EVE Energy 3.2V 90Ah LiFePO4 Batteries

Somehow, the batteries I get to test are usually relatively small, but people often ask me to review higher-capacity cells, for example, ones for UPS units. So today I’m testing 90 Ah LiFePO4 cells: I’ll inspect them, test them, and discharge them in a freezer.


Not long ago, I reviewed and ran a cycle-life test on 18650-size LTO batteries, and complained about their inconsistent quality. Some time later, a store representative got in touch, apologized for the poor-quality product, said they had taken it off sale, and offered to test some other batteries. I accepted the offer. They also offered a discount for readers, but I’ll tell you more about that at the end of the review.

Anyway, I picked up a fairly large box from the post office, containing a pair of cells.


My first impression was that someone had taken books with blue covers and attached terminals to them. The batteries really do look like the familiar “book” size :)


Specifications from the product page. Based on these, I’d put these cells in the “mid-range” class: although the continuous discharge current is listed as 270 A, that works out to just 3C for their capacity.


The batteries are wrapped in blue heat-shrink tubing, with no markings on the sides.
The terminals are threaded for M4 screws, with a depth of 6 mm. In my opinion, this type of fastening looks a bit small for the stated discharge currents; I would have expected at least M5.
The terminals differ not only in color; each plastic surround is also marked with the polarity.
There’s a sticker on the safety valves, and each battery also has a unique QR code with its serial number, which matches the number printed in plain text next to it. The cell’s capacity marking is also next to the QR code, showing its voltage and capacity.


Dimensions and weight.
Without the terminals, the dimensions are 195 × 130 × 35 mm; the terminals add about 4.5 mm to the length.
They’re pretty heavy, at around 2 kg each. Interestingly, the parcel was also listed as weighing around 2 kg, but when I carried it home, it felt noticeably heavier—and sure enough, the two batteries weighed over 4 kg together with the packaging.


On inspection, I immediately noticed that the case was a little “bumpy.” At first I thought the battery had the characteristic “swelling,” but it turned out to be more like indentations than bulges.


The stated battery impedance is 0.3 mΩ; my measurements came in closer to 0.25–0.28 mΩ.


For the test, I used long M4 screws, which I didn’t tighten all the way. To ensure good contact and prevent them from turning, I also added locknuts—and as it turned out, that was definitely worthwhile.
The thing is, these screws are very thin for the alligator clips on the EBC-A40L load. After connecting them, I had to squeeze the clips and twist them to “wear them in.” Without the locknuts, I’d either have had to tighten the screws too much, or they would have come loose.


The batteries arrived fully charged, but I still ran the mandatory “conditioning” cycle to normalize their condition.


According to the description, the batteries have a nominal charging current of 18 A and a fast-charge current of 90 A.
At the normal rate, charging took 5 hours and 22 minutes, with a fairly short CV phase. Of course, it’s short compared with 5 hours, but I could already see during charging how quickly the current dropped toward the end.


Unfortunately, without additional workarounds, I can only charge at up to 40 A. At that rate, charging takes about 2.5 hours, which is also pretty good.


I also discharged them at 18 A and 40 A. The 18 A current is 0.2C, while 40 A is close to 0.5C.
I ended up getting about 2 Ah more than the stated 90 Ah. I ran the 40 A discharge test twice simply to make the graph a little less “boring” :)


The second cell produced comparable results.


At a discharge current of 40 A, the battery case warmed up to 34–36 °C.


I also had no particular concerns about the consistency of the results. There was a small variation, most likely due to measurement error and changes in temperature.


During testing, I decided to try measuring capacity at a higher current, at least 1C or 90 A. Since I don't have a load like that, I had to improvise.
Connected in parallel were:
EBC-A40L
Ebc-A20
My main load for testing power supplies.

In the end, I got a current of 40+20+30=90 A.

The setup was certainly pushing its limits, but even so, I don't think the measurement error was all that large.


There was a small hiccup during testing: I tested the second unit first and forgot to reset the Ah counter on one of the loads (it doesn't reset automatically).

The first unit showed 41.41+20.66+30.877=92.95 Ah.


For obvious reasons, I can only show the graph in a cropped form, without linking it to the overall test, but it's still possible to assess the change in voltage. In principle, you can also calculate the capacity in Wh; I got 288.02 Wh.


These are the results for the second unit. The Wh counter hadn't been reset, but I was able to recover the starting reading from the photo: 17.808 Wh.
So the capacity was 40.91+20.41+(48.299-17.808)=91.81 Ah.


The calculated energy delivered was 289.63 Wh.


During the test, the batteries warmed up to about 42–45 degrees, which isn't all that much. The third photo shows the overall setup.


The batteries are advertised as new, and one of the telltale signs of a new battery is not just that it meets its rated capacity, but also that it has a low self-discharge current. So I decided to check that, too.
To do this, I fully charged the batteries, then repeated the charging cycle 50 hours later and measured the capacity taken in.
The first unit “took” 186 mAh, and the second even less—107 mAh—which indicates a low self-discharge current.
It's also worth noting that even if you start charging again immediately after the charging cycle ends, the batteries still take about 30–40 mAh, so you can subtract that from the results above.


But I think the freezing test is much more interesting. For that, I left one of the test subjects in the freezer for almost a day.
The battery case temperature was -16 degrees. As you can see in the photo, it immediately became covered in condensation, and the impedance increased from 0.26 to 0.43 mΩ.


I got 88.34 Ah or 252.8 Wh, and the battery warmed up to 30 degrees even though it was in the freezer during the test.


But the graph gives us more information. It shows the test results at room temperature (red) and in the freezer (blue).
You can see a significant voltage drop right at the start, followed by a rise as the battery warms itself up, and then another drop. In regular tests, the voltage drop at the end was very sharp, whereas here it was gradual.


After the test, I let the battery “rest” for a while at room temperature and continued discharging it. I got another 3.3 Ah or 9.2 Wh.


The obvious question is: are batteries like these available on AliExpress? Yes. For example, Turmera sells them under its own name for $80. There are similar ones from Litokala, but those are rated at 105 Ah and cost $170 for 4.

Turmera is probably selling decent cells, judging by the price at least, but Litokala's are noticeably worse. Reviews say they have a capacity of around 83 Ah and an impedance of 0.35–0.37 mΩ, which suggests they're probably used cells.

I also liked Litokala's managers' Photoshop work: they pasted two QR codes onto one cell and left another with no code at all.



Conclusions
Based on the test results and the condition of the terminals, I can say that the batteries are new, their capacity is higher than rated, and their impedance is lower than the stated 0.3 mΩ. The batteries also showed good consistency, making them suitable for building battery packs.
At room temperature, I easily got 91–92 Ah at load currents up to 90 A. In other words, during discharge, each cell delivered almost 300 W, equivalent to 1,200 W for a 4S pack. The temperature was about 42–45 degrees.

Discharging at low temperatures, however, produced a less impressive result. There was a significant voltage drop at the beginning, which could be critical at high load currents. But I think that at 20–60 A, the battery will quickly reach a reasonably normal operating temperature through self-heating.

In short, it should work great in a UPS. I’m not sure about using it in a car: there could be problems at low temperatures, but I can’t say for certain.

And now for the promised discount. The store offered 5% off any batteries (I hope I understood that correctly) if you provide the code "kirich" when placing your order, by any convenient method. I wouldn’t say it’s a huge discount, but I don’t think anyone would turn down a little extra savings—and on a sizable order, it could make a real difference.
I was also told that although the store is Ukrainian, you can ask about the possibility of shipping to Russia.

That’s all from me. I hope you found this useful.