I have a strip of 3VDC (two AA in seires) christmas LED lights. 10pcs in one strip. I've been using a 18650 (4VDC), to make them shine brighter; after one week first half of them dims down from bright to low-bright when turned on. The other half stays bright all the way. What tha led is going on?

  • 1
    one or more LEDs on the half-circuit ladder is damaged, allowing more current to flow through it than designed, detracting from the current delivered to the other LEDs. if the strip was deigned for high-resistance 3.0v driving (like from AA batteries, which internally limits the current compared to 18650s), it likely can't be fixed; it's blown.
    – dandavis
    Dec 30, 2017 at 19:26
  • dumb question, but are you regularly unhooking the 18650 and putting it back on the charger to recharge? Also do these 18650s have a battery protection circuit on them? Dec 30, 2017 at 22:25
  • @Harper - yeah, no Dec 31, 2017 at 20:06

1 Answer 1


An arc discharge light, like a fluorescent or metal halide, is a dead short that happens to make light. (It is an arc, after all). They would flow infinite current if you let them, which would burn then up fast. You stop that by putting a current-limiting ballast on them. That's what a ballast is, it's a current regulator.

LEDs aren't quite that bad. But they are current-hungry, and must be current limited. If you are driving a raw LED, a small increase in voltage will cause a big increase in current. Their voltage-current curve is steep.

The best LED-based appliances use an active driver circuit which actively limits current. These active drivers are "magical"; they can input a wide band of voltages from 80 to 306 volts AC and outlut exactly the current the LEDs want.

Low-voltage consumer gadgets like Christmas lights often use resistors as current limiters. Resistors are dumb, and not magical at all. They don't regulate current, they just make the voltage-current curve somewhat less steep, which is "good enough" for working with one expected voltage. They also overspec the LEDs somewhat.

You raised the voltage to 125% of normal.

  • If you did that with a linear load, that would raise current to 125% and thus total power to 156% of normal (56% more).
  • If you had done with with a highly non-linear bare LED, check the data sheet, but it would raise current to about 300% of normal, and power to 375% and kaboom.
  • The resistor eases this somewhat and gives us something in between, probably in the 200-300% power range.

So you can see your "little voltage bump" is really putting the LEDs on the griller. You are leaning heavily on whatever overdesign was built into the LED spec.

The bigger question is not why this failed, but why the other one is still working! You seem to have made yourself into a "QA department" for these lights and are discovering the ones which were not overbuilt.

  • well, the best ones (pro gear @ >$100/m) actually use a constant voltage supply and under-driven modules that avoid heat+current crawl. If you want the most light output, CC is indeed the way to go because you can push LEDs precisely up to max on that steep V/I curve.
    – dandavis
    Dec 31, 2017 at 22:39

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.