I purchased a flashlight that came with 3x AAA cells. These are branded with the flashlight brand, labeled as Alkaline (Mercury and Cadmium free), LR03 AM4. Made in China. I am dubious of the Alkaline claim, but at least it's not a "Heavy Duty" cell.

What's odd is that the flashlight was working well, at full brightness. Then it dimmed. Some percussive adjustment it went back to normal, but eventually back to dimmed. Opened the flashlight, checked for corrosion, leakage/vented or a bad connection, there is none. Moved the batteries around, and suddenly back to normal. The slightest movement cause it to go back down.

Eventually I identified one of the cells as the issue. I could rotate it in place, and at some point it would work. This was the odd part. I couldn't get it to work normally again. Worse, I measured the battery voltage. -0.3V. Somehow, it has reversed it's polarity? A second set of these batteries from another flashlight shows one of three also having the same issue. -0.1V, while the other two show 1.49 and 1.52 volts. When on, this bad cell jumps to -0.56V when the circuit is pulling 5mA.

Obviously, this is a bad battery of questionable quality, but this is the first I've seen a battery go bad like this. What would allow a battery to fail and reverse itself? A normal chemical process? The other two higher voltage cells attempting to charge into a lower voltage cell (all in series)?

  • \$\begingroup\$ It's most likely a bad electrical connection that comes and goes as you move the battery. If you take the cell apart, you will probably find the bad connection / location. \$\endgroup\$ Jul 26, 2016 at 4:54
  • \$\begingroup\$ But that would result in reverse polarity? I mean, I've opened batteries before, they are a rod in a bunch of chemical electrolyte goop. Nothing that should swap connection. I'd be less confused if it was like an open or short, but 300 mV reversed? \$\endgroup\$
    – Passerby
    Jul 26, 2016 at 5:02
  • \$\begingroup\$ Is it an LED flashlight with a built-in constant-voltage boost converter? If so that would explain its ability to stay bright until the current gets limited. \$\endgroup\$
    – Transistor
    Jul 26, 2016 at 6:18
  • \$\begingroup\$ @transistor no, it has a three way push button switch and three different value resistors. \$\endgroup\$
    – Passerby
    Jul 26, 2016 at 17:58

1 Answer 1


There are several possibilities for the reversed cell, depending on the flashlight design.

The first is the flashlight is designed as you'd expect, takes power from the battery of 3xAAA as you'd expect, and works well from 4.5v to a very low voltage.

If you take 3 random batteries, even new ones, they will be slightly mismatched. If you connect them in series, and run them down, then one of them will be more 'run down' than the others. If you run them down far enough, the weakest will be reversed by the other two.

How far do you need to run them down? If the flashlight fails at 4v, chances are you don't get a reversal. If it fails at 2v, you probably will. If your problem that the flashlight keeps going and works too well at low voltage?

You may just have been unlucky, and got a bad mix of cells, 2 out of 2 is not out of the realms of possibility. Try again with a fresh set of cells.

The other (remote) possibility is that the flashlight taps off an intermediate battery connection for bias or something, and so hammers one cell harder than the others. You opened another flashlight, was the reversed battery in the same position? If yes, that supports this possibility. If not, then it refutes it.

With care, you can fashion a probe to measure cell current in situ, ideally with a scrap of copper clad PCB, but you might be able to make something with an old credit card, tape and some foil. See if the running currents from each cell are equal.

Spring battery holders are notorious for poor contacts, corrosion at the contact point between cell and spring, and moving the cell around in the holder is the 'standard' way to get the contact made again. This issue is completely independent of the cell reversal issue.

  • \$\begingroup\$ No center tap, and iirc not in the same position. These are new, barely used flashlights. And the fail rate is 2 out of 9. 25% failure. The issue is really the reversal of the battery. In situ, the reversed one actually went up to -0.5V. \$\endgroup\$
    – Passerby
    Jul 26, 2016 at 6:01
  • 2
    \$\begingroup\$ If we've removed the possibility of a rogue tap in the battery holder, and note that the failure is occasional, rather than every time, then yes, reversal of the weakest cell is a normal failure mode when the battery is discharged far enough. \$\endgroup\$
    – Neil_UK
    Jul 26, 2016 at 6:03
  • \$\begingroup\$ @Neil_UK -- Yes, but, for two lantern batteries 4S, discharged constant current 1mA into a white LED, for the four cells, two were completely discharged, and two were left fully charged (I'm trying to prove that now - I have yet to do the final discharge). Duplicated results with a second battery. Would you please look at my question? How to prove novel conduction path in weaker battery cells in two 4s-1p carbon zinc batteries discharged at 1mA from 6V to 3V? \$\endgroup\$ Apr 6, 2022 at 16:46

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.