Samsung cell options: 40T, 50E, 50S, etc

[RagingGrandpa]

18 hours ago, alcatraz said:

Hey, is battery cooling going to be a thing on eucs soon?

Hopefully it will never be needed... EUC's are already very porky these days

[Rawnei]

19 hours ago, alcatraz said:

Please share your thoughts. The more the merrier.

On second thought maybe a heat camera won't work as well as a termistor glued to the cell, if the pack is in heatshrink. If you take off the heatshrink, it wouldn't be realistic. Heat could dissipate faster.

You'd want the delta and from it you could calculate what amp draw is safe for which ambient temperature.

Hey, is battery cooling going to be a thing on eucs soon?

Move to Sweden we have passive ambient cooling here. 😅

[RagingGrandpa]

6 minutes ago, cegli said:

It looks like the 50S are superior in every way to the 50E, except in price.  They used to be quite expensive and hard to get, which is why I don't think they weren't used previously.

Cycle life is great.  An independent test showed 4500mAH per cell left after 952 cycles with at 90 DoD at 5 amp discharge.  They are 25A rated cells, versus 10A 50e.

The Lynx looks like it's using a 36S4P configuration.  With the Lynx pack, if you're pulling 5400W on these, you'd be drawing from the cells at 10A.  At 10A and below, they look to perform similarly to the 50E (maybe slightly better from the spec sheet).  At 10A and above, the 50S does quite a bit better.  I'm guessing with hard breaking, hard acceleration, and/or high speeds, the 50s will get better battery life.  If you ease into all your acceleration and breaking, and don't go super fast, they'll probably be very similar.

50s probably has less voltage droop when the battery is low as well, so you'll be able to access more speed when the battery is close to dead.

Here's some good tests:
https://endless-sphere.com/sphere/threads/cycle-life-tests-of-high-power-density-cylindrical-cells.114473/

[cegli]

Another great comparison of 50E vs 50S:

The 50S is a monster.  Look at the lack of voltage droop and the capacity given.  Looks great.  30W is at about 8 amps, which I'd say is a reasonable analog for EUC usage.

More charts here:
https://imgur.com/a/ye3K3i1

Edited November 15, 2023 by cegli

[Tomallo]

18 minutes ago, cegli said:

More charts here:
https://imgur.com/a/ye3K3i1

Wow, what a great resource... Gotta love the flashlight enthusiast crowd

[alcatraz]

We'll see 53G cells soon too. 50S probably do better in racing wheels, and packs of 4P (up to about 1800/2400/2700Wh depending on voltage). Anything over that is better fitted with 53G cells. 

For the time being 53G cells appear to be about the same price as the 50S, or maybe a bit higher.

[alcatraz]

Then there are the Yuandong 6000mAh cells tested at 6130mAh that you might see in extreme range wheels.

20% more range for the same weight is pretty cool.

[alcatraz]

Just so people understand... I'd just mention that the 50S shine in those comparisons where the other cells produce more heat. 

If you turn down the amp draw the other cells catch up to the 50S, and some might surpass it. 

Turn up the amp draw some more and the 50S are going to drop down a bit towards the 40T. 

Just thought I'd mention this before people think they're better across the board. We've seen indications that the 50S age a smidge faster than some of the other cells out there.

Edited November 16, 2023 by alcatraz

[cegli]

Interesting about the 53G and Yuandong 6000mAH.  I'll have to look those up.

On 11/15/2023 at 5:19 PM, alcatraz said:

If you turn down the amp draw the other cells catch up to the 50S, and some might surpass it. 

Turn up the amp draw some more and the 50S are going to drop down a bit towards the 40T. 

Just thought I'd mention this before people think they're better across the board. We've seen indications that the 50S age a smidge faster than some of the other cells out there.

Do you have any charts/data that supports this?  I haven't seen anything that matches that.  The lowest draw test on the above data is 20W per cell (~5A) and they're still decently above the 50E and way above the 40T.  Cycle life is almost identical versus the 50E in the testing on endless sphere.  The 50S can take up to 6A per cell for charging, versus 2.5A on the 50E.  I don't see any reason to prefer 40T or 50E cells (besides price).

@20W per cell:
50S - 16.50WHr
50E - 15.25WHr
40T - 13.20WHr

I'm curious to see a 10W (~2.5A) test, but I bet the 50E and 50S are pretty much neck and neck and the 40T is still way behind.

I agree that 53G or 6000mAH cells might be better than the 50s, once they're available.

 

Edited November 18, 2023 by cegli

[alcatraz]

16 minutes ago, cegli said:

Interesting about the 53G and Yuandong 6000mAH.  I'll have to look those up.

Do you have any charts/data that supports this?  I haven't seen anything that matches that.  The lowest draw test on the above data is 20W per cell (~5A) and they're still decently above the 50E and way above the 40T.  Cycle life is almost identical versus the 50E in the testing on endless sphere.  The 50S can take up to 6A per cell for charging, versus 5A on the 50E.  I don't see any reason to prefer 40T or 50E cells (besides price).

@20W per cell:
50S - 16.50WHr
50E - 15.25WHr
40T - 13.20WHr

I'm curious to see a 10W (~2.5A) test, but I bet the 50E and 50S are pretty much neck and neck and the 40T is still way behind.

I agree that 53G or 6000mAH cells might be better than the 50s, once they're available.

 

I saw some data on cycle life that had the 50S age as fast or slightly faster than the competition (it's been reposted on this forum), even though the current draw was very low. Considering they put out less heat at those currents they should have slightly better cycle life. Heat = bad. 

It's not a reason to stay away. They're still on average better for many wheels. I just want to point out that there could be scenarios where you'd want something else. Like larger packs (lower currents), or extreme racing (extreme currents), or possibly wheels that you want to get a few more charge cycles out of.

This is anecdotal but I saw some 50S cells with 4400mAh tested capacity for sale already. They haven't been out for that long and that's pretty steep. In their defence, it could have been a crazy application with 1000 cycles in a short time. 

[RagingGrandpa]

10 hours ago, alcatraz said:

I saw some data on cycle life that had the 50S age as fast or slightly faster than the competition

I think everything is shown here in Page 1... supporting that 50S ages just as nicely as 50E and 50G for typical 1C discharging.
Did you mean some other test data? 

[alcatraz]

15 hours ago, RagingGrandpa said:

I think everything is shown here in Page 1... supporting that 50S ages just as nicely as 50E and 50G for typical 1C discharging.
Did you mean some other test data? 

That's not it I believe. There was a graph comparing different cell types and one axis was amount of cycles.

I remember that from it you could see that the 50S didn't have a slower capacity degradation considering they were being discharged at a lower C rate than the others.

Which led me to believe that when they're pushed to equal C rates the degradation ought to be faster than the others. It's an assumption but not an unreasonable one.

Meaning if you ride gently you'll have a comparable capacity degradation. But who would pay extra for 50S and not use the higher discharge rate? In those situations there could perhaps be a bit higher degradation that what we're used to see. But it isn't based on a lot of data . Call it just a hunch for now.

Before anyone thinks I'm out to bash the cells, I'm interested in buying some to build myself a lightweight pack for my older wheel. Like a 20S1P 50S pack for my Mten3. Save some weight. It came with 512Wh and I challenge anyone to deplete the pack before their feet get tired. 

Edited November 17, 2023 by alcatraz

[RagingGrandpa]

It's all here...

On 10/19/2023 at 6:25 AM, dimos15 said:

in the original spec sheet the battery made 250 cycles at 25A  https://www.dnkpower.com/wp-content/uploads/2022/07/SAMSUNG-INR21700-50S-Cell-Specification.pdf 

On 10/19/2023 at 6:32 AM, Critzlez said:

Yes, that's because the discharge to get that rating is 25A, which is extremely high for euc real world conditions.
They'll be more cycles if it's discharged and charged at less current, which it definitely will.
Plus the rating is based on 2.5V. Many wheels only go down to like 3.0V, so even more cycles is achievable. 1000 cycles will be easily achieved with these cells, especially with active balancing so you can charge up to 80% reliably

[cegli]

Yep, here is the 1C discharge from the first page.  This means 2 separate tests from 2 different sources showed the 50S aging better than the 50E/50G:

Regarding the 53G cells, I found this spec sheet:

Unfortunately, looks like they wouldn't be ideal (or safe?) for EUCs (assuming that spec sheet is accurate).  Charge of only 0.33C means 1.75A charge max (just like the 50g).  50E cells are 2.5A, 50S are 6A.  Braking hard on these would degrade the cells at best, potentially cause damage which could cause fires at worst.  Maybe with time, they'll make a 53E/S version.

Edited November 18, 2023 by cegli

[0000]

So there's a few things in this thread I want to take a stab at explaining so hopefully it results in less confusion.

On 11/15/2023 at 12:09 PM, cegli said:

30W is at about 8 amps, which I'd say is a reasonable analog for EUC usage.

You're not wrong, but the 30W, or any constant power test for that matter, are a little weird relative to PEV use in the sense that batteries are rated for a current (limit) not a power limit. Since voltage decreases with time, a 30W from say 4.2V to empty at 2.5V the initial current will be initial sag voltage * current = 30W (say 4.0V * 7.5A) to eventually decrease to 2.5V * 12A = 30W right as the test cycle completes. But if you're targeting an average 10A discharge, who's to say 30W constant power discharge isn't roughly equivalent as a stress test. I think I'm just biased preferring constant current tests.

I'd say conceptually we're more used to constant current which has correlates much better to EUC use in actuality as EUC motor controllers are programmed for current limits directly which can be translated directly to a maximum battery current limit. If they were programmed for constant power, you'd be able to get the same performance out of an EUC at 30% state of charge vs 100% SoC.
Edit: Nah, this wasn't quite right. EUC controller could easily be programmed at some power limit and it's really voltage drain that accounts for loss of top-speed, the most noticeable performance decrease a rider experiences as the packs drain.

Running a worst case constant current test such as 10A would be decent way to simulate "typical" 4P-pack EUC performance through this artificial test a hypothetical real world endurance racing condition or continuous top-speed runs until full battery drain.

On 11/15/2023 at 8:43 PM, cegli said:

The 50S can take up to 6A per cell for charging, versus 5A on the 50E

The 50E's are only rated for 0.5C or 2.5A/cell, faster charging being one of if not the primary reason to choose (and pay more) for 50S cells over the 50E/G/GB and other similar "standard" high capacity 21700 cells.
 

On 11/16/2023 at 11:14 PM, alcatraz said:

Which led me to believe that when they're pushed to equal C rates the degradation ought to be faster than the others. It's an assumption but not an unreasonable one.

Meaning if you ride gently you'll have a comparable capacity degradation. But who would pay extra for 50S and not use the higher discharge rate? In those situations there could perhaps be a bit higher degradation that what we're used to see. But it isn't based on a lot of data . Call it just a hunch for now.

As mentioned above, you'd also be paying extra for the higher charge current rating. It follows that capacity degrades quicker at sustained high discharge rates close to the limit, but 25A/cell is way outside of typical EUC performance concerns. Even in the 4P-pack configuration.

With regard to battery cycling using 6A - I've seen performance data to suggest otherwise, but if I was personally going to regularly take advantage of 6A fast charging, I'd rather run my own cycling test(s) to compare/contrast the impact of 2.5A vs 6A on the potential capacity degradation delta, but only if I cared super long term ownership without need for battery replacements somewhere way down the line. The 50E/G/S 1C discharge chart above doesn't capture how 6A charging could impact comparative capacity degradation as it only uses the standard 0.5C charge rate.

The possibility for faster degradation from faster charging could be pretty small, and it's a big boon to not wait around and fast-charge before a ride if you got the charge capacity to take advantage of the rating.

On 11/17/2023 at 4:06 PM, cegli said:

This means 2 separate tests from 2 different sources showed the 50S aging better than the 50E/50G:

Technically they're actually at equal capacity at around cycle 700 when the 50E overtakes the 50S and cycle 850 when the 50G overtakes the 50S, understandably it's just a little hard to parse this out as each 50th cycle allows the cells to cool.

In reality, every cycle would be much more closely matched throughout on this chart if they were given more time to cool/rest between cycles which is far more realistic for actual EUC use as no one is constantly riding/charging/riding with only 30 minute breaks for 950+ cycles .

On 11/17/2023 at 4:06 PM, cegli said:

Unfortunately, looks like they wouldn't be ideal (or safe?) for EUCs (assuming that spec sheet is accurate).  Charge of only 0.33C means 1.75A charge max (just like the 50g).  50E cells are 5A, 50S are 6A.  Braking hard on these would degrade the cells at best, potentially cause damage which could cause fires at worst.  Maybe with time, they'll make a 53E/S version.

Definitely reasonable and feasible concerns. All reasons not to hype new cell chemistry that looks good on paper from the perspective of 1 or 2 specifications alone.

 

Edited November 19, 2023 by Vanturion
added a couple corrections

[cegli]

4 minutes ago, Vanturion said:

The 50E's are only rated for 0.5C or 2.5A/cell, faster charging being one of if not the primary reason to choose (and pay more) for 50S cells over the 50E/G/GB and other similar "standard" high capacity 21700 cells.

Ah yes, my mistake!  My eyes bounced on the spec sheet down to the 1C (5A) discharge.  Let me edit my post not to confuse people further.

[dimos15]

On 11/1/2023 at 4:07 PM, Jason McNeil said:

Sent a pair of 50E/50S packs, 30s2p, to an independent lab for validation under three constant loaded conditions: 10A/20A/30A—30A was the upper limit of the 50Es. Still a bit of a work in progress for the presentation, temperature probe placement was not perfect, but here's the initial comparison graph. 

 

It is very concerning that the 50s gets hotter than the 50e..  I expected the 50s to get hotter than the high discharge cells but not the 50e..

Also what is the ambient temperature for this test? It looks like they are cooling the batteries because the 50s can reach 73°C at 20A

 

And here is another cycle life test, the 50s has the biggest energy and power loss after 50 cycles at 35A

Eucs these days are very fast and they draw more amps from the batteries. Check this speedrun , almost pulling 100A. But most importantly we need brakes . We need batteries that can take a lot of charge without overheating.  High charge/discharge batteries are the way forward in my opinion.

Edited November 26, 2023 by dimos15

[dimos15]

Oh i forgot, there is a new, very good high discharge battery made in china https://www.tenpowercell.com/wp-content/uploads/2022/12/TP-INR21700-40TG-Product-Datasheet-v1.4.1.pdf

molicells are better though

Edited November 26, 2023 by dimos15

[Rawnei]

3 hours ago, dimos15 said:

Eucs these days are very fast and they draw more amps from the batteries. Check this speedrun , almost pulling 100A. But most importantly we need brakes . We need batteries that can take a lot of charge without overheating.  High charge/discharge batteries are the way forward in my opinion.

What you see there is phase amps not current drawn from batteries, also the wheel draws current from multiple battery packs in parallel so the current is spread out.

3 hours ago, dimos15 said:

And here is another cycle life test, the 50s has the biggest energy and power loss after 50 cycles at 35A

Cycle loss at 35A is not really relevant for our use case, as it says right there in the picture you shared "torture test".

[Panzer04]

Most EUCs are currently using 50Es, which would get completely blown up in a torture test like that anyway. If you genuinely want very high current performance the 50S is not really the right cell - P45Bs are - but EUCs are probably the textbook application for a master of none like the 50S. Occasional extreme current draw/charge (accelerate/brake), with fairly low steady-state usage.

[dimos15]

6 hours ago, Rawnei said:

What you see there is phase amps not current drawn from batteries, also the wheel draws current from multiple battery packs in parallel so the current is spread out.

You are right, i thought after all these years that they changed it but no.. Still call it current..

I still think that they can draw 90-100 battery amps , around 25A per battery .

6 hours ago, Rawnei said:

Cycle loss at 35A is not really relevant for our use case, as it says right there in the picture you shared "torture test".

How is it not relevant? Even from the spec sheets you can see that the samsung 40t completes the same test with 10A higher discharge!

It is obvious that the 40t will have higher cycle life than the 50s at the same amps..

This is 50s

This is 40t

[dimos15]

1 hour ago, Panzer04 said:

Most EUCs are currently using 50Es, which would get completely blown up in a torture test like that anyway. If you genuinely want very high current performance the 50S is not really the right cell - P45Bs are - but EUCs are probably the textbook application for a master of none like the 50S. Occasional extreme current draw/charge (accelerate/brake), with fairly low steady-state usage.

Sure but we got a step in the right direction by using molicells and 40t on some gotway wheels. Why are we going backwards now?

I don't want p45, i want p42 like every other performance electric vehicle.

I want to be able to charge fast and brake harder without overheating lol https://youtu.be/j3BjHuH3QNY?feature=shared&t=2909

The 50S runs hotter than the 50e

[goatman]

the P42 is a better cold weather cell than the 40T

the 40T likes regen and 10 amps, its a performance cell you can continuously discharge/charge/discharge without suffering cycle life while doing that more than you could in a day.

the p42 likes it hold or cold but not 10 amps, it cant get up to temperature for cycle life, it would rather see 12 amps right away to get its temperature up

the p42 isnt chinese and the euc makers cant get a guaranteed supply of molicels so they quit using molicel

you should want the P45 

if you ride when it isnt always sunny and 75 or warmer you shouldnt get a 40T

you should only want these cells for 4p packs if you ride hard, theyre high performance cells that perform/degrade like crap if theyre not being used  the way theyre designed to be used.

you see the self healing every 50 cycles is more pronounced than the 50S, it would be interesting to see someone do a continuous 10amp discharge/5amp charge/10amp discharge 10x in a row to see if it suffers any cycle life loss

Edited November 27, 2023 by goatman
add chart

[Rawnei]

5 hours ago, dimos15 said:

How is it not relevant? Even from the spec sheets you can see that the samsung 40t completes the same test with 10A higher discharge!

It is not relevant because the wheel don't draw that much current from a single cell, so a cell torture test doesn't really tell us anything about our use case.

Another tidbit from that torture test the cells are discharged to 2.8V our wheels don't allow that low either.

If we want real tangible data the test have to be closer to our use case.

 

5 hours ago, dimos15 said:

I still think that they can draw 90-100 battery amps , around 25A per battery .

I don't think so, think about it that would mean that most wheels are drawing much more current than what the cells they are using are rated for and we would see many problems.

You're worried that a high discharge cell won't be enough for our use-case meanwhile most wheels today use high capacity cells...

Edited November 27, 2023 by Rawnei

[Rawnei]

On 11/1/2023 at 3:07 PM, Jason McNeil said:

Sent a pair of 50E/50S packs, 30s2p, to an independent lab for validation under three constant loaded conditions: 10A/20A/30A—30A was the upper limit of the 50Es. Still a bit of a work in progress for the presentation, temperature probe placement was not perfect, but here's the initial comparison graph. 

 

How are these tests coming along? What conclusions did you draw?