3500-mAh NCA cell, Samsung, LG, Sanyo, Panasonic

[zener]

LG MJ1 had the throne for the whole 5 days. Meet the LG MJ1 killer, the Panasonic/Sanyo NCR18650GA:

Panasonic/Sanyo NCR18650GA: 3.4Ah and 11.2Wh at 10A (down to 2.5V), 49.5g (47 grams measured). Higher voltage than LG MJ1 through the whole range. Things are moving fast those days! https://yunergybattery.wordpress.com/2015/05/29/2015-new-sanyo-ncr18650ga-vs-lg-inr18650-mj1/

Todays top 5 ebike performance cells at 10A continuous discharge:
1. Panasonic/Sanyo NCR18650GA, 3.4Ah and 11.2Wh at 10A (down to 2.5V), 47 grams measured. 238Wh/kg at 10A. Measurement is here: https://yunergybattery.files.wordpress.com/2015/05/10a-vs-lg-mj1.jpg

How can he measure 3381mAh @ 10A ! while spec. says 3350mAh @ 0.67A with a 38mOhm cell?
He also measured 3371mAh @ 1C.
Did you have seen that this https://yunergybattery.wordpress.com/ have only this cell test?
We need to know the mOhm from all these cells than we have something real to compare.

 

[riba2233]

Riba2233: Thanks, that LG HG2 cell is very interesting!

I was looking for a reliable 10A charts for BE and BD cells, but could't find any. Assumed that 35E is sucessor of 32E. 32E 10A graph is missing Wh data:

As I've already said, dampfakkus's 32E and 35E 10A charts are wrong!!


Here is lygte's test with real data:

http://lygte-info.dk/review/batteries2012/Samsung%20INR18650-32E%203200mAh%20%28Cyan%29%20UK.html

Here is a chart for Panasonic BE, although a bit older one (just repacked, but that's the same cell). Dampfakkus also has test of BD cell, but again 10A data is wrong.

http://www.dampfakkus.de/akkutest.php?id=531

He even has 20A continuous test of BE cell lol :

How can he measure 3381mAh @ 10A ! while spec. says 3350mAh @ 0.67A with a 38mOhm cell?
He also measured 3371mAh @ 1C.
Did you have seen that this https://yunergybattery.wordpress.com/ have only this cell test?
We need to know the mOhm from all these cells than we have something real to compare.

Because it's only minimal rated capacity, that's just worst case. Typical is 3450 mah. Also, at higher discharge rates curve flattens and you get more Ah, but less Wh. So, usually cells has most Ah at really low discharge rate, than least ah at ~1C rate, and than it raises again at higher discharge rates. You can see that on lygte's 32E graph. Wh are dropping with higher C rate, but Ah doesn't behave like that.

 

[fellow]

Do you know the temperature of Panasonic's 18650PF under 10A dc?

I've tested them at 10A once, NCR18650PF were cold to the touch. Even shorted, never managed them to heat up. Panasonic clames that their PTC element breaks the circuit at 50'C, so according to their pdf you can't force them hotter than that. To me 50'C sounds a bit low. Riba2233 can insure that Croatia got 40'C all the time during sommer days, but I guess Panasonic never lie... .

Riba2233: Yes, Lygte got a lot of nice tests. I just hope he soon starts measuring down to 2.5V instead of 2.8V.

 

[okashira]

Do you know the temperature of Panasonic's 18650PF under 10A dc?

I've tested them at 10A once, NCR18650PF were cold to the touch. Even shorted, never managed them to heat up. Panasonic clames that their PTC element breaks the circuit at 50'C, so according to their pdf you can't force them hotter than that. To me 50'C sounds a bit low. Riba2233 can insure that Croatia got 40'C all the time during sommer days, but I guess Panasonic never lie... .

Riba2233: Yes, Lygte got a lot of nice tests. I just hope he soon starts measuring down to 2.5V instead of 2.8V.

You must be mistaken. A PF will do 60C+ on a full 10A run, and shorted you could make one hit thermal runaway (130C+)
They also don't have a PTC. None of the cells we're discussing do.

 

[fellow]

Red line=two pf cells. Ignore the black lines. Panasonic pdf states clearly 50'C. It may be mistake, 80'C PTC feels more realistic. PTC is very cheap, it costs next to nothing. I've tested them on the bike, 13s2p and controller limit set at 21A, brake on, full throttle for 5-6seconds. No damage to motor nor batteries, no heat. Clamp meter showed 10A+.

I did several short tests at 5-6 seconds, but not on the same day. It was the 13s2p battery pack I've tested, not the single cell. It sparked blue plasma everywhere, plasma sound loud as hell. 400V AC shorts are nothing compared to 48V DC. Tested different fuses, car mini fuses arced without braking the circuit. One car fuse exploded, I was wearing a protective equipment. 6mm*30mm 125VDC ceramic fuses worked flawlessly. I believe that cell can get hotter when shorted longer/several times, my test was probably too short. After all, a lot of energy is released and cells got internal resistance that makes things worse.

 

[okashira]

Red line=two pf cells. Ignore the black lines. Panasonic pdf states clearly 50'C. It may be mistake, 80'C PTC feels more realistic. PTC is very cheap, it costs next to nothing. I've tested them on the bike, 13s2p and controller limit set at 21A, brake on, full throttle for 5-6seconds. No damage to motor nor batteries, no heat. Clamp meter showed 10A+.

I did several short tests at 5-6 seconds, but not on the same day. It was the 13s2p battery pack I've tested, not the single cell. It sparked blue plasma everywhere, plasma sound loud as hell. 400V AC shorts are nothing compared to 48V DC. Tested different fuses, car mini fuses arced without braking the circuit. One car fuse exploded, I was wearing a protective equipment. 6mm*30mm 125VDC ceramic fuses worked flawlessly. I believe that cell can get hotter when shorted longer/several times, my test was probably too short. After all, a lot of energy is released and cells got internal resistance that makes things worse.

None of that is evidence of a PTC... all the new cells have no PTC.
In fact many believe the NCR18650PD/PF is really just the old NCR18650 with PTC removed with minor tweaks.

Doesn't matter that the PTC is cheap, problem is it adds resistance and reduces performance of the cell.

 

[fellow]

Proof that they use CID (Current Interrupt Device) is written at page 3 in this document: http://www.rc-electronic.com/downloads/pdf/infos/datasheet_LiIo_Panasonic.pdf
It works indirectly, high temperature increases the pressure inside, which blows off the connection and breaks the circuit. In essence, it is PTC with less resistance. CID is the correct technical term, not the PTC as I mistakenly called it many times before.

CID/PTC is visible at the top of the cell as a strange buldge. You can see when CID breaks the circuit in the panasonic pdf graph when things get wrong. PF is PD with safety layers added, not removed. That is the theme through the whole pdf they published. CID is moved to the top of the cell.

CID explained: http://batterybro.com/blogs/18650-wholesale-battery-reviews/18306003-battery-safety-101-anatomy-ptc-vs-pcb-vs-cid
It says "Collocated (placed side by side) with the PTC" in the link above. The very same document states that "18650 batteries sold in the US are required to have CID and PTC protection. " I'm sure that the rest of the world (or EU at least) mandates the same.

CID is the reason why I always use cells in series. Cell1+Cell2+Cell3+...+Cell13. If one of 13 CID blows, all 13 cells in 48V battery are safe. CID can fail 12 times without any consequences, one sucessfull CID is enough to protect the whole string. I even recommend fuselinks between the cells, just because we can

Okashira: This one may interest you: "What kind of battery does TESLA use? Tesla uses 18650 batteries but has modified them. They have taken out the PTC and CID protection circuitry and made them truly bare-bones. Instead of relying on these protection devices, TESLA has made their own out of a type of foam that floods the battery module and prevents fire."

This means that tesla 18650 cells can reach 130'C, and NCR18650pf 50'C.

 

[okashira]

Ahhhh now i see where we are confused.
The PTC and CID are two different things. They are not related.
None of these cells have a PTC. They reduce performance, because it's a temperature sensitive resistor in series with the cell.
The CID is basically an internal pressure tripped fuse.
When the cell is extremely hot and pressure rises (electrolyte boiling) the CID [strike]will[/strike] can cut off the cell, permanently disabling it.

I have tested several PF's and had them over 50C many times.....

If you overcharge a cell, it will also release gasses and trip the CID, not related to temperature.

I have not tested if a Model S cell has a removed CID. I guess I could overcharge and find out.

Another distinction, PTC's are always there and don't disable the cell. A tripped CID and the cell is trash.
But PTC's are really not used at all anymore. That is one reason why the NCR18650A and NCR18650B suck. The PTC keeps them from "opening up," at higher temperatures like most Panasonic NCA cells do.
I know Model S cells DCIR drops by a pretty staggering 40-50% when hot.

 

[Architectonic]

What kind of current would people recommend for these 3.5AH cells for ebike applications, given it is probably not a great idea to run them at 10A?

I'm guessing these packs will be rather large/expensive for something like 80A. 15p?

 

[riba2233]

10A is not a problem with some kind of cooling... But even without cooling you can do 5A continuous and 10A peaks.

 

[TheBeastie]

I have on my 'things to do with ebikes list' to build a 26650 pack one day, but because the major brands don't really do 26650 sized cells I been looking over some of the more dubious brands.
Most of the time it just seems retarded to consider any 26650 cells because comparing the prices of a 18650 3500mah Samsung to the 26650s the prices are no more cheaper then a proper well respected brand cell, even if you choose to believe some of the dubious capacity claims of the greyish brand 26650 cells.
https://www.fasttech.com/products/1/10002180/2763200-authentic-soshine-26650-3-7v-5500mah-rechargeable

I think this Soshine cell might be my first contender though for a 26650 cell, fasttech even have tested the mah capacity them selves..that being 5298.3mAh at 1a discharge though so not a heavy duty test but I only aim to build with a lot of cells in P anyway..
Edit, hhmm just noticed there protected so might be able to draw much more then 1amp then?

 

[atomek1000]

Okay, i got a 35E cell and did a test. Unfortunately it confirmed previous "dissapointing" results.
I tested this cell fully charged to 4.2V:
10A current
2.5V cut-off
I measured 0.15V voltage drop across measurement cables so take that into consideration.
Here is a graph i got from LogView via Raytronic c60 charger.

Zeit Voltage [V] Current [A] Charge [mAh] Power [W] Energy [Wh]
15m 44s 2,497 10,018 2608,000 25,015 7,955

After a discharge a cell jumped up to 3.5V so it should have still plenty of capacity.
Cell after ~5minutes was very hot. Almost untoucable so i pointed a little fan on it because i was afraid of exploding.
Internal resistance is about 50mOhm(probably little less due to cables).

I also did tests at lower current and with 3A and 2.6V cut-off it gave astonishing 3425mAh and 3.4V avg Voltage. Barely warm without any cooling.

I am also attaching a txt with full log.

I can test panasonic 18650PF at 10A if you want to compare.

 

[riba2233]

Yeah, that is kind of disappointing. LG mj1 should be better, and especially sanyo GA. You can compare with PF if it isn't a problem

 

[markz]

So where would you put the GA?
I just read BL is 7Acont.
25r, 20Acont., 50A(5s) and 100A (1s)

How truthful are these claims?
All from http://ru.nkon.nl website.

 

[okashira]

I think the ga is the best performer.

 

[markz]

15M looks promising, 25A, 1.5Ah though, for $5.90
http://www.batterysupports.com/sams...-25a-1500mah-36v-liion-battery-cell-p-21.html

 

[riba2233]

It has funny energy density compared to today's cells, and costs the same. Samsung 30Q has double the capacity, and costs the same. But ok, it has a little bit higher IR, but has much better chemistry an will last longer.

 

[madin88]

15M looks promising, 25A, 1.5Ah though, for $5.90
http://www.batterysupports.com/sams...-25a-1500mah-36v-liion-battery-cell-p-21.html

what would be the reason to go with this cells?

 

[TheBeastie]

But look at the price! It won't see many uses at that price...at least the LG is more reasonably priced.
http://ru.nkon.nl/rechargeable/18650-size/sanyo-18650ga.html

Cheers

Yeah never noticed the 3500mah Sanyos till now.
They are on fasttech as well and I think might technically be cheaper there, maybe they held out longer before getting some stock in and got a better price.
https://www.fasttech.com/products/1420/10010486/2399401-authentic-sanyo-ncr18650ga-3-6v-3500mah

 

[riba2233]

Fasttech is more expensive, remember you have to add vat! Also nkon can probably give you volume discount, just ask.

 

[TheBeastie]

Fasttech is more expensive, remember you have to add vat! Also nkon can probably give you volume discount, just ask.

Theres no import charges or costs/tax of any kind when buying anything overseas that costs under $1000 to Australia.
One of the rare few perks of living in Australia..
It wasn't until recently I finally understood why there is a estimated value type thingy written on my packages from stuff I buy from overseas its just habit sellers have for stuff as its required when they ship it else where in the world.

 

[markz]

15M looks promising, 25A, 1.5Ah though, for $5.90
http://www.batterysupports.com/sams...-25a-1500mah-36v-liion-battery-cell-p-21.html

what would be the reason to go with this cells?

I dunno, just look at price primarly, didnt really realize its a wimpy 1.5Ah. Probably voltage sag etc.

The GA's from fastech are $28 each! WOW!

The Sanyo NCR186650GA cells are good at 72V 10.2Ah and 30A discharge. Cost is around $400.
I can buy a 25r pack, 72V 7.5Ah 60A for $476, add in a charger $80, so $550 shipped from a reputable seller. No need for DIY. Comes to about same price as cells from Nkon website.

 

[riba2233]

Fasttech is more expensive, remember you have to add vat! Also nkon can probably give you volume discount, just ask.

Theres no import charges or costs/tax of any kind when buying anything overseas that costs under $1000 to Australia.
One of the rare few perks of living in Australia..
It wasn't until recently I finally understood why there is a estimated value type thingy written on my packages from stuff I buy from overseas its just habit sellers have for stuff as its required when they ship it else where in the world.

Well, you are very lucky in that case!

15M looks promising, 25A, 1.5Ah though, for $5.90
http://www.batterysupports.com/sams...-25a-1500mah-36v-liion-battery-cell-p-21.html

what would be the reason to go with this cells?

I dunno, just look at price primarly, didnt really realize its a wimpy 1.5Ah. Probably voltage sag etc.

The GA's from fastech are $28 each! WOW!

The Sanyo NCR186650GA cells are good at 72V 10.2Ah and 30A discharge. Cost is around $400.
I can buy a 25r pack, 72V 7.5Ah 60A for $476, add in a charger $80, so $550 shipped from a reputable seller. No need for DIY. Comes to about same price as cells from Nkon website.

No lol, that's price for 4 cells :lol:

 

[tumich]

Anyone saw discharge test about Sanyo UR18650RX? (20-22A cont dc) 2000mAh and fast charging 2,5A.

 

[riba2233]

I think that's topic for another thread... And btw there are cells with same C rate but much larger capacity and faster charging available now.