure if it make sense, anyway: new scheme: ibb.co/nM86mg8 \$\endgroup\$
    – Barzi2001
    Commented Aug 8, 2023 at 8:07

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.