18650 cell test: Panasonic vs. LG

circuit

100 kW
Joined
Mar 19, 2009
Messages
1,323
Location
EU
did some cycling on these two cells... Purchased from same supplier.
DSC_1130.jpg
DSC_1131.jpg

cycle.png

dis.png

ch.png

discharge-slope1.png

dis11.png


More info on my blog (Lithuanian), translated to English (part 1 to part 4):
http://translate.google.com/translate?hl=en&sl=lt&tl=en&u=http%3A%2F%2Fe-motion.lt%2F2013%2F04%2F07%2Flicio-celiu-pasirinkimas%2F
http://translate.google.com/translate?hl=en&sl=lt&tl=en&u=http%3A%2F%2Fe-motion.lt%2F2013%2F04%2F10%2Flicio-celiu-pasirinkimas-2%2F
http://translate.google.com/translate?hl=en&sl=lt&tl=en&u=http%3A%2F%2Fe-motion.lt%2F2013%2F07%2F22%2Flicio-celiu-pasirinkimas-3%2F
http://translate.google.com/translate?sl=lt&tl=en&js=n&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fe-motion.lt%2F2013%2F07%2F30%2Flicio-celiu-pasirinkimas-4%2F
Translation is really bad, but it is possible to understand the idea.

There you will find energy density, power density, cycle life, price comparisons, for example:
ICR18650D1.png


Hope this will be useful to some. Anyway I will stay away from Panasonic cells for now, until I'm sure I did something wrong there.
 
Very interesting data. You're the first individual I know that does a life cycle test. Wow!

I have always been interested in the Panasonic NCR18650B cell, but I have also had doubt in mind about this chemistry. Looks like my doubts are valid, based on the result of your test here.

Question: Why did you choose 3A charge rate? What was the recommendation from Panasonic?
 
Here is an official datasheet from Panasonic website: http://e-motion.lt/wp-content/uploads/2013/07/NCR18650B1S.pdf

It says "charging current 1325 mA" which is 0.5C. Does not mention "fast charging".
LG datasheet says "charging current 0.5C" and "fast charging 1C". "Max discharge" at room temperature for LG is 2C, same as Panasonic. So I assumed the same fast charge applies for Panasonic.
I chose to test it at 3A just because I tested LG cell at 3A and wanted to keep conditions the same. And to keep the testing in smart timeframe... The 350 cycles at 1C took 6 weeks.
I really hope that Panasonic cell failed due to too high charging current. I do have one more cell, so will repeat the tests some day.

By the way, I was really impressed by LG cell. At this rough test it managed to give 500 cycles to 80% capacity... So in more practical 0.25C charging and average DOD of 70%, it should hold for 2000 cycles or more (just a guess).
 
circuit said:
It says "charging current 1325 mA" which is 0.5C.
That's what I expected. 1325/3400 = 0.39C

circuit said:
I really hope that Panasonic cell failed due to too high charging current. I do have one more cell, so will repeat the tests some day.
I hope so too. Panasonic has a reputation to keep. I would love to see the test done again at the 1C discharge and 0.39C charge. You only need 50 cycles to know if the high charge rate was the cause of the failure.

circuit said:
By the way, I was really impressed by LG cell. At this rough test it managed to give 500 cycles to 80% capacity... So in more practical 0.25C charging and average DOD of 70%, it should hold for 2000 cycles or more (just a guess).
I'm impressed too. But I wouldn't expect 2000 cycles. I'd happy to get 75% capacity after 1000 cycles.

BTW, the Panasonic NCR18650B is definitely a 1C battery. And Panasonic datasheet reflects it. Look at the black 2C discharge curve, the average voltage is only around 3.2V. Not acceptable for a 3.6V nominal battery. Even 1C is pushing the limit. 0.5C is what it's really designed for.
 
I guess the panasonic ones are only good with many in parallel. And at low discharge rate. From your data, I would charge slower to see what the effect is and also discharge slower.

I do like their energy density.
 
mvly said:
I guess the panasonic ones are only good with many in parallel. And at low discharge rate. From your data, I would charge slower to see what the effect is and also discharge slower.

I do like their energy density.

Yes, that is why Tesla has a crapton of them in every car. I am sure that they design the car to cruise at a 0.5-1C rate, so that the full capacity can be delivered without generating too much heat.

I have a feeling they discontinued the 40kWH model because they realized that the abuse put on that pack during heavy acceleration, hill climbing etc was just a liiiiitttlee too much :)
 
Panasonic NCR cells die early at 1C charge rates. This is nothing unexpected.

Charge them at 0,5C maximum (better 0,3C) and discharge them at 0,5C and they should live a lot(!) longer.
 
Cephalotus said:
Panasonic NCR cells die early at 1C charge rates. This is nothing unexpected.

Charge them at 0,5C maximum (better 0,3C) and discharge them at 0,5C and they should live a lot(!) longer.
This is strange, because all rates are the same as for LG. Ri of LG cell is even higher, and so C rating should be lower than Panasonic's. And LG performed quite better at 1C.
If tested at 0.5C, LG would do a lot better, too.
 
IR seems to not be a good determinate of charge rate. I have read some cell spec sheets where the charge rate was higher than the discharge rate!

That does seem to be an awfully early downward slope for that battery.
 
circuit said:
This is strange, because all rates are the same as for LG. Ri of LG cell is even higher, and so C rating should be lower than Panasonic's. And LG performed quite better at 1C.
If tested at 0.5C, LG would do a lot better, too.

I don't know anything about LG and I'm unable to comapre them, but I know that Panasonic NCR cells do not like 1C charge rates. I asume that your cells also got quite hot? Panasonic gives more than 80% capacity after 2.000 clcles for new NCR types, but not at high carge rates. High discharge >1C is also problematic, but high cahrge seems to be really bad for them.

If you need battreis for high (average) chareg / dicharge rates use other cells. Charge them moderatly and the "should" live much, much longer. Don't use regen with high cahrge rates either. Discharge in e-bike is not contionous, so this is difficult to simulate anyway...
 
circuit said:
OK, will do 50 cycles at 1C discharge and 0.5C charge.
That's great. I look forward to seeing the results.

Just a thought here, not trying change your mind or convince you of anything: Why not do the test at the actual anticipated charge/discharge rate? After all the actual charge/discharge rate is the only one that's meaningful to you. Why should you care about the other rates? All my 18650 packs (and I have a lot of them) were built based on 0.5C average discharge or less. So all the capacity tests done prior to pack assembly were at 0.5C. I never pay any attention to charging rate because I never go beyond 0.3C (4 hour full charge from empty).
 
Around 1C discharge is close to what I am planning for. As for slower charge, there are no problems to make those tests happen, except one - time. It takes a lot of time.

Anyway the purpose of my tests was not to know exactly how the cells will perform on my setup, but rather which cell is better, performance/price wise. So far LG wins the competition hands-down. LG would win this competition even if Panasonic could perform slightly better than LG... Because of lower price.
 
circuit said:
OK, will do 50 cycles at 1C discharge and 0.5C charge.

I'm looking forward to the results from this.

If you do have the equipment, setup another that charge to 4.1V and discharge to 3.1V. Rumor has it this is the charge-discharge interval Tesla uses for their packs.

These NCR cells, if they are NMC (are they?) they will take a hit from fast charge. That is one of the favourable spesifications of the "old" spinel chemistry, fast charge is no problem. What chemistry is the LG cell? Couldn't dig up any info...
 
Interesting results! Thanks for the testing.

Dont' suppose you have any temperature rise information for these cells during the tests? I'm curious to compare the two (should give some idea of internal resistance).

I can imagine that charging and discharging at 1C the entire time with only a 5minute rest had the temperatures elevated for most of the testing - skews the results from a control standpoint - but gives good indication of how the cells might perform in stressful conditions.
 
5 minutes were enough to cool them down back to near room temperature. Especially in charging, where CV took almost half the time, so it was cool at the end of charge.
As per temperature, I'd say LG was a bit hotter and peak temperature was during charging, close to switch from CC to CV. When CV is reached, current starts to fall and temperature drops. As mentioned before, maximum temperature measured with hand was around 45°C.
I will attach the probe next time, if I find it...
 
First, great work on your data collection!

NCR18650B is quite poorly suited towards most EV applications.

NCR18650PD on the other hand, is perhaps the best EV suited 18650 I've ever tested, and I think you would be very happily surprised doing the same cycling tests.
 
liveforphysics said:
First, great work on your data collection!

NCR18650B is quite poorly suited towards most EV applications.

NCR18650PD on the other hand, is perhaps the best EV suited 18650 I've ever tested, and I think you would be very happily surprised doing the same cycling tests.

I wish I could see this data on the PD instead. 2.9AH and a 20AH capacity now that would make for a killer commuter pack.
 
Back
Top