Tabless design cylindrical cells tests

DogDipstick said:
On another note, I was told to take IR on charge only, and when you hook multimeters with different AC signals comming out of the leads.... it fiddles witht he HRM 10 readings... my scope told me. maybe that is why it shows so low.. I would like to see the cell tested with that machine alone. Both cells show different from manuf published ACIR numbers? the moli and the jp40...?
Click to expand...
The standard is ACIR at 30% SoC and at rest in my experience. Not sure if JP40 specs have been published anywhere, but most cells I've tested (P45B, EVE 40PL relevantly) are pretty much dead-on (slightly better than nominal spec).
 
Souvlucky said:
Kumyang has some tabless 46mm cells (46/80*, 46/95, 46/110* and 46/120*) in the making :)
(* are under development as "Energy" variants; here's a shameless crosspost with more info.)

View attachment 358184
Click to expand...
Those copper ends look beautiful, but do you have a photo of the aluminium end?
 
riba2233 said:
Wow these look like pure garbage, how did they manage that lol?

And while 4860 looks to have a good cycle life characteristics, 4695 looks much better based on numbers alone
Click to expand...
The very annoying possibility that they used a high concentration of methyl acetate in their electrolyte to further boost power density without considerations for cycle life with graphite anodes at high voltages (>4.10V).

There are other possibilities like improper cathode:anode capacity ratios, poor electrode adhesion, etc, but the electrolyte mix being quite... unoptimal is the most likely reason for such poor performance; you can tell by the drastic increase in IR, indicating either lithium plating is happening or immense anode degradation is at play.
 
small update Tabless design cylindrical cells tests

added low power cycle life graph at 90% (4.15-3.00V) for HP cells.
- 40PL doesn't seem to have a life cycle problem at low power.
- 40P so far shows standard cycle lifetime at 5C discharge

I'm thinking there's no reason why a 40PL should have significantly worse cycle life than a 40P. I assume the active material and electrolyte is identical. Plus @CamLight has a good result in his tests. I was wondering if I could have damaged the internal tabless interconnects on the anode using Micro-TIG welding (1C test is running in BF-2A holder only) Although very unlikely, I'll still try to run the 5C test also in the BF-2A holder. Temperature-wise, it should still come out fine to about 60°C (with Cu strips the 5C discharge end temperature is around 50°C for 40PL and 60°C for 40P)

Also BAK 45D arrived, so on monday I will start with testing. BTCAP 40P (together with Lishen 58SK) are ordered but delivery time will be long like 40 days.

And finally the JP40 with two other interesting samples (not yet released) from 18650-21700 store are also on the way to me.
 
Pajda said:
small update Tabless design cylindrical cells tests

added low power cycle life graph at 90% (4.15-3.00V) for HP cells.
- 40PL doesn't seem to have a life cycle problem at low power.
- 40P so far shows standard cycle lifetime at 5C discharge

I'm thinking there's no reason why a 40PL should have significantly worse cycle life than a 40P. I assume the active material and electrolyte is identical. Plus @CamLight has a good result in his tests. I was wondering if I could have damaged the internal tabless interconnects on the anode using Micro-TIG welding (1C test is running in BF-2A holder only) Although very unlikely, I'll still try to run the 5C test also in the BF-2A holder. Temperature-wise, it should still come out fine to about 60°C (with Cu strips the 5C discharge end temperature is around 50°C for 40PL and 60°C for 40P)

Also BAK 45D arrived, so on monday I will start with testing. BTCAP 40P (together with Lishen 58SK) are ordered but delivery time will be long like 40 days.

And finally the JP40 with two other interesting samples (not yet released) from 18650-21700 store are also on the way to me.
Click to expand...
Huh, I didn't consider spot welding quality to be a factor in testing.
Maybe I just thought too far ahead with my electrolyte hypothesis since I've looked at far too many battery papers these days :p
 
BlueSwordM said:
Huh, I didn't consider spot welding quality to be a factor in testing.
Maybe I just thought too far ahead with my electrolyte hypothesis since I've looked at far too many battery papers these days :p
Click to expand...
Now that I think about it, I'm like a drowning man grasping at straws. In fact, I have not yet experienced a single problem with micro-tig welded busbars. So much more straightforward explanation can be that the significantly lower internal impedance caused by the tabless design allow to discharge the cell to a lower DoD level than with a conventional tab design.

From the discharge energy perspective, the value of 2.5 V cut-off voltage at 5C dicharge-rate for the 40P with tab design gives the same amount of energy as the value of 2.9 V for the tabless 40PL. I will try this test setup with 2.9 V cut-off voltage for 40PL first.
 
Last edited:
Are there no 5000+mAh tabless cells ? It could be a good compromise to get a low ir so it can take high discharge peaks, but still get a a little extra capacity. For my use in an enduro bike I have short peaks of 40-50A/cell, but the average consumption is more like 4-5A/cell.
 
Pajda said:
Now that I think about it, I'm like a drowning man grasping at straws. In fact, I have not yet experienced a single problem with micro-tig welded busbars. So much more straightforward explanation can be that the significantly lower internal impedance caused by the tabless design allow to discharge the cell to a lower DoD level than with a conventional tab design.

From the discharge energy perspective, the value of 2.5 V cut-off voltage at 5C dicharge-rate for the 40P with tab design gives the same amount of energy as the value of 2.9 V for the tabless 40PL. I will try this test setup with 2.9 V cut-off voltage for 40PL first.
Click to expand...
I noticed this for the first time on 18650-21700's website, they recommend not welding the JP40s directly on the spiral... however, EVE apparently does not consider this an issue.

I'm not set up for cycle testing ATM but (eventually/hopefully) can run cycling on 40PL/JP40 in the near future.

j bjork said:
Are there no 5000+mAh tabless cells ? It could be a good compromise to get a low ir so it can take high discharge peaks, but still get a a little extra capacity. For my use in an enduro bike I have short peaks of 40-50A/cell, but the average consumption is more like 4-5A/cell.
Click to expand...
There aren't any in mass production yet. Molicel's P50B isn't tabless but it will probably handle your use case, and will be probably the first one out in MP. Lishen, Tenpower, BAK, and EVE all have tabless 5Ah in the works but my guess is B-samples by EOY if we're lucky -- I've only seen early data so far.
 
So first results of another tabless competitor BAK 45D were added at the end of the first post :) Tabless design cylindrical cells tests Table results with temperatures will be updated later this week.

Also, I have already opened the BAK cell and confirmed that it is indeed tabless design. I hope to add here photos of its internal structure soon...
 
thermal_runaway said:
More 18650 tabless options should be on the market early next year, high demand from power tools market :)
Click to expand...

They should have been available even sooner, makita had some packs with tabless 2.5ah 18650s already shown in catalogues but then removed them some time after, best guess is that they were delayed for some reason.
 
riba2233 said:
They should have been available even sooner, makita had some packs with tabless 2.5ah 18650s already shown in catalogues but then removed them some time after, best guess is that they were delayed for some reason.
Click to expand...
Very interesting! I believe Makita almost exclusively uses Japanese (Sanyo, Murata) cells... lots of 'custom' cells that they don't post much about unfortunately (VX40, FormulaE as a couple examples).
 
thermal_runaway said:
Very interesting! I believe Makita almost exclusively uses Japanese (Sanyo, Murata) cells... lots of 'custom' cells that they don't post much about unfortunately (VX40, FormulaE as a couple examples).
Click to expand...

They do mostly use Murata cells but some packs also use Samsung, Sanyo and LG cells.
 
You must log in or register to reply here.

Similar threads

P
    • Like
Cycle life tests of High Power density cylindrical cells
2 3 4 5
Replies
118
Views
22,281
Pajda
P
P
    • Like
Cycle life tests of High Energy density cylindrical cells
3 4 5 6 7
Replies
173
Views
42,983
BlueSwordM
BlueSwordM
P
    • Like
Calendar ageing tests of mixed cylindrical cells
Replies
13
Views
2,996
Pajda
P
thunderheart
Samsung 21700 battery battle - 50G vs 48G vs 50E
Replies
0
Views
2,381
thunderheart
thunderheart
Sajeel
First eBike build with longevity in mind - Samsung 50G MTB Front Hub (Now rear hub)
2
Replies
47
Views
4,817
Sajeel
Sajeel
Back
Top