EVE 40135 20Ah LiFePO4 batteries

It's been quite a while since I wrote a review. I've been pretty swamped lately with work on new electronic loads. But some time ago, one of my readers sent me several different types of batteries to test, and today I'm reviewing the first ones.

A reminder: to keep up with new posts on the blog, I recommend following me on Instagram. I'll post notifications about every new article there, and might also write about something interesting from time to time - link to my account

You can also subscribe to updates and new articles on the Telegram channel - https://t.me/KirichBlog


There were four types of batteries, but they all have one thing in common: they're lithium iron phosphate cells. The first type is 20 Ah, two types are 15 Ah, and one is the familiar 32700.


A datasheet for these batteries is easy to find online, and the company is quite well known in general.


They look like typical cylindrical cells. Nickel (or nickel-plated, I didn't check) tabs are welded onto the original terminals; the welds are neat and secure.
Each battery has a matrix code and serial number. Through the heat-shrink tubing, you can also just about make out a QR code and serial number. I tried to photograph them, but couldn't.
The terminals are insulated from the casing, although there is a small potential relative to one of the poles, about 2/3 of the battery voltage.


Size: 40135, i.e. 40 mm in diameter and 135 mm long. They weigh 357 g each.


The capacity of all the batteries I received was measured, and the value is written on the label. One pair had clearly been tested noticeably earlier, and that wasn't a coincidence. Here's why: someone who sells these batteries initially contacted me to ask why some cells seemed to have something moving around inside when shaken. It feels as though the roll of electrode material is slightly smaller than the can—or, more precisely, shorter, since the rattling happens when you shake the battery along its axis, not across it.
My guess was that after a few charge/discharge cycles, the roll might expand a little and the "rattle" would stop. Personally, though, I wouldn't use them in mobile devices, especially ones that are carried around, but they might be even better suited to stationary use. To get ahead of myself, I'll say that 5-6 cycles at different currents made no difference.


All the batteries were in storage condition, i.e. charged to about 50%. The voltage of each cell was also almost perfectly balanced.




During the initial charge, I checked how much capacity the batteries took in, and it was practically the same for all of them. Unfortunately, I didn't record the results for the fourth one—I simply forgot—but the first three came out as follows:

10.02 Ah
10.00 Ah
10.09 Ah


According to the datasheet, the batteries have two specified charging modes, at 0.5C and 1C. The cutoff voltage is 3.65 V in both cases, and the cutoff current is 1 A.
At 0.5C, the graph looks great: it's in the CC phase almost the entire time, switching to CV only sharply at the very end.


At 1C, things change noticeably, and personally, I think it's better to charge at 0.5C.



Capacity was measured in four modes, at 0.2C, 0.5C, 1C, and 2C, where C = 20 A. The maximum specified current is 3C, or 60 A, but unfortunately my load can only go up to 40 A :(((
No real issues here; everything matches the specified values. I only noticed later that the datasheet gives capacity at 0.33C and 0.5C. Why the manufacturer decided to depart from the standard 0.2C is a mystery to me....


Temperature at 20 A and 40 A—no complaints here either.


The second sample gave essentially the same results.



The third and fourth samples were broadly similar, but there was "one catch"—more on that later.



By the way, these samples ran a little hotter, because their impedance was also slightly higher.
Results at 40 A for the third and fourth cells. I also noticed a temperature "overshoot": it kept rising for about 15-20 seconds after the current was switched off, then started to fall. The photos show the maximum values. This may be due to the bobbin not making firm enough contact with the strip and the casing.



For the second and fourth cells, I checked the remaining capacity after discharging them at the maximum current for my load. The remaining capacity was similar, around 1100–1200 mAh, which is good. What wasn't so good was that it was relatively high. This result helps determine the maximum usable discharge current, which in my view is around 1C.


And of course, a comparison of the results, along with the very “quirk” I mentioned above.
At 0.2C, the capacity of all the cells is almost identical, allowing for measurement error, of course. Incidentally, that error depends on the air temperature and, accordingly, the battery casing temperature.
But at 40 A, something odd showed up: the third cell under test delivered noticeably less.


My first thought, of course, was that something had gone wrong during testing. By the way, all the batteries were tested without disconnecting them from the charger: I connected one to the battery tester and didn't touch it again until the test was complete in any of the modes, so a contact issue somewhere seemed unlikely.
Still, I connected the third cell to the tester again, charged it again, and discharged it again at 40 A. As expected (I trust my equipment), I got practically the same result, with only a minimal difference.
The first test yielded 18.14 Ah or 55.70 Wh; the second, 18.19 Ah or 55.40 Wh.


Next, I decided to compare the third and fourth batteries (since they have similar dates marked on them) in all four modes to find out where the problem occurs.
The graph shows that at 0.2C and 0.5C, the results are practically identical. At 1C, the voltage of the third cell sagged slightly toward the end of the discharge, while at 2C the difference was already very noticeable.
In practice, there are no issues up to 0.5C, and none at 1C either, but the third cell shouldn't be used in a battery at 2C.



Interestingly, the batteries have different impedance. The pair with the “rattle” is at the top, and the pair without it is at the bottom.

I'll also point out that people often look only at impedance and capacity measured at a low current. But, as experience shows, you should measure performance under the same conditions in which the battery will be used, whenever possible. Unfortunately, that's not always possible, or it's expensive. I could do that, but each cycle takes at least 3–4 hours.


And, of course, the conclusions.
I've reviewed batteries from this manufacturer before, perhaps more than once, so the results surprised me a little this time. On the one hand, these are quite good batteries: the measured capacity matches the ratings specified for new cells. But two things bothered me: the bobbin rattling inside, and one cell's low result at 2C.
The first issue is absolutely insignificant for stationary use, while the second is unusual and, ideally, calls for more data.
That said, I can't rule out a simple manufacturing defect. For example, I've tested many new cells from reputable brands—LG, Samsung, BAK—and among roughly 500 cells, I came across one Samsung with a poor connection in the vent.

Otherwise, these are interesting batteries. I also liked how consistent their characteristics are, their low self-discharge (noticed during testing), and the quality of the strip welds.

That's all for today. I hope this was useful, and special thanks to the person who sent me the batteries for testing.