A Supercapacitor Might Just Light Your Way One Day

Sometimes the simplest hacks are the most useful ones, and they don’t come much simpler than the little supercapacitor LED flashlight from serial maker of cool stuff [Jeremy S. Cook]. Little more than an LED, a supercapacitor, USB plug, and couple of resistors, it makes a neat little flashlight that charges from any USB A power socket and delivers usable light for over half an hour.

It’s neat, but on its own there’s not much to detain the reader until it is revealed as a “Hello World” supercapacitor project from an article in which he delves into the possibilities of these still rather exotic components. Its point is to explore their different properties when compared to a battery, for example a linear voltage drop in contrast to the sharp drop-off of a chemical cell. In the video below the break we see him try a little boost regulator to deliver a constant voltage, with consequent severe loss of lighting time for the LED. It’s by this type of experimentation that we learn our way around a component unfamiliar to us, and the article and video are certainly worth a look if you’ve never used a supercapacitor before.

30 thoughts on “A Supercapacitor Might Just Light Your Way One Day

  1. Neato,
    I’ve been using a set of six off 3000F supercars (2.7v max each in series) to start my 2.2L 4 cylinder Toyota camry for many years now, works a treat. They seem to self balance just fine when alternator charging. Only minor issue is some self discharge (incl radio memory backup current bleed) over about 10 days if not started, so add a solar panel & simple over volt regulation. Thanks for post :-)

    1. Even at surplus prices those sound a lot more expensive than just a normal car battery…. what was the motivation?

      If you’re using the capacitors I think you are (Maxwell?)… Did you need a high-wattage ballast resistor to deal with their extremely low internal resistance, or was the Camry’s charging system able to handle it? They act like a short in your charging circuit and it’s pretty easy to fry regulators and melt fusible links….

      1. I have also considered doing this on my motorcycle. For me, the motivation was mainly “because i can”, but nice side-effects include: capacitors weigh a fair bit less than lead, the engine will crank better and the capacitors aren’t as effected by cold weather. Downsides are mainly the self-discharge, but using a small lithium battery they can be charged in a matter of minutes. A single 18650 battery can easily hold 40 kJ, so get a couple of those and a small constant-amp powersupply and you can charge it from 0v several times without issues.

        The charging-circuit in all consumer-grade vehicles i have worked with probably wouldn’t like a shortcircuit to 0v, but regularly need to supply their rated capacity.
        A healthy lead-acid battery that has lost some charge because the lights were on or the radio was playing etc. will also overload the alternator and regulator for several minutes after starting the car, and if it’s a diesel-engine it might be pulling 100% for over 5 minutes due to a more discharged battery due to glow-plugs and high compression needing bigger starter, and the glowplugs cycling on and off until some heat has build up in the engine. In my late-90’s 1.9 TDI the headlights “flicker” on and slightly-less-on (As the voltage drops from high 13 to mid/high 12) every 5-10 seconds or so the first couple of minutes in the winter. This is especially noticeable if using the high-beams.

        6x 3000F 2.7v capacitors will hold about 66 kJ of energy, or the equivalent energy of 90A@12V for 60 seconds if you charged them all the way from 0 to 16.2v.

        Charging the battery from let’s say 10v(After cranking) to 14v(Fully charged) will only take approx. 30 seconds with 45A, which most alternators can easily deliver, and in the real world it will probably go faster because the alternator will deliver closer to 90A or more depending on the alternator and what else needs power.

        I would be careful of connecting jumpercables to a completely dead capacitor-bank though! In those cases i would probably use a constant amperage supply and just run it off the other 12v battery/car. You only need a couple of minutes anyway

  2. Supercapacitors do have some nice advantages over batteries, primarily their very low charge/discharge losses, and their very high power capabilities, supercaps also don’t really suffer any damage from repeated recharges. Downside with supercaps compared to batteries is that supercaps have a lot lower energy density.

    Though, a hybrid solution can work fine in certain applications, ie have a battery in parallel with the caps. Then one gets good energy density from the battery, and good short term power delivery from the caps. (Though, in parallel can be as simple as them being connected with a pair of wires, to as complicated as them being two separate power delivery systems with their own control electronics running the show…)

    Another disadvantage with capacitors in general is their MTBF value, but with adequate deratings we can rather easily get into many years of expected life time. Though, most supercaps tends to have fairly low temperature ratings while also having low MTBF values at that low temperature.

    For an example out of digikey’s currently 1683 super caps, only 104 have better MTBF/temperature specs that 2000 hours at 65C, and all the ones that do beat that tend to be more expensive for the amount of energy stored, so in my own opinion, for super caps to really take off as a battery replacement in low energy applications, then the bog standard supercap needs to get at least a few times better MTBF specs.

    Since currently, even within areas where one mostly is interested in its high power delivery, one can’t really trust that it will survive long term due to its fairly low life expectancy.

      1. Well, 65C isn’t really a temperature one would run supercaps at either, ideally speaking.

        The temperature rating is though an important factor when it comes to derating capacitors and work out their expected life at a lower temperature. As a rough rule of thumb, the NTBF value doubles with every 10 degree C decrement in the working temperature. (Ie, a 2000 hour @ 65C cap would have an MTBF value of around 16000 hours at 35C. (about 1.8 years or 24/7 operation.))

        This spec is also important when it comes to electronics enclosures that at times can be poorly ventilated and reach rather high internal temperatures, like an ambient environment around 40-60C isn’t too uncommon, especially if the sun is allowed to shine onto the enclosure. Here one would primarily look for capacitors that are above the highest expected temperature within the enclosure.

        Though, the temperature part of the spec can also at times be related to where it permanently gets damaged due to heat. (like the electrolyte boiling off and such… And yes, this can vary depending on charge state, so storage temperatures can still be higher.)

        But how a lithium cell performs at high heat is a good question.

        I know that chemical reactions practically always react faster at higher temperatures, thereby lowering the ESR of the battery, but at some point, the heat would start to damage the battery instead.

        But in the end, the MTBF value of most batteries is rather unimportant, since it reaches many years. (usually the shelf life of the battery.) So the more important spec is degradation over a number of charge/discharge cycles. (Digikey and RS components for an example doesn’t even list the MTBF value of rechargeable batteries… Mouser at least states maximum operating temperature, but no MTBF.)

    1. MTBF specs for electrolytic capacitors on my 20 years old toyota ECU was 2000h.

      My current(2015)mazda battery is lead acid + supercaps hybrid battery, as every start-stop car battery is. Pure led battery lasts about 6-12 months under constant starting current punishment.

      My 5 years old wireless weather station uses supercaps and solar paner as power source, looking at deterioration its another 5-10 years before supercaps won’t last trough the cloudy day spell I set for a target for it. In 5 years never missed scheduled reading transmission due to power loss.

      You are reading MTBF wrong.
      a) only one in 100 000-1 000 000 will fail after MTBF rated time
      b) MTBF is determined for maximum electrical and mechanical ratings.
      You would have to torture supercaps with maximum rated g for r vibrations while charging and discharging at maximum current at maximum temperature to hit MTBF rated failure times…
      In real life industrial applications(16-24/7 operation) capacitors with MTBF of 3000h fail rarely after 10-20 years of use.

      1. There is a big difference between electrolytic capacitors, and super capacitors, even if the later is also electrolytic.

        My concern weren’t purely about it being 2000h, but rather 2000h at 65C.

        Most regular caps are first of better than 2000h, and secondly, rarely rated down at 65C. 85C and 105C are far more common. Even 125C is not all that uncommon. A temperature where 2K+h ratings are the norm. So 2000h at 65C is looking rather pathetic in comparison in terms of reliability.

        Also, MTBF stands for “Mean time between failures”, and this is an average.
        In other words. if you have 10000 capacitors rated for 2000h at 65C. And keep them fully charged, stored at 65C, then after 2000 hours, you should expect roughly 5000 of the capacitors to have fallen outside of spec.
        So your statement that: “only one in 100 000-1 000 000 will fail after MTBF rated time” is factually incorrect, if we are strictly talking about what MTBF is, if we are talking about expected life time, then we should also talk about derating.

        Since as stated in my prior comment: “with adequate deratings we can rather easily get into many years of expected life time.”

        Derating a component by not running it at maximum ratings is best practice and common. (in some applications, derating will be fairly inherent, like the drive train of a car…) But we can calculate an expected life time for these derated figures.

        For electrolytic capacitors, working temperature tends to have the largest effect on expected life time, this being that the expected life time doubles with every 10 degree C decrement. Undervolting the capacitor has an impact too, but generally not to the same degree.

        Vibrations can be isolated and tends to be a mounting related issue, and can thereby not easily be included in the MTBF spec of the capacitor. (Soldering wires to a capacitor’s terminals to provide some mechanical isolation isn’t uncommon for applications where shock and vibrations are to be expected and high reliability is key.) But vibrations and shock doesn’t tend to be included in the MTBF value of electrolytic capacitors. (ceramic multilayered ones is another story.)

        Though, this doesn’t change the fact that supercaps are still having rather low MTBF values, and needs to be heavily derated to be expected to last for any appreciable amount of time, making them rather unsuitable in a large verity of applications where they could have otherwise have competed with batteries.

        Then the next thing, we aren’t always talking about catastrophic failure, just that it has exited spec.
        This can be 20% lower capacity then the label, or 5x more ESR, or the like. In some applications it doesn’t really matter much, and for others, it is a real deal breaker.

        We can on the other hand expect supercaps with higher MTBF values in the future, since it isn’t an unlogical development. And we can also expect to see more applications use super caps instead of batteries or a hybrid solution of both, since capacitors have a slew of advantages over batteries.

        But in the end, it doesn’t change my opinion that one of the main things holding super caps back is their rather low temperature ratings and therein low MTBF value, that currently make them rather unsuitable for higher temperature working environments.

        2000h@65C isn’t impressive. 1000h@85C is a lot better in more than one way.

        1. Still better MTBF than many consumer batteries.

          I would say what is holding super capcitors back is the current enrgy density, it is hard to compete with petrol or modern lithum batteries.

          Regardless though, the ability to absorb energy rapidly works in supercarps failure, so even know they have a place as an “electronic flywheel” part of a battery, or memory holdover unlikey to leak like most batteris can at the 5-10 year mark, all over your mobo or similar.

    2. Hi guys, Grant here from Cape Town South Africa. I’m not a tech geek like you guys but here’s a thing. I have 6x 12v 1kw/h supercaps on my boat. Not any form of hybrid cap/chemical battery. These are supercaps. Charge loss is +/- 1% per month. They CAN be charged in less than 7 secs and CAN be discharged in 3 secs. They can operate at -50°C to +80°C. They have a 1000000 cycle life span. No chemicals, no fire risk and fully compostable. No memory effect like chemical batteries. I am charging them with 2x 270w solar panels. 2 caps in parallel for 12v sys and 4 in series through a 5kw inverter for the 220v sys. I have even run my 200A welder on the boat powered by supercaps. They are 99% efficient. This really does make solar power viable. No heat no mess, fuss, safe and no dangerous chemicals. They just work. There are also 0.465kw/h, 3.5kw/h. 7.2kw/h and 10kw/h units available also a AA1.5v size with an all in one usb plug.

  3. Well said, such projects are Hello World of supercap projects. I also played with supercaps and LEDs and my opinion is they’re good for hand-crank flashlights, but for everything else LiIon is far superior. Take Nitecore Tube for example, miniature and inexpensive keychain flashlight that will give you much more light for much longer than any supercap light. Supercap energy density is simply much lower than LiIon’s which makes them insuitable for flashlights.

    1. I’d contend that for some use cases supercapacitors are more suitable than lithium cells. For example, if your power output is relatively low but you’d like to recharge relatively quickly and have a simpler charging regime. This is the case for a torch I made for my children. The first version was based on a wireless charging circuit and lithium cell but the second version based on a supercapacitor battery is a much better match for our requirements.

      1. Sure, when you need to store small amount of energy quickly supercaps are the way to go. Like hand-crank torch I mentioned or toy torch you described. In case you need to store more energy but don’t mind (relatively) long charge time then liions are the better choice.

  4. It’s a useful project to understand supercapacitors, but why use a 5C supercap when you could use a 700C 2032 rechargable battery? Yeah, supercaps charge fast and could be used say as a UPS to shutdown a raspberry pi, but even in that application, seems like batteries are cheaper.

  5. Guess who bought a leading UltraCapacitor manufacturer: https://www.maxwell.com/ ??? Tesla! Generator start batteries are maintained constantly with a charge circuit until the power goes out. Then the UltraCap starts the generator.

    Some other sample uses of supercaps to speak to the points above… A lot of DashCams…. that live in very hot automobile interiors that kill batteries. The cap lets the last write happen to the memory card during power down.

    The model train hobbyists have been using supercaps on locomotives to get through dirty tracks and other continuity issues. A charge limiting resistor is required so the wheels don’t get welded to the tracks during recharge.

  6. I’ve done this same thing but added a LM317 adjustable regulator IC.
    I set it up for current regulation mode in line with the LED.
    This evened out the discharge curve and made the LED last longer.
    It also evened out the LED brightness.

Leave a Reply

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