Ultimate AA Battery Test and Comparison

I purchased 37 different AA batteries across four types (Alkaline, NiMH, LiFeS2, and buck-converted Lithium-Ion) in order to find the best for flashlight enthusiasts.

Each battery went through three tests:

  1. CBA test @ 250mA
  2. CBA test @ 2A
  3. Runtime test in Convoy T3 (519a) @ 100% output.
Full list of batteries tested:

ALKALINE

  • AmazonBasics Alkaline
  • Kirkland Signature
  • eCircuit Blue*
  • eCircuit Red
  • Duracell Optimum
  • Duracell Powerboost
  • Energizer Alkaline Power
  • Energizer Industrial
  • Energizer Max
  • Great Value Alkaline
  • Rayovac Fusion
  • Rayovac Fusion+
  • Rayovac High Energy
  • Rayovac Ultra Pro

NiMH:

  • AmazonBasics NiMH
  • AmazonBasics NiMH High-Capacity
  • Duracell Recharge
  • EBL NiMH 2500
  • EBL NiMH 2800
  • Energizer Recharge Value
  • Energizer Recharge Universal
  • Energizer Recharge Power Plus
  • IKEA LADDA 1900
  • IKEA LADDA 2450
  • Panasonic Eneloop
  • Panasonic Eneloop Pro
  • PowerOwl NiMH 2800
  • Rayovac Recharge

LiFeS2:

  • Blinkpower Lithium AA
  • EBL Lithium AA
  • Energizer Ultimate Lithium
  • Powerowl Lithium AA
  • Voniko Lithium AA

Lithium-Ion AAs:

  • Coast Zithion-X
  • EBL 3000
  • EBL 3500
  • Zepath 3600

OVERALL RESULTS:

Alkaline:

Most alkaline batteries saw a steep decline in measured capacity between the 250mA and the 2A tests - these do not handle high current draws well.

In the Convoy T3, the general trend was for a breif period of high output, and then a gradual deline in output throughout the rest of the runtime. Many of these batteries technically had a pretty long runtime, but not at meaningfully high output levels.

NiMH

Most NiMH batteries had very similar measaured capacities between the high and low current tests, which was quite impressive.

In both the CBA and the Convoy tests, NiMH batteries were able to hold their voltage very steady trhoughout the run, allowing them to produce consistent performance through their lifespan - very nice!

LiFeS2

These cells had significantly longer runtimes across the board, though they lost some capacity in the higher current draw tests relative to the 250mA test. We also see declining voltage throughout the test.

Runtimes in the Convoy light much longer than any other type, and output was not flat but still held fairly steady throughout the run.

Lithium-Ion

Here we see much more variation between the different batteries tested. Capacity between the two CBA tests was pretty close, though most lost a bit in the 2A test.

2 of the batteries (Zepath 3600 and EBL 3000) held their voltage perfectly flat until dying, while the EBL 3500 and Coast Zithion batteries did not.

In the Convoy test, we see the same results. Overall runtimes are very good, on the longer end of the scale. (Note the sudden jumps back up to full output displayed after dying for a couple of these. An artefact of the buck conversion and voltage recovery)

DETAILED TEST RESULTS

There is too much data to compile in a neat manner on this site, so I’ve created a public spreadsheet that has everything for detailed comparison. Also included are (affiliate) links to purchase every battery:

SPECIFIC RECOMMENDATIONS:

Alkaline:

  • Duracell Powerboost: Best performance alkaline AA
  • Costco Kirland Signature: Best value alkaline (note: only available at Costco!)

NiMH:

  • Panasonic Eneloop (White): Longest lifespan (2,100 charge cycles), great performance, wide availability, dry chemistry doesn’t leak
  • Powerowl NiMH 2800: Best value NiMH, includes a nice charger

I do not recommend high-capacity NiMH generally as they provide worse value, but if you want higher capacity per charge:

  • IKEA LADDA 2450: Best value high-capacity NiMH
  • Duracell Recharge: Best performance NiMH (in a single charge)

LiFeS2:

  • Powerowl Lithium AA: Best performing AA (overall), best value LiFeS2 battery
  • Anything else: The rest are basically equivalent.

Lithium-Ion:

  • Zepath 3600: Best value lithium-ion, great performance overall.
  • The rest are still good, but I don’t think there’s a reason to buy any of them over the Zepath.

CONCLUSION

Overall, I think it’s very clear that there is no reason to buy alkaline batteries - they are disposable and leak, which are obvious issues, but they have abysmal performance in comparison to everything else.

LiFeS2 batteries are really awesome for certain applications, but only for those applications… their high price and disposable nature make them unsuitable for everyday use. I would keep them basically just for emergency backup cells for long-term storage, or maybe for long-term use in super-low drain applications (like a clock or something)

NiMH and lithium-Ion batteries are actually very similar here, NiMH probably have better longevity, but you may run into compatibility issues due to their lower nominal voltage.
Li-ion AAs aren’t exactly revolutionary, and their conversion circuits can lead to some strange behaviour in certain devices. They also require specific chargers (though every battery I’ve seen and used did include a chager, so that shouldn’t be too much of a concern… just don’t try to charge in something not specced for it!)

*Notes

There are some important shortcomings in my tests that need to be noted:

Spoiler
  • I chose 100 lumens as my cutoff point for the Convoy T3 test… this was probably the thing that would cause most people to disagree with me. It was chosen because I felt that specific value offered the best balance between the sharp cutoff of some cells (such as NiMH) and the gradual falloff of others (such as alkalines). Most all batteries fall very quickly at or before hitting this value. Notably, in most electronics this is about the point where usability is really going to drop off. HOWEVER, everybody has different needs… which is why I recommend looking at the graphs for your specific battery choice!

  • Most batteries went through each test multiple times - except for the lithium-primary cells, as they are expensive and I had only four each. I ensured every test was performed with as fresh a battery as possible - for both rechargeable and disposable cells.

  • Two things I am unable to test properly: Shelf life, and Lifespan (actual cycle count and degradation for rechargeables). I am simply taking manufacturer’s words on these.

  • I have no financial incentive to sell you any particular cell, no sponsors are present in this video/post. I do have affiliate links for MOST (not all) batteries on this list. Only one set of batteries was provided for free - the Powerowl NiMH… though the company doesn’t like me because it took over a year to ever bring their batteries up in a video

  • I selected the best value option for each cell in my “Price per cell” calculation (ie, if the 8-pack was the best value, I chose that, if the 12-pack was the best value, I picked that), with the exception of large bulk orders (another reason I recommend using the linked spreadsheet!)

  • I didn’t talk about the 8-pack AmazonBasics Alkaline set, which is an equivalent price per cell to the Kirkland set. Simply put, the Costco batteries were much better, so I don’t find the AmazonBasics competitive even at that bulk price… but they are more easily available so that is worth mentioning.

  • You’ll notice that some of the batteries displayed weird fluctuations in the CBA test. In short, this was repeated and consistent on the cells that did display this behaviour. I do not know why it happens.

THE END

I don’t really have a lot more to say here, it was all said in the video and in the spreadsheet. I genuinely hope this information is valuable to you guys!

16 Thanks

Extremely informative review, well done! Just a few minor suggestions:

(1) For the plots, it would be better to use more colors other than various shades of blue and green, or various shades of yellow, which are very difficult to tell apart given the huge number of plots per image.

(2) In the end, please justify your recommendations and state explicitly your criteria. For example: the AmazonBasics NiMH AA performed 3rd-best in the discharge graphs, but is not mentioned in the 4 recommendations. So what criteria are the recommendations based on?

If by lifespan you meant runtime, it appears that the Ladda 2450 and Duracell Recharge jointly placed the highest, so I could not justify the above claim.

3 Thanks

Any plans on adding the xtar cells?

1 Thank

Ah thanks for the feedback. The plots are only supposed to give a general overview of the category, individual graphs are provided in the spreadsheet (this post is mainly supposed to move users there because it is the best way to represent all the information).

Lifespan refers to cycle count (Eneloops are rated for 2,100 cycles).

The reasons for my recommendations are given. My primary criteria is value in $ per minute of runtime over the cells entire lifespan. The AmazonBasics cells perform well enough but don’t stand out for any reason.

2 Thanks

Don’t you mean the regular white Eneloops?

Oops, yes I did.

Fantastic info - thank you!

How to put it in perspective - how would a decent 14500 like F15 or H10 behave under those tests?

Wonderful and useful set of tests. I did find the graphs to be unreadable. Maybe using more colors would help. But anyway, thanks for doing this!. I have been buying Energizer Ultimate Lithium cells for application that needs this kind of cell. I will definitely try those PowerOwl cells now.

Thank you very much for clarifying! The spreadsheets are a tremendous asset.

I saw that you had a (very thoughtful) disclaimer that you are unable to test cycle life, and for this reason assumed that by lifespan you meant runtime. Your criteria do make sense, assuming that the rated cycle life can be trusted.

1 Thank

Thanks for the testing, but no Tenergy?

Love it

1 Thank

I know Energizer lithium did not win any awards but in my experience its absolutely amazing and super relaiable. They run my thermostats for 4+ years continuously.

1 Thank

Okay, the graphs have been updated. Still difficult to read, but hopefully you can single out the notable cells now.

I did not get any Tenergy cells for this test, sorry. I do actually use their NiMH C and D cells though. If people want me to add more batteries to this test I may consider doing so, the Tenergy are available in both standard and a high-capacity versions too it seems.

I love the Ultimate Lithium, they have worked very well for me, and they do have the longest rated shelf life of all the LiFeS2 batteries. They also have the advantage of being readily available at large retailers. I didn’t find they competed well on value though, and the PowerOwl cells just performed better.

Here is a thorough test of the H10 by @HKJ: Test/review of Vapcell INR14500 H10 1000mAh (Purple) 2020

In short, it can handle vastly higher current draw than any AA seems capable of, but capacity is around 1000mAh at its highest, which is quite a bit lower than all NiMH cells tested. This does not exactly translate to better performance in a flashlight though, due to differences in voltage… basically, 14500 cells are still much better than AA batteries.

4 Thanks

This is an excellent distinction that should not be missed, and the reason why IMO “capacity”, i.e., mAh, is a poor metric of performance. It is completely meaningless without reference to voltage, and quite misleading when comparing different chemistries, or even same-chemistry cells with different discharge characteristics.

Much better is “total energy”, measured in mWh = (V)(mAh) = 3.6J. This can be obtained by integrating the voltage curve against capacity during a constant-current discharge test at low current; a useful first-order approximation is given by the product of capacity and nominal voltage. Thus 1000mAh @ 3.6V translates to 3000mAh @ 1.2V, which is more than what any NiMH AA cell achieves.

In some cases, even the first-order approximation can be faulty: a high-current 18650 rated at 3000mAh may easily out-run a low-current 18650 rated at 3400mAh during a high-current discharge.

1 Thank

Those “special chargers” for the Li cells with converter board just apply 5V (ie, usb voltage) directly across the cell.

All the smarts are in the converter board, and instead of a usb port to charge them, the “charger” just hits them with the usb voltage directly.

In a pinch, a usb cable can be snipped to just expose the red/black wires, and hit the cell with the wire ends directly, like those doodads with the magnetic ends that you can charge regular 14500s with.

PowerOwl has 4.5 stars on Amazon. However if you read the reviews, they horrible. Most people say they dont last long. Orher say the voltage was all over the place.

Once again, Amazon reviews are misleading. I dont know how they got 4.5 stars when most reviewers gave one to two stars.

4 Thanks

Great info - a lot of time and hard work!
All the Best, Jeff

2 Thanks

Thank you for review. I bet youre not in North America? It is quite strange that you omitted prob the best alcalines HDX from Home Depot. Then also where are XTARs? So many questions.

@LuxWad

Yes, if you manage to get some of the Xtar LiIon 1.5V buck cells, it would be interesting to see how they measure up to the rest. Those are the best cells of that type that I have tested to date.

1 Thank

Alkaline batteries are indicated for other much more suitable uses, where they are superior to rechargeable batteries. But always in electronics that require very few amps. Then they have no rival.