BAK 32140FS-48 15 Ah batteries

These are the third batteries I've received from a reader, and this time they're from the relatively well-known BAK. Sure, they're not Samsung or Murata, but BAK is still a recognizable name.

To stay up to date with new posts on the blog, I recommend following me on Instagram, where I'll post notifications about all new articles and maybe write about something interesting now and then: link to account

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


I've been asked about BAK battery reviews many times, and I've tested their cells on several occasions. They're fairly popular because they're not too expensive and have generally decent specs. But this is the first time I've had LiFePO4 cells of this size in my hands, so I was interested in testing them and, especially, comparing them with others.


I looked for the datasheet but only came across a screenshot from it, which is more than enough for this test. The charge/discharge specs were a little unusual: for example, charging to 3.60 V instead of 3.65 V, and discharging to 2.0 V instead of 2.5 V. To be fair, especially given the charge/discharge characteristics of LiFePO4, the difference will be minimal. Still, the voltage of these cells is shifted slightly toward the lower end.


The heat-shrink sleeve has a barcode, model name, capacity in Wh, and voltage printed on the side.
The negative terminal is slightly convex and almost flush with the heat-shrink sleeve, which made it easier to connect to the battery tester.
Otherwise, they're ordinary batteries.


The dimensions are a bit odd. These cells are labeled 32140, meaning 32 mm in diameter and 140 mm long, unlike the standard 33140 cells, which are 33 mm in diameter. But my measurements showed an actual diameter of 33 mm, not 32. Strange.
They're slightly heavier than the previous Gotion 33140 cells.


Impedance is under 2 mΩ, putting them somewhere between the Gotion 33140 and EVE IFR33140


The batteries were in storage mode at around 50% charge, and their voltages differed slightly.


I don't have a holder this size, at least not yet, so I had to buy a clamp and cobble together an improvised holder. I do have special terminals at home for a four-wire connection, like the ones on a high-power battery holder, but I need to build the holder itself. That would be more convenient and, most importantly, more stable.
I used pieces of copper-clad fiberglass board as contact plates, and connected the alligator clips on opposite sides to reduce measurement error. In practice, the only thing that affected the measurement was the contact with the battery itself.


During the preliminary charge, the batteries took 7228 and 7082 mAh, which matches the fact that the first test cell had a slightly lower voltage.


As is often the case, two charging modes are specified: standard at 0.5C and fast at 1C, but only within a temperature range of 20–45 degrees.
Nothing new here: a standard charge took 2 hours and 7 minutes.


Fast charging took almost an hour less. As usual, I recommend using the standard charging mode, as it will extend the battery's service life.


The specified maximum continuous current is 2C, or 30 A. But since the batteries were performing very well, I also tested a 40 A discharge, giving me five test modes: 0.2C, 0.5C, 1C, 2C, and 40 A, where 1C = 15 A.
The measured capacity is slightly below 15 Ah or 48 Wh, but the difference is within the margin of error. Interestingly, thanks to self-heating, the batteries delivered more at 40 A than in gentler modes, for example. At the same time, the voltage barely rises as they warm up, unlike what happens when a battery is overloaded.


Temperatures at 1C, 2C, and 40 A look very good to me. At 40 A, the temperature was just under 60 degrees.


The second sample produced similar results, so it will be easier to compare them on a single graph.


The temperature difference most likely came from the difference in room temperature.


A typical residual capacity test: I honestly forgot to run the first one, but after cooling down, the second gave up just 166 mAh, which is a very good result.


The comparison results for the two samples turned out to be a little strange, though. The second sample lagged noticeably in capacity at currents of 0.5C and above. The biggest difference was at 1C, around 2%; in the other modes it dropped to about 1–1.5%.


And that's when I realized the difference was down to the testing method, not the batteries.
The thing is, I usually run the tests like this: 1 (0.2C), 2 (0.2C), 1 (0.5C), 2 (0.5C), and so on. This time, because the holder was awkward to use, I ran all the tests on one sample first, then on the other. That would have been fine if not for the weather: the room got colder while I was testing the second sample, and a difference of a few degrees could have produced a result like this.
Also, while testing the first sample, I noticed that the plate making contact with the battery's negative terminal was getting quite hot, so I cleaned the copper before testing the second battery. This slightly reduced the battery's parasitic heating.

All of this together could have introduced some measurement error. To check my theory, I ran the test at 1C (the one with the biggest difference) under as nearly identical conditions as possible—in other words, almost back to back.

The results were much closer. The first and second samples from the original test are shown in blue and red, respectively; the second test is shown in light blue and orange.
You can see that there is still a difference, but it has decreased considerably, to about 1% between the first and second samples.

I write time and again about how important temperature is, and here I went and fell into the same trap myself. To be fair, though, in the way I usually test, temperature has less of an effect, because I'm less concerned with its impact on absolute capacity than on relative capacity between samples. My usual method minimizes this effect, but I didn't think of it this time.


And of course, let's compare them with a couple of other similar batteries. Comparison at 0.2C, 1C, and the maximum current of 40 A.
Green — the batteries under review
Blue — EVE IFR33140
Red — Gotion 33140

So EVE comes in first, the BAK batteries under review are second, and Gotion is third. The difference between EVE and the others is fairly noticeable, unlike the difference between BAK and Gotion. Though considering that the BAK batteries were discharged down to 2.0 V and the Gotion batteries down to 2.5 V, the results end up about the same.


Conclusions.
The batteries themselves are quite good. The test results show that even though their specified maximum continuous current is 30 A, they handled 40 A just fine, and the temperature stayed below 60 degrees. Their characteristics are also reasonably consistent, judging by the repeat test.
Personally, if I compare them with Gotion, I think BAK might be slightly more appealing, but the difference is so small that I'd say it doesn't really matter which you buy. BAK's only real advantage is its relatively well-known name; otherwise, they're almost identical.

That's all for today. I hope you found it useful.