Supercap Lights Your Way In Times Of Need

You won’t find [Antoine] stumbling around in the dark. He just finished working on this LED flashlight which draws power from a super-capacitor (translated). He realized that lighting a high-efficiency LED takes so little power that there are many benefits in play when deciding to move away from batteries. When compared to a super capacitor, batteries have a shorter life span, are heavier, and take up more space.

The biggest drawback of a super capacitor in this situation is the low voltage operation. The output will start at 2.7V and drop as the current is discharged. [Antoine] used one of our favorite simple circuits to overcome this issue, the Joule Thief. That circuit is commonly seen paired with an LED in order to boost input voltage to a usable level. That’s precisely what’s going on here.

The final hack in his circuit is the addition of that red LED which you can see in the middle of the board. This takes the place of a Zener diode and drops the charging voltage to a safe level. That indicator light will not come on until the cap is fully topped off. This way it tells you when the device is done charging.

46 thoughts on “Supercap Lights Your Way In Times Of Need

    1. Doesn’t matter. A small 10F capacitor will power an equally small LED happily for a good 20 minutes.

      That’s great for a small pocket light that you can just forget in your pocket. When you need it, you plug it in a car charger or a laptop for a quick top-up and you get 20 minutes of light.

      A battery powered light would have either dead cells after years of neglect, or require hours of recharging for a simple job like changing a flat on the roadside.

      I used to have one of those keychain lights that you plug in the car, but after I forgot to recharge it for a long time, it simply died. The cells went dead.

      1. Nope. Lithium Cell AA Batteries in my bug out Bag. they have a 10 year shelf life. and they will run the LED flashlight they are for for well over 100 hours.

        Until “supercaps” can do that, they are no replacement for batteries.

        for emergency use, a supercap is worthless. For a handy flashlight that lasts a short time that hangs around in the light socket always ready to be used? Yes. They are handy for day to day short use.

      2. I love google translate….

        “Check the first few uses, the burden of your capacitor because it may explode and I think a super capacitor which should make it fart damage …”

        Any circuit that has the potential of “fart damage” is a solid win in my book.

    2. That’s not wholly true, Li-ion batteries may be pushed aside by supercapacitors in the very near future. While the energy density of a Li-ion battery is 160 W*h/kg and the energy density of the type of EDLC shown here is 6 W*h/kg, MIT LEES has produced a super cap with an energy density of 85 W*h/kg and EEStor claimed to have produced energy densities up to 400 W*h/kg. EEStor seems to have evaporated from existence though with a statement from the company being declared imminent for the last 4 years so I’d severely doubt any of their claims. Still while EDLCs may not yet have the same energy storage as Li-ion batteries they are quickly encroaching on the territory. Additionally they are considerably cheaper and the Capa-bus in Shanghai has demonstrated they are ready to handle some of the large power requirements of today.

  1. I’m sure someone who know more than I will correct me, but I thought that the big advantage of super caps was that they could be charged almost instantly (at least compared to any other battery.) In which case making it USB charging seems to be completely missing the point.

    But I don’t really care. Powering a single LED is not an engineering challenge and the project has a few nifty features.

    1. Simplifying a little, 100 mA at 5 volts is .5 Watts, and the capacitor holds 36 joules, which means that a full charge can be achieved in about a minute.

      Of course some of the energy is wasted on the series resistance, but that gives you a figure of scale for how long it takes to charge it up.

    2. Since he’s provided the exact values, we can calculate that the average current will be 65 mA and the average voltage over the capacitor will be 1.35 volts, so, the charge time should be approximately 7 minutes.

      1. tl;dr? “The charging time is slowed, but we can not help it if we want to use USB as a power source. The lamp will therefore now charge in 1 minute (which is very interesting)”

      2. That can’t be right, because the current through the capacitor is at maximum 5V/56Ohm = 89 mA, but when the capacitor has, say 2 volts already in it, the current is reduced to 3V/56Ohm = 53 mA.

        At the beginning of the charge, there’s 0 volts over the capacitor, and therefore no power going into it. The efficiency of the charge starts from (arbitrarily close to) zero and picks up as the voltage of the cap increases, but that also decreases the available power because the voltage across the resistor drops which decreases the current, which slows down the charging.

        The only way you can get the calculated 36 Joules into the capacitor in about a minute is by assuming a constant 5 volts x 89 mA = 0.445 Watt across the capacitor, which yields a charge time of 80 seconds, but this simply doesn’t happen.

        Which is why I don’t believe it charges in a minute.

      3. Or we could calculate the RC constant, which is 560 seconds for 10 F and 56 ohms. That’s the time to reach 63.21% of the 5 volts, and it would reach 2.5 volts in roughly 2/3 the time, or 6-7 minutes.

    1. > Someone correct *ME* if I’m wrong, but that zener diode seems to be connected in the wrong way…

      “The diode is in the right direction. She is busy from v 2.4. It therefore protects the capacitor.”

      1. He’s wrong, the zener is connected backwards. When polarized directly a zener behaves as a normal silicon diode which has a voltage loss around 0.6-0.7 volts, depending on the current that flows into it due to it’s non linearity. To get the zener effect you have to connect it backwards: cathode to the positive and anode to ground. think of a zener as two diodes connected in antiparallel.

  2. Were he to tune the circuit to resonance by testing with a variable resistor instead of just using the 1k, he could lower current draw for the same lumens by increasing frequency… effectively making his supercap charge last longer. A bit of attention to the toriod as well… iron powder in the right band to make a high Q resonant circuit, and it can be improved even more:)

  3. Would there be any issues with excessive inrush current on the supercap when you stick it in a USB socket? Is there a risk of straining the USB host?

    If not then this is a nice design because hackers generally aren’t far from a USB socket.

  4. Thanks Hack-a-day. This is precisely what I’ve been looking for the past few days. Now I just need to scale it up.

    And BTW: This is a great mod for those Solar Pathway lights they sell in Home Depot or Walmart that die after a year. With the Cap in place of the battery, they will work much longer.

  5. Lurker here. Maybe someone can enlighten me, but the semiconductors in use here have finite resistance, so wouldn’t the capacitor slowly discharge over time? It might be negligible, but some of the posters above make it sound like it’ll never go dead.

    1. i don’t ‘think’ they meant it’ll never go dead, just that the storage cell is less likely to fail ie. cease to function. It may loose its charge, but it will still be capable of receiving a charge….

      I like it

  6. Only if the cell phone manufacturers would put in a extra LED to help light our way sometimes, often using the display isn’t enough. Call me spend thrift A 5 dollar bill, and a nickle buys a key fob light from Walmart, they seem to last me a year. For me it wouldn’t matter how much during the day I spend near a USB port, if I forget to take off the key ring, and charge it. Nice project, but stuff this inside a functional USB drive I’d really be impressed. One more less daily carry item, not that carry my portable apps drive all that often.

      1. But when the battery dies after abusing the flashLED as your flashlight, you’re sure going to be waiting a lot longer to have that flashlight back. And now you’re out your phone for a time too!
        ‘Course, if you’re using that function of your handheld for long enough to drain it’s batteries, you’ve probably got other things to worry about anyhow.

    1. Done. This is really more practical with a pair of NiCd AAA cells on a thumb drive, as mine is. The build I’m running glows dimly for months, or gives decent light output for about two hours.

  7. @ NOLKK… Any phone with a built in camera flash typically has some method of turning on the flash as a flashlight.

    Hell, my POS work flip phone has a button that turns on speakerphone, but whenheld enables the flash as a flashlight.

    Its a 3 year old phone.

    1. You’re lucky. How about those über-modern phones that intentionally have no buttons, at least not very important ones, and require both hands plus eyes to operate? :( I was lucky, too, in buying a HTC TyTN just before they became unavailable, and still didn’t found a decent replacement for it.

  8. My main experience with supercaps has been in GPS modules where they are a constant source of problems and need to be replaced with batteries to get the things working reliably.

    Seems to me that if they were the wonder alternative that some people are making out that they would be in a lot more use already.

  9. The Joule Thief can operate to low voltage, but it is not very effective (: Maybe it would be much better idea to use some charge pump like ICL7660, because they can be much better in efficiency terms. Also, no inductors :D
    The circuit is somewhat non-standard, it goes like that:
    Pin 3 – supply –
    Pin 8 – supply +
    Pin 5 – LED cathode
    Pin 2 – C1 +
    Pin 4 – C1 –
    LED anode to +
    C1 is 47-200 uF cap needed for 7660 operation.
    I cannot test it right now because I don’t have a supercap, would be happy if someone shares results. You can make a flasher as well – connect 10n cap across pin 7 and supply +.


  10. Just a thought, since the red led seems to be the only “protection” from overcharge. Couldn’t you put a micro relay in there that activates instead of a light? That would open the circuit to prevent overcharge. Just a thought, I’m not much of an electronics hacker these days but I’m just thinking from my old rat-shack 500 project electronic kit days and it seems like a good way to cut the juice before popping the cap.

  11. Hi, i was wondering how you would use one of those “caps” as a batterie? How would you get it to keep the energy stored till it is ready to use?… i have sooo many questions. I dont know much about all this but thats why im here is to learn how to build cool stuff like this. Thanks for what ever info you guys reply.

  12. Lithium cell life has so many variables it is difficult to predict but I feel it is safe to say a 20% reduction in capacity at 2000 cycles is usually achievable. Supercap cycles are measured in millions.

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.