Testing 30 brands of batteries

Batteries come packaged in bright blister packs emblazoned with vague guarantees such as “45% more pictures” and “five times longer lasting.” During his internship at BitBox this summer, [Thomas] decided to put those statements to the test. He tested thirty brands of batteries on a homebrew rig to find the batteries with the most power and the most bang for your buck.

The hardware [Thomas] used an STM32 microcontroller to perform two different tests: a high drain and a low drain condition. For the high drain, 1000 mA were sucked out of the batteries until the voltage reached 0.8 V. For the low drain, 200 mA were used. Data including milliwatt-hours, milliamp-hours, joules, voltage, current, power, and effective load resistance were all logged for both conditions for all 30 batteries.

Generalizing the results for both low and high drain conditions, lithium batteries were better than alkaline, which were both better than zinc AA cells. Perhaps unsurprisingly, batteries marketed as ‘long life’ and ‘extended power’ were the worst batteries for the money, but a brand-name battery – the Kodak Xtralife cells – were actually the best value for the money.

Comments

  1. boondaburrah says:

    Would there also need to be a test case for batteries that needed to do high drain in short infrequent bursts over a long time? Or is that covered by testing high drain? (Assuming drain inbetween bursts is either negligible or covered by the low drain test) I’m not up on my battery science, so I don’t know if this type of stress is different enough.

    • Tony says:

      Well, YMMV and all that.

      For infrequent high bursts, I’d choose something off the 200mAh list.

      These sort of tests show looking at the mAh rating doesn’t necessarily mean ‘bigger is best’, you need to look at the recommended discharge rate for the battery (which can be hard to find).

      Generally speaking, to extract the most energy from a battery you need to do it slowly.

      You do have to wonder how many fake batteries he tested – genuine Ford batteries (?) from a discount shop? Hmmm.

      • Thomas says:

        They were definitely a licensed Ford product, the brand has been licensed to RMS International Ltd., and sold in a reputable discount chain – “99p Stores”. Since there were also Daewoo and Hyundai batteries for sale, I guess car manufacturers don’t care for their brands any more. (I don’t think Daewoo exist any more.)

      • Tony says:

        Weird. Maybe Ford will start branding HDMI cables etc soon.

        Daewoo (who did electronics as well as cars) fell over a couple of years ago. GM got the car.

        There are a lot of fake batteries around though. To add to the hassle even the big brand manufacturers change formulations (lowering capacity) occasionally, invalidation old results. It would be interesting to see yearly tests.

      • mojo-chan says:

        I wouldn’t call 200mA a low drain test. Things like clocks and wireless mice use an order of magnitude less at least. Even my little personal cooling fan uses only about half that.

        Also temperature is very important for some types of battery, particularly lithium. The biggest issue with lithium is that the voltage remains fairly constant until suddenly it drops at the end of life, meaning there is no easy way to tell how much energy you have left. There are different types of lithium battery too, spindle and spiral, each with its own unique properties.

        Lithium is nominally 3.6V so 1.5V lithium cells use some weird chemistry that is non-optimal, so if you are planning a project consider 3.6V lithium instead of 1.2/1.5V cells.

  2. Bogdan says:

    The only place where i still have non rechargeable batteries are remotes and clocks/watches.

    Batteries labeled as “long life” or similar are designed for long term usage in things that drain them a little, like clocks. They are designed for low leakage so that the battery doesn’t self discharge more than the device discharges it. So, the tests here are not so relevant, they discharge the battery way too fast.

    Some years ago(maybe 6-7) i measured the current consumption of digital cameras running on 2 AA alkaline bats. When taking a picture and charging the flash current went to around 3-4A. So maybe the high power batteries should be tested at higher current too.

    Still, the most interesting thing is the very small difference in capacity between the poorest alkaline and the best alkaline. But they have a high price difference making the cheap alkalines way better in terms of price/performance in low consuming devices.

  3. Steve says:

    Ford batteries FTW!

  4. loopingz says:

    It would be interested to test Nimh and Nicd too.
    I would also be super interested on some more info on the hardware + software behind as I am looking forward to uild these kind of stuff with 4 batteries at the same time (on different graphs).

  5. Siiiigh says:

    Grown man conducts middle school science fair exhibit, finds same result as middle school students only with prettier graphs…

  6. Chambon says:

    Summarizing the results:

    1) capacity-wise pretty much all zinc batteries are the same, pretty much all alkaline batteries are the same, pretty much both Li batteries are the same.

    2) capacity-wise Li > Alk > Zn (duh!)

    3) capacity/cost: just get the cheapest alkaline battery and win.

  7. Andrew says:

    Tony, did you test more than one battery of each brand?

  8. obsoehollerith says:

    I have a specific kind of need in batteries – as a musician, I’m constantly forgetting to replace the batteries in my stomp boxes. I have found that the top of the line Alkalines have the best ‘old battery comes back to life’ spec when they are left for a few hours unplugged and unused. The bargain alkies just give up,and that’s that. For my lazy, forgetful dollars, (few and far between), that’s pure gold. Good work though, for everything else, I’ll buy the dollar store brands. Thanks!

  9. 0xfred says:

    Very useful. A while ago I was in a store and trying to decide between two version of Duracells – I couldn’t find anywhere that actual capacity was stated, making it very hard to decide between “megaultra” and “supermagnifico” or whatever stupid hyperlatives they’d gone with.

  10. Brett_cgb says:

    I would have liked to see the chemistry type listed in a column in the raw data results. Sometimes its not easy to determine chemistry from the battery name.

    Nicely done!

  11. Joe C says:

    It looks like a good test setup, but I would have been interested to see some rechargeable batteries in the mix. Since I discovered pre-charged NiMH, they’re all I buy, unless it’s for something really infrequently used, like a remote control.

  12. Ren says:

    I’m thinking about something to test rechargeable batteries. As some of mine age, their capacity drops and while the charger will show it as full, the battery won’t supply power for a few seconds or minutes when I swap it back in. I would need to be able to test NiMH, NiCd, and even the “non-rechargeable” alkalines the Buddy-L recharges.

  13. NewCommentor1283 says:

    this is the data we needed, thank you

  14. Eddie says:

    My problem with Duracell AA’s is that in my experience, they leak electrolyte.

    This never happened 10+ years ago.

    Another thing that bothers me are AAA batteries. For a little bit more in space you get 2.5 times as much power and useful life.

    Actually I HATE AAA batteries.

    • Chris C. says:

      +1. I’m already using rechargeables for anything where capacity is an issue, and alkalines when I need longevity. But I keep finding my alkalines have taken a big messy dump inside my electronics. Thinking about doing away with them and using lithium instead.

  15. ftorama says:

    I found by myself (only through using them) that Varta High Energy were very good batteries, but when I think about the ads from Energizer…

  16. snow says:

    if i am not mistaken its not the first time a batteries test was done on HaD and like i pointed out the last time: when testing batteries you should consider that not all of them were manufactured at the same time but chemicals start to react as soon as the batteries are manufactured therefore a constant decline in capacity over time occurs.its hard to know how long the batteries were in the shelf before he bought them and in all fairness this should be reflected in the results.

    • Thomas says:

      They were all purchased within 1 week of each other and each chemistry had comparable expiration dates. Besides, on a 7 year life, a month is less than 2%.

    • DainBramage1991 says:

      Many batteries come with a “sell by” or “best if used by” date printed on the package that can be used to narrow down the date of manufacture, which can be particularly useful when doing comparisons such as this.

      My own experience with batteries has been that I get the most bang for the buck out of Ray-o-Vac alkaline batteries, but I tend to us NiMH rechargeables as much as possible.

  17. TechDud says:

    I would love to see them compare alkalines for their rechargability.

    Note that when charging alkalines (with an alkaline charger), one need leave the vent pointing up to avoid leakage. NiCd & NiMH cells have the vent by the positive terminal instead.
    After charging, leave the cells upside-down for a day for any gasses to vent.

    I’m still trying to find a datasheet that explains where the vents are located on a 9V battery.

  18. chris says:

    Can we see Rechargeables next :D

  19. bothersaidpooh says:

    I want a rechargeable V23A aka those sucky little 12V lighter batteries used in the Peak and many other handheld instruments.

    Yes, the “Godp” 3k mAh NiMH ones royally suck @DainBramage1991, I tested them and got capacity half what they were rated even after repeated cycling to 1.1V and 1.43V
    The internal resistance isnt too bad though so they are fine for little Johnny’s R/C car but not for something expensive.

    I tried one again after three month’s light duty in a solar light and it had under 600mAh :-(

    Little tip. Weigh them, the capacity roughly correlates to weight for similar chemistry.
    4Ah NiCad is roughly twice as heavy as a comparable 4Ah NiMH, due to the cadmium.

  20. ansil says:

    i like this now i just need someone to do 9v

  21. Greenaum says:

    Didn’t we do this whole story a couple of months ago? Unless two people have done the same project.

    • TechDud says:

      nobody has done as comprehensive a job.
      Now, if they would just test the rechargability of alkalines. This subject could be put to rest.
      I’m done with NiMH & using rechargable Alkaline for medium to low-drain devices. That 1.5V seems to give things a little more oomph.

  22. evilspoons says:

    It’s nice to see the cheap Costco batteries are nearly identical to the Energizers that cost twice as much on the next shelf up.

  23. JourneymanWizard says:

    Dropped a line to the guys at BitBox (seem good folks) for details on the setup:

    Query sent:
    —————–
    Some questions on your testing (if you don’t mind): your testing method stated you applied a 200mA or 1000mA load to the battery until the terminal voltage dropped to 0.8V. Is your load dynamic (e.g. holding at 200mA) or constant (e.g. 200mA at initial nominal voltage)? When you calculate mW/h capacity, how do take into account drop in voltage (and current if constant load used) in your calculation, or do you? Would you be willing to release schematics, partial schematics, firmware, code, or a block diagram of your BitBox battery device?
    —————–

    Response received:
    —————–
    The battery tester was one of the projects completed by our summer intern.

    He designed a board with an active load (a FET with a control loop wrapped around it). The code he wrote would present a straight resistive load, constant current and constant power load (to simulate a switched mode power supply). We used a processor module rather than sticking a processor down on the board for convenience.

    The embedded board was a slave to a python program running on a crusty old laptom. The embedded board can be set up through a serial port to apply any load you want. When the test is run it sends results back at 1 second intervals showing the achieved current, terminal voltage, mA/h accumulation, mW/h accumulation plus some other bits… The mW/h and mA/h accumulation is calculated on the embedded board (rather than the laptop) and it does use the actual terminal voltage and achieved current in its calcs.

    We chose 0.8V as the end point arbitrarily. One battery manufacturer specifies 0.8V as an end point in its data sheets – so we stuck with that.

    I will happily publish all of the source designs, schematics, code etc. etc. The caveat is that it was written and designed by an 18 year old intern! (possibly better than our code????!!!). The problem is that I don’t have much time to update the web site now he has gone back to continue his studying.

    Due to the hack-a-day link and life-hacker we have had 53000 hits our little site :-)

    Anyway, thanks for your interest.

    Quentin
    —————–

  24. Eben Spurdo says:

    OP is an obvious shill from Kodak as the thread is pure hearsay.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 91,915 other followers