If I've got a device that uses a single battery, and with that single battery, the device lasts an hour, and I want the device to last for two hours, I'm assuming I can wire two batteries up and it should work. But how? Do I want to double the voltage (so serial) or double the current? (so parallel.)

Either way, if the device takes two batteries normally, and I want to double the running time, I do a combo of 2 parallel series of 2 serial batteries, right? But if I want to triple the run time, is that 3p2s or 2p3s?



When you connect two batteries in parallel, you are effectively halving your discharge rate while doubling your capacity -- effectively, doubling battery life, as you are intending to do.

Serial connection does not increase life, but rather, increases operating voltage, which some devices need. If your device has two serial batteries, then you need to connect another two serial batteries in parallel.

If they are not going to be recharged in the configuration you are putting them in, then you should be able to connect them in parallel. They will end up using a small amount of power by charging each other a small amount, but you will probably see more than a doubling of battery life.

If you are pulling a lot of power from a battery, you will find that its battery capacity decreases and causes excess current draw to decrease life and decreased capacity to decrease life. There are some excellent explanations about batteries online. The link I gave is more lithium focused, but the general knowledge applies to most batteries. This link is specifically about discharge rates and their effect on battery life.

Hope this helps. If you comment on anything that is not clear, I will edit and expand on it.

  • \$\begingroup\$ Glad to hear it, I really suggest you spend some time on battery university. \$\endgroup\$ – Kortuk May 17 '10 at 17:29
  • \$\begingroup\$ @Kortuk: "They will end up using a small amount of power charging each-other small amounts". What, they charge each other? \$\endgroup\$ – 0xakhil May 24 '11 at 11:36
  • \$\begingroup\$ @oxakhil, yes, if you just connect two batteries you can have differences in batteries and in connections cause small charge movements between the batteries. \$\endgroup\$ – Kortuk May 24 '11 at 14:56
  • \$\begingroup\$ It's absolutely false that "serial connection does not increase life". You get twice the energy whether you wire two cells in series or parallel. You do have to handle a different voltage range as the input to the voltage regulator. You can wire lithium-ion batteries in parallel if you're careful about using balanced cells (new, same lot, etc) but other chemistries are not as friendly about it. No product that takes AA alkalines wires them in parallel. \$\endgroup\$ – Matt B. Jul 2 '11 at 16:53
  • \$\begingroup\$ @Matt B., you get the same amount of current at a higher voltage. The amount of power is increased, but if you are using a resistive circuit, for example, you will see your current double and your lifetime will be almost the same. If you are using a high efficiency regulator, lets say 100% for the sake of discussion, then connecting more batteries means more life. In devices you can depend that the operating voltage is the operating voltage. Most devices we build are designed for a specific voltage and will vary resistance with need. \$\endgroup\$ – Kortuk Jul 3 '11 at 1:15

If the device is an off-the-shelf product, and you have no way of knowing what's inside, connecting extra cells in parallel is an option that's safe on the circuitry inside the box, as it will keep the voltage range of the battery the same. For 2x battery life, wire 2 in parallel, for 3x battery life, 3 in parallel. Not 3p2s or 2p3s etc. as adding any in serial will increase the battery pack voltage, which the device's circuitry may not be able to handle.

However you should note that connecting cells in parallel is not without danger to the battery itself. If the cells are not well matched, like if they are different capacity (size), age, state of charge, temperature, and so on, you can get a large current going from one cell to the other when you first connect them, until they equalize in voltage. This current can be high enough to do permanent damage to the cells.

Additionally there is the problem of slightly mismatched capacities, where at the end of a discharge cycle one of the cells will be empty before the other one, so the one with a little energy remaining will provide all the current. The same thing happens at the end of charging, where one gets full earlier and the other is being charged at twice the nominal rate. Unless you take the trouble to put a current meter on each cell, it's an invisible problem, but over many charge/discharge cycles it can do serious damage to the overworked cell, potentially even fire or explosion. When Sanyo recalled some lithium-ion batteries from Lenovo, it was the 3s3p extended packs that were the problem, the same cells in 3s2p were ok. The more cells in parallel, the worse the imbalance problem gets, which is why you generally don't see >3p in commercial products. The battery probably won't catch fire the first time you use it, but statistically over lots of shipped product there can be a higher failure rate. For a hobby project, you can mitigate this by using new cells (not salvage from different old laptop packs) and adding a resistor between the parallel cells as a current limiter at least until they equalize.

All of the above only applies to lithium-ion. If you're using NiMH or NiCd, you probably don't want to connect them in parallel at all. Lithium-ion is easy to charge because you just connect it to a power supply that is both current- and voltage-limited, and turn off the charger when the current falls below some threshold or you get past a timeout (or a safety flag such as temperature). NiCd/NiMH is a lot messier to charge (quickly), you need a circuit that can watch the voltage profile, temperature, etc. (unless you're trickle charging). The imbalance problem would be more difficult. It's probably been done, but when laptops ran from NiCd the packs had a higher voltage because the cells were all in series. Disposable alkalines are always in series because you can expect the user to pick any old cells from the drawer without regard for state of charge. They can already leak if overdischarged, connecting them in parallel would only make the problem worse. Imagine someone mixing NiMH and alkaline AA of various capacity and fullness randomly in both series and parallel - the product will soon be ruined. For a hobby project, nobody's going to care if you ruin a few Energizers, but for anything you'll hand to someone else you should be more careful.

There are other ways to improve your battery life besides connecting extra cells in parallel. Probably the easiest way is simply to use larger cells. For alkaline or NiMH/NiCd, switching from AAA to AA will more than double battery life, and for even more capacity you can use C or D cells. For lithium polymer, there are lots of sizes to choose from, including custom if you need it. If you're using disposable AA or AAA cells you can double battery life by switching from alkaline to Energizer lithium. They cost more than double, so they're more expensive in $/Wh, but they also weigh half as much and they still fit inside the device.

The other posters seem to assume the device is a black box, so perhaps it's not worth discussing too deeply what would happen if you had control over the whole design and added extra cells in series rather than parallel. Basically you'd have to redesign the entire power section - from a higher voltage wall adapter, somewhat different battery charger and protection circuit, different input voltage range to any regulators, possibly an entire redesign to anything powered directly from the battery rather than a regulator.

  • \$\begingroup\$ I do not think that the batteries specified in question are necessarily rechargables. Still good information. I would probably note that there are circuits out there to deal with battery isolation. \$\endgroup\$ – Kortuk Oct 10 '11 at 15:21
  • 1
    \$\begingroup\$ Non-rechargeables are addressed in paragraph 4. \$\endgroup\$ – Matt B. Oct 11 '11 at 17:13
  • \$\begingroup\$ I know, but I thought it worth pointing out. \$\endgroup\$ – Kortuk Oct 11 '11 at 17:34

You want to add batteries in parallel, keeping a constant voltage.

You are correct that if a device takes two batteries in series, you want to add additional series pairs in parallel.


The capacity (i.e. milliamp-hours or mAh) is what governs the lifetime. Capacity adds in parallel, so wire + to + and - to -.


Connecting two identical batteries in parallel will often increase the lifetime by a factor of at least two, and may increase the lifetime by even more than that (not only will the batteries be drawn down about half as fast as would be a single battery, but they may allow a device to keep working past the level of depletion that would cause a device to fail if only using a single battery). There is, however, an important caveat: if one battery is stronger than the other, it's possible that the stronger battery may try to push power into the weaker one. If the batteries are not rechargeable, such action may be detrimental to both batteries; even if the batteries are rechargeable, they may be damaged if they are connected while the difference in charge is too great.

Connecting batteries in series will cause them to push more voltage into the device they are using. The effect of this may vary with the device:

  • Some devices will be damaged or destroyed

  • Some devices will work better with the higher voltage, but draw more current (possibly leading to shorter battery life than if using one battery, but perhaps allowing one to get more work done in that time).

  • Some devices will work about the same at any battery voltage above a minimum threshold, and draw about the same current regardless of voltage. All of the batteries in series will get depleted at the same rate regardless of how many there are, but the level of depletion required to make the device unusable will be extended by having more batteries in series. For example, a device which will become unusable when its input voltage falls below 5.5 volts could be used with four AA batteries that were less than ~10% depleted, or six that were up to 90% depleted. Using more batteries in such a situation may be economical, up to a point, though if one had twelve batteries one might achieve optimal economy by using three groups of four (along with a couple of dummy cells or a jumper wire) until they no longer worked, and then using two groups of six.

  • Some devices use a constant amount of power, and will in fact use less current when supplied with a higher voltage. Putting more batteries in series when using such devices will be economically advantageous provided that the voltage is not increased beyond the device's ability to handle it.

  • \$\begingroup\$ Thanks for these caveats. In fact, I expect these possible concerns are the reason why specialized add-on power packs are designed and sold (for use with, e.g., smartphones). \$\endgroup\$ – boardbite Jun 30 '13 at 13:14
  • \$\begingroup\$ @boardbite: I would like to see a widely-adopted open standard for rechargeable batteries, though I don't know that anyone in the industry has any interest in such a thing. Rechargeable battery packs are sold with integrated electronics for a few reasons. Among other things, the only way to distinguish a new battery that's 80% drained from an old battery that's nearly full and has 70% as much useful charge as would a new full battery, is to know something of the battery's charge/discharge history. If a device has non-removable battery, the information can be kept in the device, but... \$\endgroup\$ – supercat Jun 30 '13 at 16:25
  • 1
    \$\begingroup\$ @boardbite: ...if a device uses interchangeable battery packs it's necessary to keep the information in the battery somehow. This may be done either by having a simple EEPROM in the battery which is kept updated by the device, or by having electronics in the battery that can monitor its operating conditions. The former approach has the disadvantage that if the battery is charged or discharged by something that doesn't know how to update the information properly, the information may become inaccurate. It is, however, cheaper. \$\endgroup\$ – supercat Jun 30 '13 at 16:27
  • \$\begingroup\$ @boardbite: Even before batteries included such intelligence, though, bundling them in packs meant that it all the batteries should be from the same manufacturing lot and exposed to identical conditions, and should thus generally have very similar states of charge. \$\endgroup\$ – supercat Jun 30 '13 at 16:29

protected by Dave Tweed Dec 11 '14 at 14:53

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.