A More Powerful Boost Charger

[Meseta] built a powerful boost charger to top off his portable devices. He was inspired by the Minty Boost but wanted to overcome the rather limited capacity provided by the two-celled product. He ended up building his own DC to DC boost circuit rather than using an all-in-one IC. As you can see, the result uses four 18650 lithium-ion cells, normally found in laptop batteries, and can power two USB devices at the same time.

28 thoughts on “A More Powerful Boost Charger

  1. Interesting approach…but I have a few suggestions:

    First, he claims that monolithic chips designed for boost converters have very limited current output.
    Thats 100% false.

    Second if he is already using 4 cells why not connect them in series and use a buck converter. That generally a easier direction to go (lower switching current means smaller mosfet) and more common (meaning the chips are cheaper).

    Despite my gripes its still an interesting project and I like the hand built nature of it.

  2. I realize these sort of devices are considered Chargers. I always found that idea a bit foolish however, why bother charging one set of batteries, only to use them to charge a second set of batteries?

    You lose electricity to heat at both junctions, even more is lost if you have to massage the raw electric potential into a different / constant voltage source to keep the second charging circuit happy.

    Why not develop the peripheral devices to operate in an external battery mode, where operating current is drawn directly from an external battery pack for operation, but not wasted in charging the internal batteries?

    I also wonder just how hot the components in his portable charger get during refill / operation if he were to keep that case zipped up ^-^

  3. @smilr
    it isn’t very foolish if you need a device to run for 8 hours and its battery lasts five, and you are miles away from an electrical outlet.

    To do something like you suggest, you could try removing the battery from the device when the external power pack is attached, making it so that the device only draws power to run. That would probably have mixed results dependent on each individual device’s design.

  4. @Andrew: Probably due to the cost of the charging circuit/device. A single cell charger on DX is $7, whereas a 1-6 cell balance charger is $35. Granted the second will charge much faster.

  5. “Why not connect the batteries in series?”

    Because then you’d have to build a complex battery management and charging system that follows the voltage of the individual cells and balances them.

    When they’re all parallel, all you need to do to charge them is a constant current source, say 1 Amps, with a voltage limiter at 4.2 volts. The cells will balance themselves.

  6. Parallel or series, doing it either way without the appropriate safety circuitry is not reasonable. In fact, it is more difficult to do it safely in parallel–this is why you’ll notice that most RC battery packs tend to be wired in series before parallel (in which case you’d have to match the cells. I don’t even know how you’d do battery management in series if the cells are mismatched–that’s just ridiculous.

    You CANNOT safely or properly charge four cells in parallel in a one cell charger.

    The cells will have their voltages balanced in parallel, but their currents will not be. In fact, if one cell’s voltage drops considerably faster than the other cells, that cell will actually be CHARGED by the higher voltage cells (the easiest way to see what happens is to model the cells as arbitrarily slightly different voltage sources in series with arbitrarily slightly different internal resistances in series). You’ll note that in some cases current will flow backwards into a cell, which is NOT good, since you’re not safely limiting the charge current. Of course, this behavior also further reduces efficiency. In addition, as noted before, with most converters, you tend to get considerably more losses if you’re working with higher currents instead of higher voltages (mostly I2R losses).

    Really, look around you–do you see any (well-designed; i don’t mean cheap chinese led flashlights or something) products that use batteries in parallel? There’s a good reason for that.

  7. sorry, i meant:
    “. . .match the cells). I don’t even know how you’d do battery management in *PARALLEL* if the cells are mismatched–that’s just ridiculous. . .”

    This is because you’d have to sense the current into each cell, of course–you’d need current sense resistors for each cell, and then a current sense circuit and a current rather than a voltage balancer. If you’ve ever tried to play around with current sources, you’ll know that it’s harder to design good high efficiency current sources than it is to design voltage sources (I2R losses, mostly)

  8. I like the innovation, one thing I’d add to it would be a MAX1555 IC for each cell you could charge them up without needing to take them out of the unit. Though this would mean extra circuitry to isolate each cell whilst charging.

  9. “I don’t even know how you’d do battery management in series if the cells are mismatched–that’s just ridiculous.”

    There are ways. One is to shunt the current over a cell once a certain voltage is reached. Another is to have each cell charged separately. In a series configuration, you always have to play by the weakest link, and once you detect one cell going under the safety limit, it’s full stop for the entire battery.

    With sensible load currents, having the cells in parallel will not see any of them draining so fast that the others would dump uncontrollable amouts of current into them. Same thing with charging. Besides, lithiums do tolerate high currents both charging and discharging. It’s the excess voltage that they don’t like.

    If you’re charging a 4 cell pack limited at 0.5C, the worst thing that happens in the hypothetical situation that three cells disconnect completely, one cell would get 2C. It should be able to handle that.

    RC packs made with NiMh cells can take some punishment without dying immediately, even to a point where a cell could reverse its voltage and still come back after a recharge. Lithiums don’t work like that. If you step out of the safe zone, the cell is dead.

  10. “I don’t even know how you’d do battery management in *PARALLEL* if the cells are mismatched–that’s just ridiculous”

    They balance themselves. You don’t need fancy stuff.

    If one cell loses voltage quicker than the others, less current starts to flow through that particular cell and it drains slower.

  11. “There’s a good reason for that.”

    Yep. It’s called not having to step up voltage, because it’s cheaper that way. The downside is, that almost none of the products actually do any sort of battery management, because it’s cheaper that way, and they expect you to buy a new one in three years time anyhow.

    There are no well designed products on the market. They’re all compromizes, even the high-end brand ones.

  12. Of course, really cheap stuff usually doesn’t have battery management, and this is usually true. I was thinking more MP3 players, cellphones, etc. . .where power management is pretty important, and is the dominant compromise. You typically see a step down or two for the different voltages needed (1.8-ish, from 3.7v). You’re right about step ups typically being more expensive than step downs for the same power load, and that’s partly because you’re doing more current instead of voltage (so i2r losses in the mosfet & inductor, thicker inductor needed, and more expensive mosfet needed)

    It’s probably less of an issue with lithium-ions–perhaps they are better matched and better capable of handling those currents. I have never attempted to work with them in parallel–only in series with charge balancing. Certainly I’d imagine one battery charging the other would shorten the battery life (it’s essentially unnecessary charge cycling), and would make the single charge life lower because of the i2r losses in the battery while one battery charges the other. It’s definitely something I avoid (in general, too) but even more so for replaceable batteries when designing anything consumer oriented, because you know someone’s going to put a dead battery with a fresh one, and if they’re in parallel one’s going to charge the other one to death, and possibly start a fire. we try to avoid the lawsuits as much as we can ;)

    In the end, in this particular case, I don’t think anyone could make a very strong case that putting the cells in parallel with a step up is better than putting them in series with a step down.

  13. A thing to keep in mind is voltage

    Rechargeable NiMH AA batteries tend to output 1.2V. 4 in series would yield 4.8V, which might be too low for some devices, hence a Step-Up would be needed for any voltage range. This voltage would also lower over time.

    Li-Ion batteries output a nominal 3.6 volt/cell. So, you could do 2x series 2x parallel to achieve 7.2V, and the voltage probably wouldn’t drop below 5V before killing the batteries.

    My guess is this is why he decided to use a Step-Up system.

  14. @smilr

    It’s because device manufactureres first hate consumers, and second utterly hate people that modify things. No cellphone exists that has a external battery plug to do this. all have only a built in charging circuit that you must satisfy.

    If you could open up your cellphones and rip out the chargers so the terminals are direct ot battery then your idea is sound, but simply because of the utter hatred that device makers have for the people that buy their items we will always have no way of extending the run times of a device without wasting a lot of energy in the process.

  15. I don’t see the problem with making a portable device default to charging it’s own battery first, because that ensures you have a full charge on your internal battery when you disconnect from external power.

    However, once the internal battery is charged the device SHOULD go into a mode whereby the internal battery is merely maintained while power goes primarily towards operating the device.

    If the device DOESN’T do this then I really would have to wonder what the manufacturers were intending.

  16. Hey guys, many thanks for featuring this on Hack a Day. And it’s great to see all the constructive criticizms

    I would like to just defend some of my decision:

    To make things clear: this does not charge 18650 cells; it uses 18650 cells to charge other devices via USB cable.

    I use parallel cells because the circuit is designed so that it can work with a single cell (and hence why I have decided with step-up although I had initially designed the circuit based on the SEPIC topology)

    I have decided to not use battery balancing because I’m charging batteries externally using a dedicated charger (charge balancing is not as important for discharging). and as Einomies mentions, they balance themselves out (also Li-ion have a 80-90% charge efficiency vs. 66% of NiMH according to Wikipedia)

    @threepointone you will find that laptop batteries are wired in parallel before series).

    @Andrew: I agree with you! However, here’s what I actually said: ‘Unfortunately, because they’re “micropower”, they’re not designed to output higher power’.
    There are plenty of Monolithic DC/DC ICs that handle higher power outputs, and I had originally thought about using the LT1707, which will do 10A at 5V. But it costs $15, and doesn’t include any kind of low battery detection. And also, I wanted to make something with what I had in hand (which include low pin-count PICs and Transistors ripped from a defunct CD drive).

  17. “Certainly I’d imagine one battery charging the other would shorten the battery life (it’s essentially unnecessary charge cycling)”

    This is a problem with NiMh and NiCad chemistries. Lithiums don’t have significant self-discharge, which would cause one weak cell to drain the entire pack. The weak cell would be topped up, and then the current stops flowing. If it starts to truly leak anything and heat up, the internal fuse will blow and the cell disconnects.

    You control lithium cells by voltage, not by current like with NiMh. The voltage of a lithium cell is a very good indication of the state of charge. It’s not linear, but generally you can plot a curve where n Volts equals m Amp-hours remaining. The only thing that affects the voltage beyond that is the V=RI of the internal resistance, so the cell voltage reads consistently higher or lower depending on the current going in/out.

    This is much unlike NiMh which exhibits weird spikes and voltage drops depending on the current going in/out and the state of charge at which it happens. Knowing the voltage of a nickel cell doesn’t help, and the varying voltages mean that the currents keep circulating inside the pack.

    That’s the reason why parallel lithium pack tends to self-balance, because any cell that lags behind in voltage tends to get charged faster until it catches up, and any cell that gets full faster than the others will have its voltage up, and that diverts the current to the other cells. Likewise in discharge, weaker cells get drained slower and better cells get drained faster, leaving all of them in roughly the same state of charge despite the differences in capacity, so a parallel configuration is very suitable for lithium cells.

    Because lithiums have tight tolerances in acceptable voltage levels, they’re difficult to use in series configuration because you have to stop once the weakest cell runs out of voltage. All cells in the series are then essentially the same size as the weakest cell.

  18. “I don’t see the problem with making a portable device default to charging it’s own battery first, because that ensures you have a full charge on your internal battery when you disconnect from external power.

    However, once the internal battery is charged the device SHOULD go into a mode whereby the internal battery is merely maintained while power goes primarily towards operating the device.

    If the device DOESN’T do this then I really would have to wonder what the manufacturers were intending”

    I’m not sure what devices you’re specifically talking about,but everything I’ve ever owned did exactly that,charge the internal battery then run off the external power source.

  19. @ Ray: That’s kinda what I was saying, I just put it badly.

    “They should all do that and if they don’t they’re screwy for one reason or another.”

    ;)

  20. Re: the airport experience, I made three minty boost kits a few years back (so that my wife and son didn’t cockroach mine in-flight). I’ve gotten into the habit of throwing all of the carry-on e-toys into one bag for security to focus their attention on. Inevitably, they’ll begin to mention a desire to give it a closer look, and I’ll acknowledge that I’ve got a lot of interesting crap in there. If there’s to be security theater, I can at least make my part of the play go smoothly.

  21. @ threepointone

    most high capacity laptop batteries have cells wired in parallel, then those sets wired in series. The packs are monitored only by group, the series monitoring. The parallel cells are just left to get on with it.

  22. I’ve done plenty of stupid things with Li-Ions and i haven’t had one blow yet. Unless shorted or exposed to a naked flame i can’t see anything bad happening.

    Short circuits and naked flames are things that any battery may explode in, so he’ll be okay.

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.