Samsung 30Q INR18650-30Q 3000mah 15a cell

[jk1]

It will be interesting to see what the NCR18650GA's will do at 7a they should be ok at that since they are 10 a rated cells.

 

[Pajda]

It will be interesting to see what the NCR18650GA's will do at 7a they should be ok at that since they are 10 a rated cells.

The holy trinity of Samsung INR18650-35E, Sanyo NCR18650GA and LG INR18650 MJ1 as a highest capacity cells available on the market are good performers up to 3C discharge rate but they hate faster charging than standard 0,5C (this value of CHG current is recommended in datasheets of all three cells). Here you can see the results of cycle life tested at 1C-1C:

I can tell you that Sanyo NCR18650GA have the lowest DCIR, but LG have superior cycle life under all conditions. I have also very bad experience with "High capacity" Samsung cells ("E" and "A" series) like 35E and 32A, they all have very poor cycle life. So I am a little surprised that "High power" ("Q" and "R" series) Samsungs cells are such a good performers.

 

[okashira]

It will be interesting to see what the NCR18650GA's will do at 7a they should be ok at that since they are 10 a rated cells.

The holy trinity of Samsung INR18650-35E, Sanyo NCR18650GA and LG INR18650 MJ1 as a highest capacity cells available on the market are good performers up to 3C discharge rate but they hate faster charging than standard 0,5C (this value of CHG current is recommended in datasheets of all three cells). Here you can see the results of cycle life tested at 1C-1C:

I can tell you that Sanyo NCR18650GA have the lowest DCIR, but LG have superior cycle life under all conditions. I have also very bad experience with "High capacity" Samsung cells ("E" and "A" series) like 35E and 32A, they all have very poor cycle life. So I am a little surprised that "High power" ("Q" and "R" series) Samsungs cells are such a good performers.

You're charging too fast. Run again charging the GA at 1.5A not 3.5A (!!!)

 

[okashira]

It will be interesting to see what the NCR18650GA's will do at 7a they should be ok at that since they are 10 a rated cells.

The holy trinity of Samsung INR18650-35E, Sanyo NCR18650GA and LG INR18650 MJ1 as a highest capacity cells available on the market are good performers up to 3C discharge rate but they hate faster charging than standard 0,5C (this value of CHG current is recommended in datasheets of all three cells). Here you can see the results of cycle life tested at 1C-1C:

I can tell you that Sanyo NCR18650GA have the lowest DCIR, but LG have superior cycle life under all conditions. I have also very bad experience with "High capacity" Samsung cells ("E" and "A" series) like 35E and 32A, they all have very poor cycle life. So I am a little surprised that "High power" ("Q" and "R" series) Samsungs cells are such a good performers.

You're charging too fast. Run again charging the GA at 1.5A not 3.5A (!!!)

ALl of those cycle life's are terrible. It's nice that the MJ1 outperformed in this test, but it's still unacceptable performance so these results can be mostly discounted. You need to reduce charge current substantially

 

[Pajda]

ALl of those cycle life's are terrible. It's nice that the MJ1 outperformed in this test, but it's still unacceptable performance so these results can be mostly discounted. You need to reduce charge current substantially

Yes, I was doing those tests with this expectation. In short the meaning of this test was to determine what impact will have a fast charging in BEVs for those industrial grade cells in comparsion to the Tesla cells. And as you have already mentioned the impact is "devastating".

You're charging too fast. Run again charging the GA at 1.5A not 3.5A (!!!)

As you wish :wink: 0,5C equals to 1700 mA (datasheet value of "Standard" charging current for both cells):

 

[Science Faction]

Hi All, Great work and data in this thread.
Where are we upto in terms of suppliers?
What is the most reliable/cheapest supplier? especially for small units as I am doing a quadcopter project.
If anyone has bought them in bulk and wants to sell me 4 for testing I would be most appreciative.
Actualy I wouldn't mind 4 x 25R's as well.
Cheers
Paul

 

[ridethelightning]

best supplier for small quantities that I know is fasttech, 2x30Q ~$12(?) free shipping.
others have cheaper prices but you need to buy larger quantities to get a discount and to make the expensive shipping worthwhile

 

[TheBeastie]

ALl of those cycle life's are terrible. It's nice that the MJ1 outperformed in this test, but it's still unacceptable performance so these results can be mostly discounted. You need to reduce charge current substantially

Yes, I was doing those tests with this expectation. In short the meaning of this test was to determine what impact will have a fast charging in BEVs for those industrial grade cells in comparsion to the Tesla cells. And as you have already mentioned the impact is "devastating".

You're charging too fast. Run again charging the GA at 1.5A not 3.5A (!!!)

As you wish :wink: 0,5C equals to 1700 mA (datasheet value of "Standard" charging current for both cells):

Pajda, Quite fascinating to see how much poor performance you get with your cells when you charge them at a high rate..
I am wondering if other people get the same results if they charge at such a high rate...

I can see how its a worth while test because so many posts on on ES in general people buy/use/talk about massive high amp power supplies/charger setup so I can't help but assume they do in fact charge at these high rates.. I am guessing that its due to the logic that they feel its a better way to go if they bulk charge etc as they can be fully around with little waiting while its charging instead of very slowly charging overnight while they sleep or when not around at all..

Also the increase in resting time between chargers showing a larger increase in capacity is also very interesting, I remember in the final 6 months of using my last lipo pack that when I had left it on its ideal storage voltage for a good few months and came back to use it as a once off it had great performance for that single cycle and I figured leaving it rest a long time was some how internally 'recharging its batteries' so to speak but I never bothered mentioning it to anyone in the past.

Okashira, when I look at your charts/posts I haven't been able to see what charge rate you charge at?
http://i.imgur.com/LVpBOH3.jpg

 

[okashira]

ALl of those cycle life's are terrible. It's nice that the MJ1 outperformed in this test, but it's still unacceptable performance so these results can be mostly discounted. You need to reduce charge current substantially

Yes, I was doing those tests with this expectation. In short the meaning of this test was to determine what impact will have a fast charging in BEVs for those industrial grade cells in comparsion to the Tesla cells. And as you have already mentioned the impact is "devastating".

You're charging too fast. Run again charging the GA at 1.5A not 3.5A (!!!)

As you wish :wink: 0,5C equals to 1700 mA (datasheet value of "Standard" charging current for both cells):

Pajda, Quite fascinating to see how much poor performance you get with your cells when you charge them at a high rate..
I am wondering if other people get the same results if they charge at such a high rate...

I can see how its a worth while test because so many posts on on ES in general people buy/use/talk about massive high amp power supplies/charger setup so I can't help but assume they do in fact charge at these high rates.. I am guessing that its due to the logic that they feel its a better way to go if they bulk charge etc as they can be fully around with little waiting while its charging instead of very slowly charging overnight while they sleep or when not around at all..

Also the increase in resting time between chargers showing a larger increase in capacity is also very interesting, I remember in the final 6 months of using my last lipo pack that when I had left it on its ideal storage voltage for a good few months and came back to use it as a once off it had great performance for that single cycle and I figured leaving it rest a long time was some how internally 'recharging its batteries' so to speak but I never bothered mentioning it to anyone in the past.

Okashira, when I look at your charts/posts I haven't been able to see what charge rate you charge at?
http://i.imgur.com/LVpBOH3.jpg

I've been using ~2A until 3.8V, 1.5A to 4.15V for the 30Q

1.6A until 3.77V then 0.9A until 4.15V for the Model S cell

 

[cwah]

Yes i noticed that that's why i charge them as low as possible rate

 

[Pajda]

I've been using ~2A until 3.8V, 1.5A to 4.15V for the 30Q

1.6A until 3.77V then 0.9A until 4.15V for the Model S cell

and cut-off currents pls?

Did you find any significant benefit by using "Tesla AVC" instead of standard CC-CV method? I hope I can start "2 stage AVC" tests for LG MJ1 cell samples this week ("Only" 2 stage AVC because the chinese tester ZKE EBC-A10 is limited to program 6 steps only). I think I can start with "1C-1C" equivalent at 100% DoD, so the time to full charge with 100mA cut-off is 1h 35min. I have already tried this settings:

1) CC-CV: 5,0A to 3,9V with 2,5A cut-off
2) CC-CV: 2,5A to 4,2V with 100mA cut-off

With this AVC setting the cell reach 4,2V within 45min (almost the same time as CC phase with 1C at 3,4A) but the finishing CV phase starting on 2,5A is much longer so the total charging time is 1h 50min.

 

[okashira]

I've been using ~2A until 3.8V, 1.5A to 4.15V for the 30Q

1.6A until 3.77V then 0.9A until 4.15V for the Model S cell

and cut-off currents pls?

Did you find any significant benefit by using "Tesla AVC" instead of standard CC-CV method? I hope I can start "2 stage AVC" tests for LG MJ1 cell samples this week ("Only" 2 stage AVC because the chinese tester ZKE EBC-A10 is limited to program 6 steps only). I think I can start with "1C-1C" equivalent at 100% DoD, so the time to full charge with 100mA cut-off is 1h 35min. I have already tried this settings:

1) CC-CV: 5,0A to 3,9V with 2,5A cut-off
2) CC-CV: 2,5A to 4,2V with 100mA cut-off

With this AVC setting the cell reach 4,2V within 45min (almost the same time as CC phase with 1C at 3,4A) but the finishing CV phase starting on 2,5A is much longer so the total charging time is 1h 50min.

0.15A cut off for Model S cell
0.2A cut off for 30Q

 

[okashira]

I've been using ~2A until 3.8V, 1.5A to 4.15V for the 30Q

1.6A until 3.77V then 0.9A until 4.15V for the Model S cell

and cut-off currents pls?

Did you find any significant benefit by using "Tesla AVC" instead of standard CC-CV method? I hope I can start "2 stage AVC" tests for LG MJ1 cell samples this week ("Only" 2 stage AVC because the chinese tester ZKE EBC-A10 is limited to program 6 steps only). I think I can start with "1C-1C" equivalent at 100% DoD, so the time to full charge with 100mA cut-off is 1h 35min. I have already tried this settings:

1) CC-CV: 5,0A to 3,9V with 2,5A cut-off
2) CC-CV: 2,5A to 4,2V with 100mA cut-off

With this AVC setting the cell reach 4,2V within 45min (almost the same time as CC phase with 1C at 3,4A) but the finishing CV phase starting on 2,5A is much longer so the total charging time is 1h 50min.

I have not done any comparisons to see if there is a benefit to the AVC method, however as you can see it's exactly what I am using.

I did cycle test using a supercharge algorithm, my tester does 10 steps so I am able to emulate it closely. I got really bad cycle life doing that.

 

[okashira]

See this post where I tried a fast charge
https://endless-sphere.com/forums/viewtopic.php?f=14&t=68556&start=50#p1058819
I was using a pretty conservative supercharge, and it was the sole reason for the poor cycle life.
Reducing charge current fixed the cycle life

I am pretty sure you will rape the cell with the 5A / 2.5A charge setup, but give it a try.

My guess is Tesla uses pulse charging (ie, 8A for 100ms, 0a for 100ms) to cause the cells to warm up as much as possible at the beginning

 

[Pajda]

See this post where I tried a fast charge
https://endless-sphere.com/forums/viewtopic.php?f=14&t=68556&start=50#p1058819

For the next time I should read the whole thread before I ask for what was already mentioned

I was using a pretty conservative supercharge, and it was the sole reason for the poor cycle life.
Reducing charge current fixed the cycle life

hmmm...

I am pretty sure you will rape the cell with the 5A / 2.5A charge setup, but give it a try.

I also think there will be a short end. But the cell perform very well even at 5A charging current. The phase of 5A CC takes only 10min and after that the cell is only a barely hot (I usually have IR camera and temperature data logger but not at this time) but I can easily hold my hand on the cell.

My guess is Tesla uses pulse charging (ie, 8A for 100ms, 0a for 100ms) to cause the cells to warm up as much as possible at the beginning

That is a new information for me... But I do not believe that pulse charging with average value of 4A will generate much more heat inside the cell than 4A continuously.

 

[parabellum]

But I do not believe that pulse charging with average value of 4A will generate much more heat inside the cell than 4A continuously.

Not much more, just double.

 

[okashira]

See this post where I tried a fast charge
https://endless-sphere.com/forums/viewtopic.php?f=14&t=68556&start=50#p1058819

For the next time I should read the whole thread before I ask for what was already mentioned

I was using a pretty conservative supercharge, and it was the sole reason for the poor cycle life.
Reducing charge current fixed the cycle life

hmmm...

I am pretty sure you will rape the cell with the 5A / 2.5A charge setup, but give it a try.

I also think there will be a short end. But the cell perform very well even at 5A charging current. The phase of 5A CC takes only 10min and after that the cell is only a barely hot (I usually have IR camera and temperature data logger but not at this time) but I can easily hold my hand on the cell.

My guess is Tesla uses pulse charging (ie, 8A for 100ms, 0a for 100ms) to cause the cells to warm up as much as possible at the beginning

That is a new information for me... But I do not believe that pulse charging with average value of 4A will generate much more heat inside the cell than 4A continuously.

I did a supercharge and recorded temperature here:
https://www.youtube.com/watch?v=vFrTUlYIQww

Heat is not the problem with high charge current. Actually, the hotter the better because the lithium ions can diffuse faster.
http://lithiumbatteryresearch.com/Plating.php

 

[okashira]

Updated 30Q, 7A model S.

IN RED - Also added another Model S cell. This one I charge to 4.05V with a 0.2A cut. Discharge to 2.9V with 5A. The first few cycles I set the charge cut wrong so it was high.

Remind me to fix the labels and descriptions...

 

[okashira]

Updated\
Seems the Model S cell has some sort of self healing characteristic. Very interesting.
Any significant pause in cycle testing results in improved capacity.
All of the same pauses are done on the 30Q, and it only results in the same or reduced capacity.

 

[Wheazel]

That is very intresting indeed. Wonder what process lies behind this data...

 

[bigbore]

Is there any comparison between Model S and 30Q at 10A discharge (not cycle testing)?

 

[riba2233]

I have both cells at home, but can only test up to 5 A...

 

[bigbore]

I have both cells at home, but can only test up to 5 A...

Do you want to look at my poor discharge setup? I could make a little video tomorrow
These days I've been checking my old zippy lipos cells to make a 3s?p pack to use with your spot welder and I can discharge single lithium cells at 10A +/- 0.100A

 

[riba2233]

That's great, now you only need cells

 

[hillzofvalp]

Anyone willing to cycle the HG2 at 2A charge, 7A discharge?