A123 prismatic cells - THE solution?

[999zip999]

My A123 20ah cells can only deliver 18ah. But for 840 cycles. Oh so far. And still 18ah. No bms very hard regen. May be that the regen is awaking them up. With a charge pulse. Just a guess.

 

[thunderstorm80]

My A123 20ah cells can only deliver 18ah. But for 840 cycles. Oh so far. And still 18ah. No bms very hard regen. May be that the regen is awaking them up. With a charge pulse. Just a guess.

Regen pulses have nothing to do with awakening a cell. It's just that those cells are the only commercial Li-Ion technology which has the least resistance, the least hysteresis, and the most life cycles.
They are also the only commercial Li-Ion which can accept up to 4C charge rate (by the spec), which also explains your impression.
I also don't use BMS - I charge them to 3.5V@cell, and do manual balance charge once in a while. I learned that even 3.4V@cell is enough, but then the CV stage is longer and you may have some cells which are undercharged (<3.4V@cell) if you are out of balance.
I am still surprised why people use gun powder as their power source.... (LiPo's)

 

[Hillhater]

I am still surprised why people use gun powder as their power source.... (LiPo's)

b
2 reasons mostly..
Cost
Weight

 

[999zip999]

The high pulse regen Awakening the cells was a bit of a joke but these cells can take a high regen without suffering. 1,054 cycles all 24 are 3.48v - 3.56v bulk charging @ 16amp. Safe and sane. But heavy and big.

 

[madin88]

LiIon has about two times or more watt hours per weight and volume. this means if you have 1000 charge cycles on a LiFe battery, you would have only 500 on a LiIon battery.

I once thought about using A123 M1B for one high power bike because they can be charged extremely quick and are ery safe, but if you compare all Pros and Cons, LiIon makes much more sense for EV so i went with such.

 

[999zip999]

I was looking at nmc cells like weststart and sent money but they said a company had got all there production and sent me back my money plus 52.00. I didnt understand. Would like a smaller lighter 20ah cell @ 10c.

 

[thunderstorm80]

LiIon has about two times or more watt hours per weight and volume. this means if you have 1000 charge cycles on a LiFe battery, you would have only 500 on a LiIon battery.

I once thought about using A123 M1B for one high power bike because they can be charged extremely quick and are ery safe, but if you compare all Pros and Cons, LiIon makes much more sense for EV so i went with such.

Li-Ion has LESS than two times the energy capacity (200Wh/Kg vs 133Wh/Kg for A123), but even that only until the aging effect AND the cycle-count brings it in PAR with A123. This takes around 2 years. It doesn't stop there, but keeps getting progressively lower while A123 declines WAY SLOWER.
I can give you a solid example: I have A123 26650 cells I bought in the beginning of 2011, almost 7 years ago. Each cell rated as 2.3Ah.
Today, I get 1.88Ah from each cell. This is a 20% decline in almost 7 years, which is about 3% loss per year.
3% per year while the cells have always being kept at 100% SOC when not in use, and I brutality pulled out of them 20C continuous discharge rates and 4C discharge rates during regen. (I measure 1.88Ah at 10C discharge rate by taking the E-bike uphill without pedalling. With slower discharge rates I assume I would have seen higher numbers)
I don't know of any other Li-Ion chemistry which can accept 4C charge rates without degrading it's life.
Regarding your two times capacity statement, it's true only for a certain types of Li-Ions which trades even further their life-cycle count and the allowable C discharge rates in favour of capacity. I find those even less appealing, unless you have the money to replace your battery pack every 1.5years or so. Those cells diminish in capacity even faster than the standard 200Wh/Kg packs which I talked about earlier.
And I didn't talk yet about the safety, and the much faster balance-drifting issues with non A123 cells.
For me, it's only cons. The weight save is not a "pro" because it's only very temporary.

 

[madin88]

Some years ago the A123 M1A/M1B cells had been hyped for RC model usage because of theire robustness and the possibility for charging / discharging them at extreme high levels. I personally have used them too, but replaced them later by LiPos because of the much higher energy and power density.

If sombody is looking for robustness, safeness and extremely high cycle life, A123 is a winner for sure.


Lets compare the M1B (76g) with a Samsung 30Q (48g) at 10A discharge:

regarding http://www.dampfakkus.de, the 30Q performs following: 2,93 Ah, 9,84Wh (down to 2,5V) -> 205Wh / kg

As it will degrade quick when using it 4,2-2,5V, i usually program my controller between 4,15-3,3V which results in about 15% less usable Wh, but will give 3 times higher cylce life (many times proven).

ok, than it has 175Wh / kg

The M1B performs following from my calcuation when looking at discharge graphs: 7,65Wh at 10A when discharged down to 2V
https://images-na.ssl-images-amazon.com/images/I/711o+u+UACL._SL1500_.jpg

so it has 101Wh / kg which is only 57% of the 30Q

now think further:

you could use 75% more 18650 cells in terms of weight (and probably volume too) so the stress on the battery will be reduced further (like 6A discharge current per cell vs the previous 10A, less stress during regen), and in the end you have many many times the cycle life of initial specification :wink:

M1B is no more attractive than if all considered, and after working with more than 3000 18650's (mostly from Samsung) i can give following adive:

If the cells come from the SAME CHARGE, if the pack is BUILT PROPERLY (mostly i terms of equal current flow through every single cell), and if you don't always squeeze out all the capacity, it will stay balanced very very well and will last for years.

check out this diagram about degradiation of tesla batteries:

less than 10% in 8 years is something i could live with.

https://electrek.co/2016/11/01/tesla-battery-degradation/

 

[thunderstorm80]

Some years ago the A123 M1A/M1B cells had been hyped for RC model usage because of theire robustness and the possibility for charging / discharging them at extreme high levels. I personally have used them too, but replaced them later by LiPos because of the much higher energy and power density.

If sombody is looking for robustness, safeness and extremely high cycle life, A123 is a winner for sure.


Lets compare the M1B (76g) with a Samsung 30Q (48g) at 10A discharge:

regarding http://www.dampfakkus.de, the 30Q performs following: 2,93 Ah, 9,84Wh (down to 2,5V) -> 205Wh / kg

As it will degrade quick when using it 4,2-2,5V, i usually program my controller between 4,15-3,3V which results in about 15% less usable Wh, but will give 3 times higher cylce life (many times proven).

ok, than it has 175Wh / kg

The M1B performs following from my calcuation when looking at discharge graphs: 7,65Wh at 10A when discharged down to 2V
https://images-na.ssl-images-amazon.com/images/I/711o+u+UACL._SL1500_.jpg

so it has 101Wh / kg which is only 57% of the 30Q

now think further:

you could use 75% more 18650 cells in terms of weight (and probably volume too) so the stress on the battery will be reduced further (like 6A discharge current per cell vs the previous 10A, less stress during regen), and in the end you have many many times the cycle life of initial specification :wink:

M1B is no more attractive than if all considered, and after working with more than 3000 18650's (mostly from Samsung) i can give following adive:

If the cells come from the SAME CHARGE, if the pack is BUILT PROPERLY (mostly i terms of equal current flow through every single cell), and if you don't always squeeze out all the capacity, it will stay balanced very very well and will last for years.

check out this diagram about degradiation of tesla batteries:

less than 10% in 8 years is something i could live with.

https://electrek.co/2016/11/01/tesla-battery-degradation/

Sorry for the very late reply...
I would like to know where you can buy such cells as the Tesla ones.
Even the famous 30Q Samsung cells will lose significant portion of the capacity after 250 cycles if you fully discharge them and at their max discharge rate.
Your calculations regarding the specific energy are wrong - First, the prismatic cells of A123 have 133Wh/Kg, but we can assume 120Wh/Kg on practice. It's the 26650 cells that sit around 100Wh/Kg. Second, if you lower your maximum charging voltage to increase the life cycle so you are at around 175Wh/Kg, then you are not having double the density of A123 - even if we talk about the 26650 cells. With the prismatic cells, which can be brutality discharged, the difference is 175/120 - much less than a factor of 2.
And last, after 1-2years you would have the same density as A123 pack, due to aging and life-cycle capacity loss, but your clock doesn't stop there. You will keep getting lower and lower capacity while the A123 chemistry stays almost completely fresh.
I am sorry, but I would like someone to show me, A REAL, and PURCHASABLE battery cell, which has similar properties to the Tesla spoken above. I didn't find one. It's either high specific energy (220Wh/Kg or so) by paying for much less cycle life, or lower density of 170Wh/Kg but still getting 500 cycles if you use your battery to it's max.
It doesn't come any closer to my A123's thousands, if not 10000, of life cycles.
As I said before - my old 26650 2300mAh cells bought in the beginning of 2011, have currently 1850mAh remaining. Do the math and see for yourself. Yes, they have now only around 78Wh/Kg, but we talk about 8.5 years of abusing them with up to 20C discharge currents, 4C charging regen currents, and almost daily full cycles. They still accept those high currents, and are barely warm after each such deep full discharge cycle.
Any other chemistry is not economical at ALL in the long run. The glory of that high-energy density is a very temporary one, with a sad end (mostly to your bank account).

 

[liveforphysics]

To my knowledge, not a single vehicle OEM in the world is using A123 products today for an EV traction battery (I think a couple Chinese cars use them for start-stop ICE batteries today).

A123 was unquestionably an industry leader when the M1 cell first emerged 11-12 years ago.

While holding that tech/industry leadership position, they decided rather than evolve the product to stay relevant, to instead join a handful of big industries in getting fleeced by lawyers selling them "IP" kool-aid and lawsuits (the function of IP).

Anything you choose to cling to and "protect" in the exponential age is your own ensured obsolescence. However much you involve lawyers in your product or company is how much you want to get fleeced in exchange for non-contributing harms creation for yourself and others.

Today A123's LFP cathode tech is abandon-ware and someone there with vision partnered with solid-energy which uses a modern cathode interfacing a solid electrolyte, which is at least a useful RnD path forward, and I wish them the best.

 

[thunderstorm80]

To my knowledge, not a single vehicle OEM in the world is using A123 products today for an EV traction battery (I think a couple Chinese cars use them for start-stop ICE batteries today).

A123 was unquestionably an industry leader when the M1 cell first emerged 11-12 years ago.

While holding that tech/industry leadership position, they decided rather than evolve the product to stay relevant, to instead join a handful of big industries in getting fleeced by lawyers selling them "IP" kool-aid and lawsuits (the function of IP).

Anything you choose to cling to and "protect" in the exponential age is your own ensured obsolescence. However much you involve lawyers in your product or company is how much you want to get fleeced in exchange for non-contributing harms creation for yourself and others.

Today A123's LFP cathode tech is abandon-ware and someone there with vision partnered with solid-energy which uses a modern cathode interfacing a solid electrolyte, which is at least a useful RnD path forward, and I wish them the best.

I still find their abandon-ware the best up-to-date "ware" for propelling a hungry E-bike with large regen surge currents.
I live in a hilly city, and I see regen powers in the order of 2KW when I am using the motor (TC4080) to brake. If you happen to go on -10% downhill at 50-60kph, then this 2KW can last more than just a few seconds, and the A123 prismatic cells I have swallow it all without a problem. I barely have to apply the mechanical brakes as there is no regen current limit needed. I routinely see regen currents on the CA in the order of -40A or so. This would kill any other Li-Ion in no time, unless I parallel several of those which will result in a very, VERY large pack...

Here is another story:
I bought those prismatic cells second hand from a friend who used them on his electric motor-cycle. He was very pleased with the performance but wanted more range, so he bought one of those "non abandon ware" modern Li-Ions they use in modern electric cars. He quickly found he made a terrible mistake, as his range got progressively shorter and shorter, to the point where he sold the electric motor-cycle and bought a gasoline one. With the A123's the range he achieved was staying constant, but lower compared a new modern "non abandonware" Li-Ion. However, after some time, the modern Li-Ion declined in it's range below what the A123 gave, and this is my point. (Unless you are insanely rich and don't mind replacing your entire pack every 1.5-2 years)

This is why I said - I would like someone to show me where you can buy cells with performance as the Tesla cells shown above.

 

[liveforphysics]

Glad you like them. I've never made a bike with less than 1.8kWh onboard, so regen loads have never been a concern, but no doubt if you have a tiny pack with aggressive charge and discharge, they can work out well.

 

[madin88]

Sorry for the very late reply...
I would like to know where you can buy such cells as the Tesla ones.
I am sorry, but I would like someone to show me, A REAL, and PURCHASABLE battery cell, which has similar properties to the Tesla spoken above. I didn't find one. It's either high specific energy (220Wh/Kg or so) by paying for much less cycle life, or lower density of 170Wh/Kg but still getting 500 cycles if you use your battery to it's max.

It is nothing special, they just don't charge and discharge the cells to 100% and take care about the temperature, so they get many many times the cycle and calendar life of initial specifications
The BMS of your normal e-bike battery won't stop charging at 90% and discharge at 20% e.g, and the current flow often is not equally shared through all cells, so you usually need to replace it every few years.

Tesla packs are made of 18650 size Panasonic PF or Panasonic B cells and thats NCA chemistry IIRC. Just the same as the higher current Samsung 25R or 30Q cells which are a great choice for high powered e-bikes.

So forget the datasheets because it is bullshit. The manufacturers do not want the cells to last a very long time because they want to sell every few years a new battery to you.
--> used as specified means you are already abusing them, but you can really easy extend the lifetime by a huge factor if you know what should be done.

Any other chemistry is not economical at ALL in the long run. The glory of that high-energy density is a very temporary one, with a sad end (mostly to your bank account).

Yes you have right about the factor two that it isn't correct for the larger 20Ah pouch cells, but if you do a comparison with the newest 20700 or 21700 format it will come close to that.
I am aware that A123 have an awesome cycle life and can take alot of abuse, and the difference to most other LiIon manufacturers is that you CAN TRUST the specifications they make.
However, if you know for yourself what you can do to extend the lifetime of a LiIon battery, it clearly is winning for EV applications.

About 3 years ago i built one 20s12p 2kWh battery out of 25R cells.
I take over 140A (10kW+) during acceleration or when climbing hills. Most of the time i fill it up with 1,6kW (3/4C), and during hard regen i see 3kW+
The internal resistance is the same as it was 3 years ago (40-45mOhm), and i only noticed a very small drop in range.
Cell drift is less than 5mV which is also same as from beginning on (NEVER EVER balanced!)
It seems to follows the chart from tesla^^

 

[Cephalotus]

What happened to this company and why everybody recommends LG/Samsung cells that although have an initial higher specific energy at the beginning of their service, they lose the advantage to the A123 after 1.5-2 years and keep declining while the A123 stays intact. (The A123 with their LiFePo4+nano technology have a negligible aging effect)

My Sony and Samsung cells that are now 6-10 years old are still good, too.

My oldest Li-Ion cell is a prismatic LCO cell from Sharp that is noew 19 years old, hasned#t been in use since 18 years but offers still more than 90% of its initial capacity.

 

[Cephalotus]

Sorry for the very late reply...
I would like to know where you can buy such cells as the Tesla ones.
Even the famous 30Q Samsung cells will lose significant portion of the capacity after 250 cycles if you fully discharge them and at their max discharge rate.
Your calculations regarding the specific energy are wrong - First, the prismatic cells of A123 have 133Wh/Kg, but we can assume 120Wh/Kg on practice. It's the 26650 cells that sit around 100Wh/Kg. Second, if you lower your maximum charging voltage to increase the life cycle so you are at around 175Wh/Kg, then you are not having double the density of A123 - even if we talk about the 26650 cells. With the prismatic cells, which can be brutality discharged, the difference is 175/120 - much less than a factor of 2.
And last, after 1-2years you would have the same density as A123 pack, due to aging and life-cycle capacity loss, but your clock doesn't stop there. You will keep getting lower and lower capacity while the A123 chemistry stays almost completely fresh...

A LG 18650HG2 starts at 3,0Ah and after 600 cycles at 15A discharge (5C) and 4A charge from 4,2V to 2,0V(!) you will still have 2,4Ah of the initial capacity.

https://www.nkon.nl/sk/k/hg2.pdf

So the smaller 18650 cell will have higher energy densitiy comapred to a NEW larger(!) 26650 LFP cell after 600 heavy abuse cycles.

I never need more than 5C in an electric vehicle (what for? The battery would be empty after 12 minutes) and 1,33C charging is much more than most ppeople well ever need for an ebike.

Such a cell costs less than 50% than your A123 cell and is easier to charge.

Name me one single reason why I should built a battery pack from A123 cells that is worse tan a very old and abused pack of those cells after many years, is more difficlut to solder and costs more than twice the price?

For stationary ultra lifetime applications and ultra power application a Tohsiba SCiB or sony LTO is the prefered cell.

there is some reason why nobody in the vehicle world is intereestd in those A123 cells and why they went bankrupt.

 

[999zip999]

Ouch we know A123 is best suited for a 3-wheel or golf cart but still 16s is great for a bike. At a good price yes they're very expensive. But I got 20ah pouch cell for 25usd shipped 4 years ago, but now 70.00 ea just to much. But now I'm looking for a light weight option like high quality 18650 or a NMC 20ah pouch. Did look at weststart but they took my money than said have all production for One customer now. Then sent all my money back and then sent another $52 to me. Still looking for a good battery option I just don't like all the interconnections of that 18650.
So not good for you maybe a great fit for you wife or father-in-law there's something to be said for Fail-Safe.