The 3500mAh heavyweight division: Sanyo NCR18650GA vs Panasonic NCR18650GA vs LG MJ1 vs Samsung INR18650-35E

[thunderheart]

Hi guys! I’ve got 4 3500mAh batteries from top manufacturers and I’ve tested and compared them. I think it’s interesting not only for me, but also for you, the guy who is reading this:)

For those who prefer watching than reading I’ve made the video version of this test:


The cells were bought from Queen Battery, a Chinese supplier of genuine batteries, who specializes mainly on EV and eBike market. Queen Battery has a branch in Europe, but they also work with customers from Americas, Russia, etc…

As always, I've tested with ZKETECH EBC-A20 and a self-made battery holder. It's a PC-connected battery tester supporting 4-wire measuring and discharging at up to 20A.


I've followed all the prescriptions of the IEC61960-2003 standard concerning battery's capacity measurement. Before each discharging cycle each battery was charged at standard current mentioned in its datasheet to 4.2V (cut-off at 0.1A, which is the lowest supported by EBC-A20). Before each discharging or charging i've held a 1-1.5hrs pause. The environment temperature was 20-25°C (21-23°C to be honest).

Discharge cut-off voltage was 2.5V for Sanyo, Panasonic and LG and 2.65V for Samsung (following their datasheets). Discharging currents were 10A, 8A, 5A, 2A and 0.2C which was 0.67A for Sanyo and Panasonic and 0.68A for LG and Samsung.

I’d like to mention, that Sanyo and Panasonic are the same company now and I use different brands just to distinguish Japanese made NCR18650GA marked as SANYO from Chinese made NCR18650GA which doesn’t have brand marking but is designed in Panasonic style. So let’s start!

Sanyo NCR18650GA

This cell’s marking is SANYO L NCR18650GA 7401. The “L” means that it was made by Sanyo Energy Higashiura Co. LTD, Japan. Here is its datasheet.


The main specs:
Rated capacity: 3300mAh (0.67A discharge at 20°C)
Capacity (min): 3350mAh (0.67A discharge at 25°C)
Capacity (typical): 3450mAh
Discharge cut-off voltage: 2.5V
Max continuous discharge current: 10A
Standard charging current: 1.675A (I’ve charged at 1.68A due to EBC-A20’s resolution restrictions)
Max charging current: 3.35A
Charge end voltage: 4.20V ± 0.03V

Here are the results I’ve got:


As you can see, at 0.2C (0.67A) it gave out 3491mAh / 12.64Wh which is much more than the typical capacity of 3450mAh declared in the datasheet. The results at higher discharge rates are:
2A: 3351mAh / 11.88Wh
5A: 3323mAh / 11.27Wh
8A: 3286mAh / 10.76Wh
10A: 3295mAh / 10.60Wh

Almost 3300mAh at 10A! And the curve has no sudden falls. I think it’s a brilliant result, bravo Sanyo!

Panasonic NCR18650GA

This guy is the twin brother of Sanyo. The wrapping is in traditional Panasonic style and has “Made in China” and D 75051W markings.


The cell was made in Suzhou, China, at Panasonic’s new battery plant. It shares the same specifications and the same datasheet with Sanyo NCR18650GA, which is not a surprise as far as they have the same model number.

The Chinese NCR18650GA showed the following results:


At 0.2C (0.67A) it gave out 3448mAh / 12.53Wh - exactly the same capacity which is mentioned as “typical” in the datasheet. It’s lower than the Japanese one’s result, but still a great one. At higher rates:
2A: 3305mAh / 11.69Wh
5A: 3295mAh / 11.13Wh
8A: 3286mAh / 10.73Wh
10A: 3250mAh / 10.31Wh

The results are a bt lower than those of the Sanyo, but the difference is not critical. Sanyo’s results can also differ from cell to cell, so the main thing is that it complies with the datasheet’s typical capacity. At 8A there is almost no difference and at 10A the difference is only 45mAh.

LG INR18650 MJ1

This battery is made in South Korea (I guess) and the wrapping has the following marking: LGDBMJ11865 P274I301A1. The datasheet is here.


The main specs:
Capacity (nominal): 3500mAh (at 0.68A discharge)
Capacity (minimum): 3400mAh (at 0.68A discharge)
Discharge cut-off: 2.5V
Max continuous discharge current: 10A
Standard charging current: 1.7A
Max charging current: 3.4A
Charging end voltage: 4.20V ± 0.05V

The standard charging and discharging currents are a bit higher than NCR18650GA’s, but the cut-off voltage is the same.


At 0.2C / 0.68A the capacity was 3384mAh / 12.32Wh which is lower than the minimum declared in the datasheet. Not a good start for a heavyweight player. The results at higher rates:
2A: 3275mAh / 11.62Wh
5A: 3258mAh / 10.98Wh
8A: 3208mAh / 10.39Wh
10A: 3205mAh / 10.09Wh

I can’t call these results great or even good. Maybe they could be good if the cell was marked as a 3400mAh one, but for a 3500mAh cell it’s a fail. Surprisingly the results at 8A and 10A are almost identical.

Samsung INR18650-35E

This cell is also South Korean and it is marked INR18650-35E SAMSUNG SDI 2G25. Here is its datasheet.


The main specs:
Capacity (min): 3350mAh (at 0.68A discharge)
Capacity (typical): 3400mAh
Discharge cut-off: 2.65V
Max continuous discharge current: 8A
Standard charging current: 1.7A
Max charging current: 2.0A
Charging end voltage: 4.20V

Pay attention to the discharge cut-off voltage – it’s higher than that of others. Max charging current is only 2A which is slightly higher than the standard charging current and significantly lower than 3.4A allowed for Sanyo/Panasonic and LG. Max continuous discharge current is limited by 8A, but I have tested it at 10A for comparison.


At 0.2C (0.68A) the second Korean guy shows 3488mAh / 12.63Wh even having discharging cut-off at 2.65V! I want to remind the result of Japanese made NCR18650GA with 2.5V cut-off: 3491mAh / 12.64Wh – the difference is only 3mAh / 0.01Wh!!! Bravo Samsung! The results at higher rates:
2A: 3384mAh / 11.90Wh
5A: 3317mAh / 11.08Wh
8A: 3248mAh / 10.47Wh
10A: 3205mAh / 10.14Wh

At 10A it’s slightly higher than the LG MJ1, but don’t forget that the Samsung’s max discharging current is 8A and the MJ1 was discharged to 2.5V.


COMPARISON

Looking at the heroes one by one is interesting, but more interesting is to compare them. I will not compare at 0.2C because the “0.2C” is not the same for all 4 batteries.

So let’s start with 2A:


Samsung continues its surprises! Even with 2.65V discharge cut-off it managed to overtake the Sanyo NCR18650GA! Sanyo is the second, Panasonic – the third and LG is the last. The difference between Samsung and LG is 109mAh / 28Wh.

5A discharge:


Sanyo takes over Samsung thanks to it’s 2.5V discharge cut-off voltage. The battle between Samsung and Panasonic for the second place is lost by Samsung because although it has higher capacity (3317mAh vs 3295mAh), Panasonic has higher energy: 11.13Wh vs 11.08Wh. So Samsung is the third and LG is the fourth as expected.

8A discharge:


The two NCR18650GAs showed the same capacity, but the Japanese one has 0.03Wh more energy given out, so the first place is occupied by him. The third is Samsung’s nimble 35E and LG is… yes, the fourth.

10A discharge:


Sanyo proves that he is number one and there can’t be another opinion. The difference between Japanese and Chinese NCR18650GAs is 45mAh / 29Wh. Panasonic is the second, LG has lost the third place to Samsung despite its specs.


CONCLUSION

The Japanese made NCR18650GA’s triumph was not a surprise for me, but the brilliant show by Samsung was! Nevertheless, Sanyo NCR18650GA is the number one, Panasonic NCR18650GA is not the same but very close to its Japanese bro and is the number two. Samsung showed an exceptional performance at 0.2C and 2A but at higher discharge rates Panasonic was better. LG is the outsider of this division. It’s not a bad battery, but the competitors are better.


The prices (without shipping) at the moment of purchase were the following:
Sanyo/Panasonic NCR18650GA: US$3.70
Samsung INR18650-35E: US$3.10
LG INR18650 MJ1: US$3.10
For the latest pricelist, shipping rates and discounts drop an e-mail to wangxin570@gmail.com (Queen Battery’s sales manager). They can also make battery packs on demand.

You can also check out my YouTube channel with reviews of battery chargers and other stuff. The Miboxer C4-12 (3.0A x 4 slots) review is coming soon. Subscribe to not miss the new reviews.

I’m looking forward to read your thoughts, suggestions and questions in the comments!

[public_enemy]

Very nice!

[brumbrum]

Thanks for this.

Which one got the hottest and which stayed the coolest at 10amp discharge?

[thunderheart]

brumbrum, I didn't measure the temperature. I didn't even touch them while in progress, to not affect the results. I allowed myself to touch them only after discharging was done and they all were quite hot. I'd not recommend using any of them at 10A without active cooling.

[madin88]

Thanks for sharing.
I believe i already have seen cylce tests posted by someone in an e-cigarette board..

[chuttney1]

Link to li-ion cell data

http://lygte-info.dk/review/batteries20 ... arator.php

[thunderheart]

Thanks for sharing.
I believe i already have seen cylce tests posted by someone in an e-cigarette board..

Maybe.. I haven't seen any 3500mAh 4 battery test with comparison.

Link to li-ion cell data
http://lygte-info.dk/review/batteries20 ... arator.php

It misses Chinese made NCR18650GA, if i'm not mistaken

[DVDRW]

To see cycle test for them is my ultimate dream
So then I can decide which goes to my future i-miev 40kWh battery extener.

[thunderheart]

DVDRW, nice dream:)

[Pajda]

To see cycle test for them is my ultimate dream

I look at the cycle test results of this "holy trio" every day and yes, it's a very nice look...

btw. Great job thunderheart, your testing methodology is exactly as it should be.

[thunderheart]

Great job thunderheart, your testing methodology is exactly as it should be.

Pajda, thanks for appreciation!

[rojitor]

Nice review. Thank you.

[The Battery Doctor]

Interesting! I did a test of the same cells including the new Sony US18650VC7, it was published 5 dec. 2017. However i used 40 cells from each brand, as you can see in my data one cell is NOT ENOUGHT for any kind of data cells the capacity can vary between 200-360mAh in brand new cells. If you look at my graph at ca 4:00 you can se that ANY CELL could have won if just used ONE sample!

[youtube]https://www.youtube.com/watch?v=Qpm91Itv6k8[/youtube]

Also Sanyo and Panasonic is NOT THE SAME COMPANY! They are different companies but Panasonic OWNS Sanyo! Its a huge difference! Sanyo cells are also made in Japan but at different factories, except for some new cells (2017 with other heat shrink plastic, like the one you got).

Your test is internering BUT not scientific since you used just ONE sample. The result is more or less RANDOM. Please use at least 10 cells in further test,please. And DONT call sanyo cells panasonic! They are NOT the same! Both GA cells are made by SANYO but the china version is made in china!

[thunderheart]

rojitor, thanks for appreciation!:)

The Battery Doctor, I've stopped watching your video when you said you used LiitoKala Lii-500 as tester . You didn't follow the prescriptions of the standard, you didn't even follow the manufacturers' prescriptions. You can use 40 cells or even 400 cells of each brand but if your testing equipment is LiitoKala Lii-500, then your results worth NOTHING. And you call my test "not scientific"?

Panasonic has bought SANYO and owns its thechnologies and rights to use them in his own production (which he does). The plant in Suzhou, China belongs to Panasonic, not to its SANYO branch, so the second GA is Panasonic, not SANYO, and they bear the same model number and are sold as just NCR18650GA, which proves that they are the same company with factories in different parts of the world.

And please stop spamming. You've copied your post under my video on YouTube for what? Are you going to send me text messages, e-mails and letters with the same text?

P.S. If you have really bought 40 cells of each brand then you can afford buying a normal testing equipment i think. Throw your Lii-500 out and use something serious before calling others' tests "not scientific"

[rojitor]

The battery doctor.... you are kinda right. You got a legit point. Nevertheless I must disagree. Correct me if I am wrong but afaik all cells have a quality test and those under the standards are discarded and sold to third party companies. They are rebranded as whateverfire and sold for cheap.
So "maybe" I repeat "MAYBE" testing more than one could help. If you need 10 cells to test a top class brand....That brand Is just crap.

[999zip999]

Wow the heat is on. So at what rate would you use these cells at in an ebike pack for decent life ?
10amp ? 6amp ?
Or the cells hotter then this conversation ?

[thunderheart]

So at what rate would you use these cells at for decent life ?

No matter at what rate you gonna use them. The only thing which matters is the quantity of cells in the pack. If you are going to buy a couple of cells for a powerbank, then any of them would be good. If you gonna build a pack of 100 cells, then the difference will be noticeable.
For example the 0.51Wh difference between Sanyo and LG @10A would result in 51Wh in the pack, which means 5 batteries. So instead of 100 Sanyos you would have to buy 105 LGs.

[Joachim]

The temperature have big effect on the energie(wh)you can take out of a cell.

i like to see how meny Wh,s these cells give@ 30/40 degree celsius

[999zip999]

I want a 72v 20ah pack @ 70amp . would like it to fit in a em3ev triangle bag. So 10p I would need or 12p. This is getting big. 240 cells. I think samsung 25r. (25a cell )for less heat and less parallel for 70amps. But a bigger battery.
I do want 800 cycles.
Meaning I need cycle life and low sag.

[thunderheart]

Meaning I need cycle life and low sag.

What about Boston-Power Swing? It's an impressive cell in terms of cycle life.

[999zip999]

I think heat is the problem for cycle life and sag. So is a 25amp cell better suited for higher amp ebike pack.. If I had a emoto for 40ah of battery. I see this. Always a balance and a compromise. Till we get the 5 volt cell.

[Pajda]

The heat generation and voltage sag are caused by the cell internal resistance (IR). You can choose two ways how to deal with this issue. You can use high power (HP) cells with low IR such as your mentioned Samsung 25R. Then you can operate at 3-5C average discharge current without significant effect on cycle life. Or you can use high energy cells (HE) such as this three cells but you need to stay at 1C max average discharge current for good cycle life. The average discharge current already counts with using short (up to 15sec) peaks with maximum allowable current per cell.

However, for cycle life, no matter if HE or HP cell is used, the most important parameter is average Depth of Discharge. For example I measured when only 50% DoD is used (in the range 100%-50%) all of the modern cells, again no matter if HE or HP, loss only about 7-10% of its nominal capaity after 1000cycles! But when you are using 100% DoD most of the HE cells are practicaly dead after 400-500 cycles. HP cells can do 1000 cycles with about 70% of remaining capacity.

Also when using a low average DoD you can allow much higher average currents or using 1C fast charging without significant effect on cycle life.

[999zip999]

I was looking at this pack http://powervelocity.com/index.php?id_p ... er=product as it made of the panasonic cells. Not to many 72v volt 20ah packs on the market. Usa base.

[madin88]

Heat basically leads to "faster chemical reaction" which increases the performance.

-> less internal resistance, so less voltage drop and more Wh

If you going to abuse your battery, it makes a lot of sense to warm it up (RC model pilots/drivers usually doing this with the LiPos for records and such extreme things).

The bad side about a battery that is heated is that aging (calendar life) will suffer ALOT.
Thats also at least one big reason why the cycle life gets so worse if you use the cells at specified C rate (like 20A on a 25R or 30A on Sony VTC5) because the cell becomes VERY HOT at the end and this hurts.
It's usually the internal resistance which will get worse and worse than..

For long storage it would be optimal to put the battery in the fridge, and during usage around 20°C.

[youyoung21147]

Perform the same test after 300 cycles and you will be surprised with the results !

GA is excellent in the first 50-100 cycles but the internal resistance rises so quickly that even an old 18650PF beats it after that.