I selected two non rechargable CR123A lithium-manganese battery to power my project due to their ability to operate under extreme temperature. A single battery is 3V with a rated capacity of 1550mAh. Since I need more capacity I decided to wire them in parallel so that the total capacity is 3100 but the voltage is the same.

Bonus question: Is it bad to wire them in parallel for a consumer product? :)

I read somewhere in this site that non-rechargable batteries might explode if someone inserts a fully charged one with a semi charged. I have not a safety circuit in there.

Now I need this to regulate them into 3.3V for my project.

Since I want efficiency two options exists:

  1. Boost converter
  2. Buck/Boost converter

Theoretically we want to go from the battery pack's 1.2-3V into 3.3Volts so we go up. A boost converter is the best choice since it can accept cut-off voltages down to 1.2Volts squeezing every little drop of energy from the batteries.

Practically I measured a battery bought from a local supplier and its voltage is 3.25Volts. What if a weird CR123A manufacturer makes a battery more than 3.3 Volts? Should I use a buck/boost converter and sucrifice the efficiency? Buck/Boost converter start from a Vin min=1.8 Volts and I would rather not use them if I could avoid them...

How should I solve this?

I could use a boost converter to regulate 3.3 from 3Volts but what if a user inserts a 3.5Volt CR123A bought from a weird chinese manufacturer from eBay?

Do I need a buck-boost converter to regulate 3.3V from rated 3V battery? I am afraid that the actual voltage in the batteries might be more than 3V in the beginning.

  • 1
    \$\begingroup\$ Never connect batteries in parallel. Small differences in discharge curve in individual batteries will lead to self-discharge. \$\endgroup\$ Commented Dec 28, 2016 at 18:07
  • \$\begingroup\$ you can easily join matched CR123A's with SMD fuses. Just verify Voc before assembly. <<1% is achieveable \$\endgroup\$ Commented Jan 3, 2017 at 21:18

2 Answers 2


I wouldn't worry about a battery exceeding 3.3 V. The voltage comes from the chemistry. Someone would have to use significantly different chemicals to make a higher voltage.

You might consider putting the batteries in series, then using a buck converter. By the time the cells get down to 1.65 V each, there is so little energy left that you're not really getting any advantage draining them to 1.2 V.


Parallel is fine if you use schottky diodes to OR them.

If the 0.25V Vf drop is too much for application, use ideal diodes such as LM66100 (see datasheet for proper configuration when using two of them).

With resepect to the voltages, you'd be much better off using components that work down to 2.5V than to use a boost (or buck) converter. With smart low-power design, you should be able to get sleep currents down to uA's. I do it all the time (with CR123's!). I do use boost regulators occasionally in those designs to accomadate white LEDs and such, but in sleep the processor is the only thing running, and it's tied directly to the battery.

Sorry for the necro post for those that are offended by that.

  • \$\begingroup\$ Well, I learned a new word: necroposting. \$\endgroup\$
    – P2000
    Commented Jul 10, 2020 at 4:20

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.