Test/review of Liter INR20650 3400mAh (Blue)

Liter INR20650 3400mAh (Blue)

Official specifications:

  • Model Number: NCR20650

  • Nominal Capacity: 3400mAh

  • Nominal Voltage: 3.7V

  • Charge Method: Constant Current Constant Voltage

  • Charge Voltage: 4.2V

  • Charge Current: 2065mA

  • Weight: 58g

  • Size:20x65mm

  • Made in: Korea

A 20650 battery with not that much capacity for the size, how much current can it deliver?

These discharge curves do not look very good, the batteries are a bit mismatched and they can barely deliver 7A.

At least they stay fairly cool.


When batteries get larger I generally expect more capacity or more current, or even both, but this battery has less of both. A good 18650 has the same capacity and can deliver more current

Notes and links

How is the test done and how to read the charts
How is a protected LiIon battery constructed
More about button top and flat top batteries
Compare to 18650 and other batteries

Someone probably made a mistake mixing the chemicals :FACEPALM: :smiley:

Very bad performance for this size…

Thank you HKJ!

Not bad performance at all actually.

This is a Chinese 20650. Its energy density is actually quite good at 200Wh/kg. Not as good as the big boys with 18650s at 250Wh/kg and 21700s at 260Wh/kg, but still great energy density.

The main attraction of this cell is that it is very inexpensive., at 1,87$ if you buy 100 pcs, and even less if you buy more.

I’ve actually tested mine, and it got similar results to HKJ’s testing, except I went down to 2,5V, and got the rated capacity of 3350mAh at 0,2C.

Its internal resistance is a bit lower than HKJs though, which I find strange. I can actually draw more current from them before the Vdroop gets too high, indicating HKJs perhaps got a lower binned sample than mine.

Where to buy at that price? Does your figure includes shipping costs? ( ͡° ͜ʖ ͡°)

Regarding internal resistance differences, what you say only reaffirms the lack of consistency problem. If building a large battery pack it is advisable to bear in mind overall performance figures dependant on the weakest links in the chain, or make an effort to classify the cells via a controlled, timed in series discharge (or other valid method) to attain an ordered listing and then balance - combine them to uniformize the series battery pack chain links. This can add a bit of cost.

Cheers :-)

Depends on where you live.

Since I live in Canada, and want to build a bigger capacity Ebike battery soon enough before winter, I decided to ask the nice folks at the manufacturer directly, and asked for 200 pcs.

Here is what they said:

Thanks for your inquiry of our 200pcs of these FST 20650s.

Its unit price: USD2.8/pcs free delivery with Fedex express to Canada.

200pcs total cost:USD560;

Best Regards.

Did you went that way already? 200 pieces is a large number but those cells' specific power is rather low at ≈20W max. Makes me think you aim to build some sort of small electric scooter (2KW power?) with good range.

Do you have an spot welder at hand? ( ͡° ͜ ͡°)

Cheers ^:)

No. But I have beryllium copper springs on hand now ( ͡° ͜ ͡°), with prototype battery holders.

I just need to make another big spring batch since I don’t have enough large springs anymore :smiley:

It is true that they are not famous these 20650s, we would expect much better results on this cylindrical battery format, we can see that its internal resistance is catastrophic and that from one cell to another their capacities vary enormously.

As far as I am concerned, I acquired three Chinese-made 20650-format SZNS cells, which have similar specifications on paper to those tested by HKJ, ie: 3350mAh of typical capacity and a maximum discharge rate of 2C officially therefore 6.7 Amps maximum theoretically, but after several tests they have proved much better than indicated in their official specifications. Because I get the following results:

Measured capacities (Discharge 4.2v-> 2.5v@0.2C (670mA)):

  1. 01: 3465 mAh
  2. 02: 3476 mAh
  3. 03: 3445 mAh
    note: their actual capacity are rather regular

Internal Resistances (measured in ACIR @ 1kHz on YR-1035)

  1. 01: 35.0 milliOhms
  2. 02: 33.4 milliOhms
  3. 03: 34.7 milliOhms
    note: here too it is rather regular and especially not too high for a cell having officially, a maximum discharge rate of 6.7A (besides it is these results of IR that pushed me to test levels of discharges higher on these cells).

Maximum temperatures reached with a discharge rate of 4C (13.4 Amps)
(measurements with SENFU probes A10b (plated probes directly in contact with the center / middle of the metal cylinder of each cell)):

  1. 01: 63.2 ° C
  2. 02: 64.9 ° C
  3. 03: 62.7 ° C
    note: after I did not want to test higher in discharge by fear greatly reduce their life cycles, but it’s already not bad for cells that have official specs of 6.7A maximum discharge to reach double ( 13.4A) without incident.

I put official photos this cell model SZNS 20650, because mine have wraps in sorry states (I absolutely must rewrap them ;-))

That is good energy density and good price, but as others have said, not ideal because of the high IR and questionable quality control. This won’t help you now since you already purchased them, but the shockli 5500mAh 26650s (PLB 55A) have good energy density and power density (and good documented cycle life in the datasheet) and also pretty good price. When I built my ebike pack I bought 24 of them from shockli at $4.89 each, but with expensive shipping at $55.

A lot better figures than those of these 20650 Liter energy cells. Your measured temperature delta at 13.4A discharge is nice & moderate, less hot than a high capacity 18650 when discharged at 10A. Their DC IR must be close to your measured AC IR numbers. Their maximum specific power is for sure more than double that of Liter 20650s (45+W). Nice find if the price is right.

Cheers ^:)

@EasyB, never said I actually purchase 200pcs of the stuff.

I said I inquired a price directly from the manufacturer.

I initially found 24x of the cells inside of the powerbanks, and all of their internal resistance were about the same, along with capacity.

For the price, I do not know, it’s an Australian contact who sold me for about 3usd / piece and I had more expensive shipping than batteries, but the curiosity has taken over the reasonablity at the time (lol), I do not know how much he had paid or even how he got them, I searched everywhere on the net and I found no place where he sells this model, even if they are better than the “liter 20650”, it is not worth the quality / performance of Sanyo or LG, moreover it’s been more than a year that LG 20650 “HH2” is expected but still nothing available today, they are supposed to give 3500mAh typical capacity and 20 amps continuous. As well as an LG 20650 “M42” giving 4200mAh typical capacity and 15 continuous amps, and still nothing available. After I heard several rumors that are substantially all the same: like what the 20650 is a format that would sell rather badly. It’s sad because I think it’s a good alternative to the 26650 format.

If planning a big battery pack build one potential problem with big brand batteries is price and availability. On the other hand there are some lesser known OEMs making decent cells with good availability and price. PLB and Lishen are two of them. Lishen makes the LiitoKala rewrapped Lii-40A, a run-of-the-mill 3C max 4000mAh 21700 cell (HKJ reviewed). It could be interesting given its low price (AliExpress 8 piece pack example). Maybe in Alibaba better deals can be made when going big quantities and/or even buy first brand OEM cells, but no experience in this respect.

Cheers ^:)

Ah, I see. Well, the shockli cells have worked great for me.