Giving A Crank Flashlight A Super Capacitor Overhaul


[Caleb] was given a tiny LED flashlight which has a crank used to charge it. Unfortunately it wasn’t holding a charge, and constant cranking didn’t work very well either. He cracked it open to find a single lithium button cell. Instead of using a drop-in replacement he soldered in his own super capacitor.

The stock device is remarkably simple. It uses a standard DC motor as the generator. It’s connected to the crank using a set of gears, with the two red wires seen above connecting it to the control board. Four diodes make up a bridge rectified and apparently feed directly into the battery. No wonder that cell went kaput!

But this orientation isn’t bad for using capacitors. They can be charged directly and the switch which attaches the LEDs to voltage doesn’t interfere with their operation. The last problem was making room for them in the case. [Caleb] considered a few different approaches, but ended up just heating the plastic enclosure until it could be deformed to make room for the additional parts.

25 thoughts on “Giving A Crank Flashlight A Super Capacitor Overhaul

  1. What makes a capacitor by definition a supercapacitor. Is there a difference in materials or construction? Is a regular capacitor considered a supercapacitor depending on how it is utilized (storage)?

      1. Supercaps/Ultracaps are usually made of materials that make them in some ways a hybrid between rechargeable batteries and capacitors. As a result they have much higher energy density than most caps – but much lower voltage ratings. Most supercaps are rated 2.5V, and anything rated more is multiple units in series.

      1. I just spin it until it seems bright enough. LEDs give thousands of hours of life and if I’m over driving them and only get hundreds that’s fine. Also the plastic gears will most likely give out before anything else.

      2. I would put an LED in series with (a lower voltage) zener to provide the voltage limit to the super caps; that way, I have a handy indicator that my cranking isn’t putting any more charge into the caps.

  2. I suggest the next upgrade to be schottky diodes as the ones in all those cheap wind-up torches have a voltage drop of around 0.6v whereas you can get schottky types with a voltage drop of only 0.15v

    1. If it’s a custom setup for personal use, there’s a pretty good chance you always spin it the same direction anyway. Make a note of that direction and skip the rectifier completely – the motor produces DC naturally.

  3. Well anything that improves the quality of these cheapo chinese crank lights is a good thing, and this does that job well, but it also points the way to upgrading something more expensive/important, like an emergency crank radio.

  4. Seeing this hack gave me an excuse to tinker with an old wind-up torch of mine with a dead LIR2032 cell inside.

    I did away with the original circuitboard and made an AC-DC bridge with 0.15v drop schottky diodes connected to the motor, the output of which going to a DC-DC boost converter set to 14v, then hooked that into a 20v-24v 1 farad capacitor (the huge type used for helping the amplifiers in car audio systems).

    I initially connected up a multimeter to see how far the voltage would go before trying an LED or two with appropriate resistors.

    Unfortunately the gears were already a little worn out and they just got hot and slowly seized up so I only got as far as 6.5v. I have another identical torch but don’t wish to tinker with it as the battery works and I don’t want to ruin the gears.

    Are there any wind-up torches out there NOT made out of plastic gears?

    1. Side note: The 1 farad capacitor has a horrible self-discharge rate that would make it useless as a power storage source for a wind-up light, not to mention the sheer size and weight of the thing!
      Using some LSD (low self discharge) NiMH cells would be far more appropriate for such a task as they hold their energy for months.

    2. Windup lamps are mostly a novelty item, anything that mechanically complex isn’t reliable enough for emergencies, also at the rather crap light levels you get from hand power you could simply remove the gears and fill the cavity with cheap single use lithium batteries at the same materials cost and have a lamp that will run constantly for weeks and a shelf life of 10 years.

      I guess it would be possible to make one with all proper metal gears, a good modern generator, proper batteries, proper charging circuit etc. but it would cost the same as a bicycle…

      If you want something for an actual end-of-civilization scenario you probably don’t want batteries in it at all, or gears or any other mechanical parts. I would probably go for an array of supercapacitors, a mid-sized solar panel, proper charging circuits to protect the capacitors from overvoltage, some kind of magnetic or touch-controllers to avoid mechanical switches, and then seal it hermetically in a very sturdy case.

      1. You’re absolutely right about the novelty aspect of these cheap wind-up torches as I mostly buy them out of curiosity and they’re only a quid or two, but the one torch I carry with me everywhere clipped to a belt loop is a Photon Freedom Micro, I’ve been buying & using their keychain lights for 9 years now and swear by them:

        Of course solar panels are one of the best sources of non-mechanical ‘endless’ electricty, I have small Kyocera 30 watt panel, it looks like it’s only a year or two old but was made in 1983! I tested it a few years back and in full summer sunlight it still generates 24 watts.

        But that still doesn’t stop me wanting a decent dynamo-charged torch (maybe with a USB charging port too) which uses metal bearings/gears and a proper dynamo not a ten cent motor.

        1. Solar panel from 1983, did they even make them back then! I’m joking of course. I’m surprised your getting that much power out of them after that long. I thought they lost ~30% after 10 years. Going on your end of the world scenario I would have a beehive and steal honey and wax from time to time. Also trees burn and make light :)

  5. What about just storing the energy in a long persistence phosphorescent material, such as SrAlEuO5 and using NUV LEDs with 1K parallel resistors to foil EMP and motor spikes?

    If heated this gets brighter, so you could dump the extra power from the motor into a heater for those times you need more light,

    Also worth mentioning, I also tried the supercap method with my “squeeze light” and noticed the horrible self discharge.
    The fix is to “train” the capacitor by holding a stable DC voltage on it for several hours which helps a lot.
    The typical supercap works best at 2.50V and any higher just makes it die faster.
    So adding a crowbar to stop overvoltage would be useful, or switch between multiple supercaps for more power storage.

  6. Cost as much as a bicycle, no just as much as a bicycle dynamo light setup. I don’t bother to turn mine off during the day.
    I would like a very heavy spring unit that has to be stepped on with 100 pounds or so and several inches motion but would run for many minutes or hours.

  7. I have one of these. Just dug it up from the parts bin and removed the cell. No markings on it at all, not even under the black shrink wrap. Since it never held any charge no matter how long you kept winding, I just assumed it was a primary (non-rechargeable) button cell.

  8. I actually recently bought a few of those to try make some fun wind up electronics projects, they actually do recharge when you spin the handle hehe :) found them on dealextreme, they also have similair gadgets (cheaper and expensiver wind up stuff) that may be of interest.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.