HAKADI 26700 3500mAh, Na-ion batteries

Na-ion batteries based on a relatively new chemistry (though that depends on how you look at it) have finally made their way to me. I bought them specifically for a review and to satisfy my own curiosity, and now I can share the results of some tests with you. I think you'll find them interesting.

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 my account

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


I'll start with the order. At first I wanted to order the popular 18650 size, but first, their capacity is rather low, and second, I found test results for these cells, so I went for the larger 26700 size. The 3500 capacity also seems more useful to me. The store offers several lot options; I chose the smallest one, 4 cells, which cost around 16 dollars. If you add them to your cart, Ali will later show them in the coins section with a good discount.

Right after I placed the order, the seller messaged me saying it would take a long time to arrive. I replied, “OK, I'll wait.”
By the way, there's a warning about this on the store page. In short, don't worry, you'll get your order; just wait.


To be more specific, you'll be given a tracking number, and after a while you'll receive an envelope or postcard, most likely at a Ukrposhta branch.
In my case, the envelope contained a nut, though there have been cases where people received just a postcard. It depends on the store.


But after a while, the parcel will show up in your Mist Express app (if you have one), usually from Trade44. That's your batteries, or devices containing batteries.
Anyway, I picked up the package containing my test subjects at a Mist Express branch.


A little about this type of battery. They're still new in the battery world, but according to Wikipedia, they've actually been around for many years. At some point, though, it was decided to develop the familiar lithium-ion batteries instead, and these were “put on the shelf.”



Key features include:
1. Price. Since these batteries contain no lithium and soda is cheap, they should eventually cost less, but that will require large-scale production.
2. A wider low-temperature operating range. These batteries are said to perform better in the cold.
3. They can be discharged to zero, so they effectively don't need over-discharge protection.
4. Greater safety, again thanks to the absence of lithium and their ability to be discharged to zero.

There are drawbacks, too:
1. Capacity is roughly on par with LiFePO4, so for maximum capacity at minimum weight/volume, conventional Li-ion is still the better choice.
2. The operating voltage range is too wide: from 1.5 V (discharged) to 3.95 V (fully charged), i.e. more than a twofold voltage swing. For a nominal 12 V battery, this would look like this:
“Conventional” Li-ion, 3S: 7.5–9 V to 12.6 V
LiFePO4 4S: 10 V to 14.6 V, with the operating plateau actually between 12 and 13 V.
LTO 6S: 9 V to 16.8 V
Na-ion 4S: 6 V to 15.8 V.

Such a large voltage swing requires a compatible load. In other words, your load must operate correctly across this range, which isn't always convenient and is unlikely to suit typical devices.



I couldn't find a datasheet; the only information available is on the product page:

Battery type: Sodium battery
Nominal voltage: 3.1V
Standard capacity: 3500mah
Weight: 82± 50g
Size: 26.4*71mm
Charge voltage: 4.1±0.05V
Discharge cut-off voltage: 1.5±0.05V
Internal resistance: ≤20mΩ
Standard charging current: 0.5C
Standard discharge current: 0.5C
Max. Charge Current: 1C
Max. Continuous Discharge Current:3C
Temperature of discharge: -30~60℃


In my opinion, the specified charging voltage is incorrect. 4.1 V is the maximum; the end-of-charge voltage for batteries like these is 3.95 V. Be careful!

They look like ordinary batteries, with very sparse markings and a flat positive terminal.


The only difference is the slightly less common size: 26700 rather than 26650, but that's not important. I had four cells, so I tested all of them, though I ran the first one through a wider range of load currents.


The product page has a photo showing an impedance measurement.


But my measurements gave a different result—and the batteries in the photo above are clearly in the same condition as mine, judging by the cell voltage. It seems to be within the stated limits, but it's still disappointing...


Let's start with the basic tests, beginning with a conditioning cycle that consisted of these steps:
1. Discharge at 1C
2. Rest for 5 minutes
3. Charge at 0.5C
4. Rest for 5 minutes
5. Discharge at 1C
6. Rest for 5 minutes
7. Charge at 0.5C

When taking the screenshots, I forgot to scroll the results log on three of them, so the first discharge test results aren't shown. In reality, though, the difference was minimal: the batteries initially had about 10% charge, meaning they were almost discharged.



Two charging modes are specified for these batteries: standard at 0.5C and fast at 1C.
I'll note right away that the CV phase is relatively short, more like what you'd see with LiFePO4 batteries, while the CC phase graph is more linear, which is more like “conventional” Li-ion. This makes it possible to roughly estimate the state of charge from the cell voltage. As a reminder, that's a problem with LiFePO4.



The battery also handled a 1C charging current just fine. Nice.


Now for the discharge test.
The specified discharge current is up to 3C continuous and up to 5C for short periods. In general, these particular batteries are most likely what are known as high-capacity cells, since high-current Na-ion cells are rated for discharge currents of up to 20C.

Still, I ended up testing the first pair not only at 0.2, 0.5, 1, 2, and 3C, but also at 4 and 5C, since the temperature stayed well within acceptable limits.
Unfortunately, I never saw the claimed 3500 mAh; the maximum was around 3150 mAh.


Temperature at discharge rates of 2, 3, 4, and 5C. I think the claimed maximum of 3C is specified with a little margin to spare.


The second cell showed roughly the same results as the first.


Temperature at 3, 4, and 5C


I tested the third and fourth cells at currents up to 3C. I saw no point in going higher, since the tests on the first pair were more than enough, especially as those were outside the specified operating conditions.
By the way, the capacity here is even lower, around 3100 mAh.


Temperature at 2 and 3C


The fourth cell came in at roughly the same 3150 mAh as the first.


Temperature at 2 and 3C


As usual, I measured the remaining capacity after testing at the maximum current and letting the cells cool down.
The first pair was tested at 5C, and the second at 3C, but the results were roughly similar: 100–140 mAh, which is quite good.


I checked whether the results were consistent at 0.2C and 2C. Overall, not bad, but the capacity was disappointing: the minimum at 0.2C was 3095 mAh, and the maximum was 3144 mAh :(


During testing, I charged the batteries on the second tester to speed things up. It doesn't really matter, but at one point I charged a battery and left it in the tester for a relatively long time. Then I started charging again and it "put in" another 274 mAh, which puzzled me a bit.
I was also puzzled by a phrase in the Wikipedia description of these batteries:



All this prompted me to run an additional test: measuring capacity after several days in storage, specifically after about 60 and 65 hours.
The batteries were fully charged, and I ran the first test 6 hours later, then the second one another 65 hours after that.
For batteries 1 and 3, I checked how much charge they would "take in," while for 2 and 4, I checked how much they would "give out."

But no, everything is fine: after 60 and 65 hours, the batteries had taken in only about 30–45 mAh, which is very good.


The discharge test also showed a difference of only about 70–75 mAh. I compared the earlier results with the initial ones, taken when the batteries were almost "fresh out of the box," and the latest results, taken after all the tests, including discharge at 5C.

The point is, I thought these batteries might need to be "broken in," but it turns out they don't—just like any other Li-Ion batteries.
By the way, I've said this many times: if someone tells you your new Li-Ion battery will "break in," don't believe them. They're just trying to make money off you.



For a more extensive test, I decided to check the batteries at low temperatures. More specifically, I put them in the freezer and accidentally left them there for about a week—which is even better, as it more closely reflects real-world use.
The impedance went up a little, as expected, though not by much in my opinion—only about 10–15%. The voltage, however, stayed at around 3.90 V.


I tested the batteries at 0.5, 1, 2, and 3C. First, I tested battery 1 at 1C, battery 2 at 2C, and battery 3 at 3C. Then I decided to test at 0.5C rather than 4C, so battery 4 was tested at 0.5C.
You can see some steps in the graph; unfortunately, I noticed them later. Since the computer and the freezer are in different rooms, I had to control the load and log the data using a tablet. It's an old one, and the software would occasionally freeze for a bit before continuing. This introduced a small error in the Wh measurement, though it wasn't critical and didn't affect the Ah measurement at all, so I didn't rerun the test.

By the way, I ran into a problem with the tablet that I'd encountered before but had forgotten how to solve. I installed the load software and ran the tests, saving the test files to the C: drive. Then I went to look for them, but they weren't there. When I opened the software, the files were visible in it. This isn't a hidden/system files issue, but a file permissions issue. Unfortunately, it was easier to copy my files onto the tablet and generate the comparison graphs there. I couldn't figure out how to retrieve those 4 test files, even after trying various options :((((



But it's more interesting to compare the results with the initial ones. The measurement results are shown below the graph, ordered from the lowest current to the highest; the first measurement was taken at room temperature.
Of course, the batteries also warmed up in the freezer, which is why I decided to test at 0.5C: at that rate, the battery heats itself less, so the effect is smaller.

In my opinion, the batteries perform pretty well at low temperatures. At 0.5C, the capacity difference was about 8–10%, and the energy difference was about 15%. That's with a temperature difference of around 40 degrees.




As I wrote, the specs said the batteries could be discharged "to zero." Well, let's test that.
First, I discharged one battery using the load, but its minimum setting is 0.1 V. After disconnecting the load, the voltage returned to 1.5 V fairly quickly. The battery also delivered about 10% more than its rated capacity.


Naturally, that wasn't good enough for me, so I simply shorted it out and left it that way for about 15 hours. A bit brutal, but what can you do? :)


After 15 hours, I connected it to the load in monitoring mode. It can't measure below 0.1 V, but I think you can still see that the voltage started rising again. After 50 minutes, it was already around 1.16 V.


I wasn't sure whether the voltage was dropping across the load's own input resistance, so I also measured the impedance and voltage using a separate instrument and a multimeter. After almost an hour, I got a reading of about 1.25 V. But the impedance was even more surprising: it was almost the same as at the very beginning, when the batteries had about 10% charge.


The battery was then fully charged at 0.5C, i.e. in the same mode as before. The charging did take a little longer, but that’s understandable.


Then I discharged it at 0.5C and plotted the results alongside those from a standard 0.5C test. The difference was within the margin of error.


Next, I wanted to compare the discharge curves of different battery chemistries.
1. Sodium-ion
2. Lithium iron phosphate
3. “Regular” lithium-ion
4. Lithium titanate


But since the batteries above had quite a wide spread in capacity—especially the lithium titanate ones—I dug through my test archives and also used another person’s test results to plot batteries with more similar capacities, around 1500 mAh.
The order is the same:
1. Sodium-ion 18650
2. Lithium iron phosphate, 1600 mAh
3. Lithium-ion, 1500 mAh
4. Lithium titanate, 1500 mAh.

The discharge curves are quite different, especially for the lithium iron phosphate cell.


The sodium-ion cell data above came from an interesting website. The author not only tests batteries but also shares an archive of the test logs. Since I have a load from the same series, I can use the data for comparison plots.
They tested 18650 cells rated at 1500 mAh, although for some reason they were tested as 1300 mAh. I think the author measured their actual capacity and used that as the C-rate reference; I usually use the rated capacity.

The site also has tests of a 12 V pack and 50 Ah sodium-ion cells, along with some conclusions. I recommend giving it a read.




That’s all for now. I may run a cycle-life test later, but I haven’t decided yet.
My personal take: these batteries are interesting, but very niche. If they do become widely available and their price comes down, and inverter manufacturers start making devices that support such a wide voltage range, they could become a compelling option.
For now, they’re just an interesting toy for testing. Their advantages include:
1. Better performance at sub-zero temperatures, which can be critical in some cases
2. No need for over-discharge protection, and the batteries themselves are tolerant of over-discharge. So far, only nickel-cadmium batteries could be stored fully discharged (down to zero).
3. Greater safety, though only time will tell.

What bothers me is the wide operating voltage range. A nearly threefold difference between fully charged and discharged is a lot.

That’s all for now. I hope you found this useful. I’d be very happy to hear your questions or just your comments. I may add to the review later :)