Here's what I'm trying to do.

I'm trying to make it where I can run any circuit requiring 5 volts off of 7.2 volts. I get 7.2 volts from combining two 3.6v ni-cd phone batteries but for simplicity I show it as one center battery in my circuit.

The source power shows as a battery but I plan to use a wall adapter that outputs either 9VDC or 12VDC, but I figure 9VDC is more efficient.

The battery shown on the right is the output (5V). I could replace it with an LED and resistor in series but I used a battery symbol for simplicity.

Here are my questions:

  1. If I used such a circuit, would I be able to successfully charge a ni-cd battery to max charge without blowing up the battery or is my circuit flawed?

  2. If my circuit can be successful, is there a way to modify my circuit so that I can connect a wire to a GPIO pin of an AT89C2051 so that it can detect when a battery is charged/still charging without negatively affecting the charging process.


  1. Could I change the variable voltage regulator (LM317) to an 8V fixed voltage regulator to charge the 7.2V battery with a resistor in series? If so, what resistor value?

I'm asking all this because I want to make a low-priced in-circuit battery charger for my microcontroller projects and I don't want to have to remove the battery in order to charge it.

attempted circuit

Why this question isn't duplicate

I want to build the charger circuit using discrete components that are commonly available in anyone's junk box. I don't want to have to purchase a special IC and wait weeks for it just in order to charge batteries. Additionally, I'm looking for a modification to this circuit to make my charger more efficient. I adapted parts of it from talkingelectronics.com.

  • \$\begingroup\$ The answer to the question in the link isn't what I'm looking for because I want to use simple discrete technology and I want the source power for charging to be a low voltage dc supply, not an ac mains \$\endgroup\$
    – user116345
    Nov 20, 2016 at 21:34
  • 1
    \$\begingroup\$ Then you haven't read it. \$\endgroup\$ Nov 21, 2016 at 0:21

1 Answer 1


I would personally look for a battery management IC and modern battery technology that is not an environmental disaster (Cadmium)... But if you insist on doing NiCd, you have some things to watch out for:

  1. NiCd batteries can explode if over charged.
  2. NiCd batteries suffer from memory effect, they should be fully discharged then fully recharged otherwise their capacity is greatly reduced.
  3. Power dissipation in your regulation system, you will not be able to provide any large amount of power, nothing close to being able to charge a phone if that is what your plan is. Assuming even the lowest 500mA charge rate of a phone, your 5V regulator will be dissipating 1.5W. Move upto the 1.5A of a modern smart phone and your at 4.5W, even with a large heat-sink that will be hard to manage.

Continuing on, assuming a full open circuit charge of 1.3V, you should plan to charge the batteries to 7.8V or 8V as you suggested. Monitoring the charge will be challenging without some way to interrupt the charge and check what the floating voltage is.

Also two items about your circuit, I did not spend much investigation but your variable voltage regulator seems overly complex with the addition of Q1 etc. Second, I would choose a low drop diode for D2, like a schottky, otherwise your 5V will look more like 4.3V and not power any USB based device (USB spec is 5V 5%).

Instead of redesigning the wheel with really old technology, you should look into how a modern Lipo based system works with a boost or buck converter. You could still build it yourself and it would be efficient enough to charge a smart phone.

  • \$\begingroup\$ I considered a variable regulator to convert a higher supply voltage (9VDC) to 7.2VDC so I don't overload the battery, then the second regulator converts to 5VDC. 4.3V may be fine with me since I'm actually powering an AT89C4051-24PU and its lowest requirement is 4V. I don't have immediate access to new technology IC's. I want to use what's in my junkbox to make one. \$\endgroup\$
    – user116345
    Nov 21, 2016 at 0:11
  • \$\begingroup\$ @Mike, if this is just for a low power device, it should work ok... Only last issue I see is that the batteries could charge unevenly, especially if the packs are not identical. You may want to add some sort of an over voltage protection to each battery pack, like a zener diode... \$\endgroup\$
    – MadHatter
    Nov 21, 2016 at 0:19
  • \$\begingroup\$ The packs are 100% identical \$\endgroup\$
    – user116345
    Nov 21, 2016 at 2:35