Advice for charging a NiMh pcb battery

Hello all

I have a DS3231 RTC in my circuit design. I have not included the 200R and in4148 as usually included in the RTC modules, as this is apparently a bad way to charge the LIR2032.

I could use a standard 2032, but I would like to have a rechargeable option (yes, I know... the RTC will take years to run down the standard 2032).

The Vbat pin on a DS3231 can take 2.3v - 5.5v, so a 3.6v NiMh would seem suitable (and I have loads of them).

But, I cannot find a suitable circuit to run from 5v to charge the battery. I hunted around the RS site for battery management IC's, but they are biblically expensive (£8 ish) and not exactly what I need.

Anyone have an ideas or at least a basis on an approach?
Many thanks

I would just use a resistor to feed current into the battery. Based on your description from 5V to 3V6 and with negliable load maybe 1mA, so (5-3V6) / 0.001 = 1k4 Ohms, or thereabouts.

You can buy a rechargeable version of the 2032

When NiMh were in wide use, it seemed to be accepted that a permanent trickle charge of C/20 to C/50 was OK.

I thought the rechargeable version of the 2032 was the LIR2032... which is a Lipo?

I may be wrong, but I thought Lipo's were a bit more temperamental for charging that NiMh.
Also, mildly less chance of a fire event

You are correct, lithium batteries are very fussy about how you charge them. NiMh you can just trickle some current into them indefinitely. As your load is tiny you just need enough current to keep it charged.

NiCd were even more forgiving, but they are bad for the environment.

And tended to destroy the PCB after a coupe of years in use.

Thanks for the help. Throwing together a basic trickle charge circuit. That should be fine.

Sorry.... as a side question....

I have my Arduino feeding a MCP23017 multiplexer over I2C, which in turn uses it's outputs to switch NPN transistors through a 1k to the base.

The ground is obviously switched on when the MCP23017 pin is high.

Is it essential to have a 10k to pull that base low when off? I just don't have the room on my PCB.
Is the low signal from the multiplexer good enough to hold those transistors off?

Otherwise my PCB just ain't big enough.... that that then impacts other things.


No, for 2 reasons:

  • When the outputs of the MCP23017 go low they will pull the bases low through the resistors.
  • Bipolar transistors are current operated, with no current into the base they turn off anyway.

Thank you Perry. You have confirmed what I suspected

Additional information that I think might help:

Bipolar transistors have a low input resistance, they need current to be fed into the base for anything to happen, no current, nothing happens.

MOSFETs, which I think are confusing you, have a very, very high input resistance to the gate, so high you can consider it to be open circuit for most purposes. They do have a little bit of capacitance on the gate though. The result is that unless you actively pull charge out of the gate with something (such as the resistor you suggested) it will stay there and the MOSFET will not behave as you expect. Because of their high input resistance they pick up noise from the environment if left floating (not connected) and end up with some random voltage on the gate and an unpredictable output on the drain.

After many years of tinkering, I still have not grasped the basics :roll_eyes:

At 53, my tired, weary head reads these electronics maths books I buy and just turns to mush.
Next morning.... forgotten it all.

53 is just a "toddler" around here. Life starts at fifty. :grin:

(That makes me a late teenager, very occasionally I may imagine being such. :rofl:)

1 Like