Tabless design cylindrical cells tests

riba2233 said:
I just miss developments from big four, Panasonic, sony, lg, Samsung... They are seriously behind and they used to all be in sync with releasing new tech, capacities, power levels etc. Ok sony/murata released 4ah tabless but that's about it (Samsung's 4.5ah is already behind). They are better with HE cells but still, not to mention large format like 4860...
Click to expand...
To be fair, that's mostly because they're concentrated on pouch and prismatic large format cells. They also have very nice tech stuff coming relatively soon (specifically K-On with Hyundai), but it seems like their cyclindrical cell development has indeed slowed down.

It's just difficult to improve energy density today without massively improving packaging efficiency, using better cathodes/anodes, or changing the base tech entirely.

So much research seems to lead to nowhere and yet, all that research that was deemed unprofitable (IE, a cancerous view when it comes to research) has suddenly become profitable and suddenly, most established companies are finding themselves with their pants down.
 
The Christmas update is finally here:

News:
- added cells:
Samsung 45T (CC4503F101) 4500 mAh
18650-21700.com AM04 (Ampace JP40) 4000 mAh

Update:
- added more cycles in all graphs
- added new Table results

My comments:
- Merry Christmas to you all and thanks for your support 🎅

- The Samsung 45T (along with the 58E) arrived just before Christmas so so far only 5C and 10C discharge characteristics with temperatures and a few first cycles are available. With the 45T it is now clear that it is not a tabless design. It seems that Samsung has finally matched the Molicel P45B at a better price. But from my point of view, it's too little, too late because next to the tabless BAK 45D both of these cells don't make much sense.

- The 18650-21700.com (that's the name of Austin Marhold's company) AM04 (that's a model designation) is, according to available information, based on the Ampace JP40 design. I also got samples of original JP40 from Austin and I can confirm that they are almost identical as far as discharge curves are concerned. The reason why I am not showing the original JP40 results yet is that unfortunately most of the samples arrived discharged below 2V and the two samples that were above 2.5V show an increased rate of self-discharge. Austin responded that he was aware of this and was working with his supplier to make sure it didn't happen again. However, the milk has already been spilled. And it doesn't add to the confidence that it will be easy to get Grade A quality of JP40. So here's my disclaimer, even though my tests of any cell model come out very well at the single cell level, it may still be problematic to produce a battery pack from them (that's the main difference between Grade A and B cells, which unfortunately cannot be seen in my tests). In any case, JP40/AM04 is currently one of the best demonstrators of HP cell technology in the 21700 format. Interestingly, the tabless design is not a guarantee for good cycle-life with fast charging (3C). AM04 heats up only minimally during 20min 3C charging (to 34°C at 24°C ambient temperature without cooling) but after the first 50 cycles, it's finished. It should be noted that the datasheet states max 2C charging for JP40/AM04 similar to EVE 40PL (On the other hand, Molicel and BAK allows 3C charging and achieve a decent cycle-life when 3C charging is used)
 
Pajda said:
The Christmas update is finally here:

News:
- added cells:
Samsung 45T (CC4503F101) 4500 mAh
18650-21700.com AM04 (Ampace JP40) 4000 mAh

Update:
- added more cycles in all graphs
- added new Table results

My comments:
- Merry Christmas to you all and thanks for your support 🎅

- The Samsung 45T (along with the 58E) arrived just before Christmas so so far only 5C and 10C discharge characteristics with temperatures and a few first cycles are available. With the 45T it is now clear that it is not a tabless design. It seems that Samsung has finally matched the Molicel P45B at a better price. But from my point of view, it's too little, too late because next to the tabless BAK 45D both of these cells don't make much sense.

- The 18650-21700.com (that's the name of Austin Marhold's company) AM04 (that's a model designation) is, according to available information, based on the Ampace JP40 design. I also got samples of original JP40 from Austin and I can confirm that they are almost identical as far as discharge curves are concerned. The reason why I am not showing the original JP40 results yet is that unfortunately most of the samples arrived discharged below 2V and the two samples that were above 2.5V show an increased rate of self-discharge. Austin responded that he was aware of this and was working with his supplier to make sure it didn't happen again. However, the milk has already been spilled. And it doesn't add to the confidence that it will be easy to get Grade A quality of JP40. So here's my disclaimer, even though my tests of any cell model come out very well at the single cell level, it may still be problematic to produce a battery pack from them (that's the main difference between Grade A and B cells, which unfortunately cannot be seen in my tests). In any case, JP40/AM04 is currently one of the best demonstrators of HP cell technology in the 21700 format. Interestingly, the tabless design is not a guarantee for good cycle-life with fast charging (3C). AM04 heats up only minimally during 20min 3C charging (to 34°C at 24°C ambient temperature without cooling) but after the first 50 cycles, it's finished. It should be noted that the datasheet states max 2C charging for JP40/AM04 similar to EVE 40PL (On the other hand, Molicel and BAK allows 3C charging and achieve a decent cycle-life when 3C charging is used)
Click to expand...
That's rather weird. If it doesn't heat up much at all and it doesn't have a large voltage change, that indicates it isn't anode, electrolyte, or interconnect limited...

That actually sounds scary, since this possibly indicates lithium plating from an imbalanced anode:cathode loading ratio that only happens during fast charging (something that can and does happen in literature)...
 
BlueSwordM said:
That actually sounds scary, since this possibly indicates lithium plating from an imbalanced anode:cathode loading ratio that only happens during fast charging (something that can and does happen in literature)...
Click to expand...
Hot spotting and electrolyte degradation (creating gas and byproducts) is potentially an issue as I had two AM04’s pop their CID’s when run at a high current level (within rating) up to the datasheet’s max external temp spec.
 
CamLight said:
Hot spotting and electrolyte degradation (creating gas and byproducts) is potentially an issue as I had two AM04’s pop their CID’s when run at a high current level (within rating) up to the datasheet’s max external temp spec.
Click to expand...
1- Perhaps, but this could be why Ampace (CATL/ATL) only advertised the JP30, never the JP40.
I'm starting to believe that all 1st generation battery products are just cursed, as the EVE40PL, JP40 and AM04 have all had issues.

The failure modes that you experienced seem to be both related to lithium plating, but for different reasons.

Lithium plating failures during fast charging tends to happen mainly when the graphite anode, or electrolyte, can't keep up; graphite potential vs Li+ drops below 0V, and it becomes favorable for lithium metal to be deposited and eventually, the electrolyte gets consumed, too much lithium gets consumed and the cell just dies.

Considering the cell didn't get hot, this is actually somewhat likely and compounded with my previous statement regarding anode:cathode ratio: after all, if the cells had been heated to 35C beforehand, it's possible Padja's cells wouldn't have died after 50 cycles at 3C CCCV.

Why wouldn't the P50B or BAK 45D be affected? Well, they likely have much higher Si loading in their anodes, which would allow them much more room before lithium plating can begin, even if they aren't as good in other places :)

2- As for your case Mooch, it's probably due to lithium salt or electrolyte depletion. Now that the interconnect isn't much of a bottleneck (tabless), cathodes and electrolytes are pushed to the limit during discharging. Hotspots and cathode degradation aren't much of a worry on my side, but electrolyte depletion at high rates is, which leads to gas generation, etc.
|
| 
---> I forgot about another factor: the current collector tab to the top contact. The electrodes may be all-tab (tabless), but not what gets connected to the outside. There's also the separator :)
 
Last edited:
BlueSwordM said:
Now that the interconnect isn't much of a bottleneck (tabless), cathodes and electrolytes are pushed to the limit during discharging. Hotspots and cathode degradation aren't much of a worry on my side, but electrolyte depletion at high rates is, which leads to gas generation, etc.
Click to expand...
All of the “tabless” 21700’s have a single cathode tab though (from the current collector plate to the top contact) and hotspotting should still be a potential issue IMO. The anodes are tabless except for the 45D.

Though I enjoy the discussions here one day it would nice to not be limited to our guessing and have some of the engineers/researchers available for a few hours of AMA. 😂
 
BlueSwordM said:
1- Perhaps, but this could be why Ampace (CATL/ATL) only advertised the JP30, never the JP40.
Click to expand...
There was a huge amount of press on the JP40 and a big trade show release of that cell. I haven’t seen nearly as much for the JP30 (and only the JP40 is on their web site) but the first cell in a new line always seems to get the most attention.
 
Small update

News:
- pls check new data charts with cell body Temperatures 🥶 if you like it, I can add them in the first post.

My comments:
- Data are available only for cells measured on NEWARE BTS-4000 system since spring this year.
- Data are measured using a miniature T-type thermocouple, fixed by kapton tape in the middle of the cell body, with 1 sec sampe-rate
- My cycle-life test is running in a basement room with a relatively stable temperature and no significant air flow. The room temeperature (RT) stability is usually within +/- 2 °C, the problem is more the summer-winter transition when the RT changes from about 20°C (Table data are measured at 25°C) in winter to about 23°C in summer. This error can be partially compensated by the second "dT" data chart, where the cell body temperature at the beginning and end of the test is taken and the dT=Tend-Tbegin is calculated from it.

5C_cycle_life_test@Temperature_rise.PNG

5C_cycle_life_test@deltaT_rise.PNG
 
Last edited:
Yeah, but only the JP30 is advertised on their YT channel :)
CamLight said:
There was a huge amount of press on the JP40 and a big trade show release of that cell. I haven’t seen nearly as much for the JP30 (and only the JP40 is on their web site) but the first cell in a new line always seems to get the most attention.
Click to expand...
 
CamLight said:
All of the “tabless” 21700’s have a single cathode tab though (from the current collector plate to the top contact) and hotspotting should still be a potential issue IMO. The anodes are tabless except for the 45D.

Though I enjoy the discussions here one day it would nice to not be limited to our guessing and have some of the engineers/researchers available for a few hours of AMA.
Click to expand...
You're right, and I should probably find a way to ask if we could have an interview on here :)
 
BlueSwordM said:
The different letter suffix for the P70 makes me think they're probably loading the cell anode to the gills with Si and utilizing a different cathode material entirely...
Click to expand...
They’re already using Si in the P50B so I’m thinking the X suffix is for their first “tabless“ cell.
I agree, cathode chem tweaks are inevitable though, to one degree or another. Probably done for every cell since their release schedule is pretty spread out.
 
Bengy22 said:
Every year molicel teases an ever better cell that makes me want to wait for it. I guess I will build my new pack when the p100x comes out
Click to expand...

I'm in the same place.

Batteries got 5% better per year for the last decade.

Now.. semi solid state batteries are hitting the market in 2025. Solid state is hitting the market in 2026. One company has produced 400whrs/kg solid state and is working on 500whrs/kg now.

Exciting times, but it's worse than buying a gaming rig in the mid 90's.
 
Hello, I'm looking for good cells to build my first battery pack.
I need 20S ?P for 84v 25-35Ah ~100A continuous ~200A max discharge.
I need batteries with not some huge voltage sag when discharge is at maximum.
Any suggestions please ?

Edit : according to the curves (thank you author for your work) and the quality-price ratio I chose the BAK 45D. A 20S6P pack will give me 360A max discharge, that's way more than I need (controller 200A max) but at least one charge will last a loooong time! I mean, I could use full throttle without "low battery" message on the screen after a few secs.
 
Last edited:
Ziip0 said:
Hello, I'm looking for good cells to build my first battery pack.
I need 20S ?P for 84v 25-35Ah ~100A continuous ~200A max discharge.
I need batteries with not some huge voltage sag when discharge is at maximum.
Any suggestions please ?

Edit : according to the curves (thank you author for your work) and the quality-price ratio I chose the BAK 45D. A 20S6P pack will give me 360A max discharge, that's way more than I need (controller 200A max) but at least one charge will last a loooong time! I mean, I could use full throttle without "low battery" message on the screen after a few secs.
Click to expand...
Molicel P50B for max capacity high power, Bak 45D for maximum power with lower capacity.
 
BlueSwordM said:
Molicel P50B for max capacity high power, Bak 45D for maximum power with lower capacity.
Click to expand...
Yes Molicel P50B is the best, but the price difference is huge for 3Ah gain.
20S6P BAK 45D 27Ah 360A max : 450€
20S6P Molicel P50B 30Ah 360A max : 834€ 😅
 
neptronix said:
Batteries got 5% better per year for the last decade.
Click to expand...
Let's start the New year with some controversy. :D According to my long term measurements this was not exactly the case, at least for the cells commonly available on the retail market (i.e. no vaporware). From my observation to date, the small cylinder format (18650) was the carrier of the best mix of highest energy density combined with availability for the end customers/DIYers (only a few years ago it was finally caught up and overtaken by the larger format 21700). The problem with this "5% law" is in the emergence of what I call "the big battery development gap" roughly between 2012-2022. This was a period when the increase of the best available energy density slowed down significantly. The increase over the whole decade was only about 5 %. However, it wasn't a complete disaster and at least we got significantly better overall cell-life, DCIR and of course a huge price drop. Now it looks like they are pushing the energy density again with desired speed, e.g. the just announced Molicel P70X (of course, if it will be a cell in 21700 format ;))

energy_density_development@Map.PNG
 
From the article:
"PRiMX50U-Power can charge to 80% capacity in just 15 minutes"

Sounds like a high discharge, high capacity cell from here :cool:
 
Wow.. it was worse than 5% per year..
Let's put a term on it: battery malaise era
 
You must log in or register to reply here.

Similar threads

P
    • Like
Cycle life tests of mixed-technology cells
Replies
8
Views
1,376
Pajda
P
P
    • Like
Cycle life tests of High Energy density cylindrical cells
4 5 6 7 8
Replies
178
Views
45,174
Pajda
P
P
    • Like
Cycle life tests of High Power density cylindrical cells
2 3 4 5
Replies
119
Views
24,330
yantra yatra
Y
P
    • Like
Calendar ageing tests of mixed cylindrical cells
Replies
13
Views
3,199
Pajda
P
GreenTecAuto
Valence U27-12XP Lithium Ion Battery $370 per Module
Replies
1
Views
1,211
GreenTecAuto
GreenTecAuto
Back
Top