Thundersky charged to 4.2V and Normal LiFePO4 to 3.7 ?

Doctorbass

100 GW
Joined
Apr 8, 2007
Messages
7,495
Location
Quebec, Canada East
Thundersky cell charged to 4.2V :shock: ... or 3.7V ??? THAT'S THE QUESTION


I am starting this debate here because i still have not found a decent explaination about that.

1: EV People know that the Thundersky cells have specs that suggest to charge them to 4.2V

2: We always known .. or learned.. that the recommanded LiFePO4 cells charge voltage is max 3.7V

3: The confusion reside about the fact that conventional Lithium cells ( LiCo,LiPo,LiMn) charge to 4.2V

Recommanded Lithium cell Voltage Range, the sumarry :

LiCo: 3.0 to 4.2V

LiPo: 3.0 to 4.2V

LiMn: 3.0 to 4.2V

LiFePO4: 2.5 to 3.6... some say 3.7.. best is 3.65V



The MAX CHARGING VOLTAGE OF THE TS CELL confusion reside about many infos:

-the fact that Thundersky may have forget to update their specs for LiFePO4 that is 3.7V instead of 4.2V max.. since ALL conventionals lithiums charge to 4.2V.. ONLY THE LiFePO4 CHARGE TO 3.7V

-The fact that some are saying that charging a TS cell to 3.7V will not fully give the rated capacity... ( only like 85%)

-We know that overcahrging a lithium cell above the recommanded voltage may make it become indtable and have some plating and damadge it.

Now here is the arguments:

-Some are assuming that charging the cell until it reach 4.2V and then STOP the charge(no CC-CV mode is used) is the same as charging it in CC-CV to 3.7V and then to cut when current reach C/20 ( conventional chrging method for Lithium) SEE THE ATTACHED GRAPH


But since charging LiFePO4 cell to 4.2V seems to damadge it... Why TS NEVER SPECIFY their SPECIAL voltage or advise people about that particular difference?

Why Ts would be so different than ALL the other LiFePO4 cells?? How can it be usefull to recommand to charge these cells to 4.2V?? I mean.. They are the only manufactur to have specs that suggest to cahrge to 4.2V instead of 3.7V like any others.

I know many persons would use the conventional BMS that top or shunt the charge to 3.7V per cell.. so why it is so complicated with TS ?

Why not just recommand to charge in CC-CV to 3.7V to everybody just like all other LiFePO4 Manufacture?

I still believe that keeping max current until 4.2V is certainly not helping for the cell life ! right ?

According to Battery university website, charging to 4.1V instead of 4.2V a conventional 4.2V lithium cell really prolong the cell life:

So I just imagine how it can be for a 3.7V to be charged to 4.2V !!! :shock:

Most cells are charged to 4.20 volts with a tolerance of +/?0.05V/cell. Charging only to 4.10V reduced the capacity by 10% but provides a longer service life. Newer cell are capable of delivering a good cycle count with a charge to 4.20 volts per cell.



Any great source for a perfect answer to that?

Any opinion?

Doc
 

Attachments

  • TS vs LiFePO4.JPG
    TS vs LiFePO4.JPG
    24.9 KB · Views: 5,245
Glad you asked. I am curious about this also. Now that the 20ah cells are available more ebikes will probably be using TS cells.
 
guys,

Go here and download this video for an excellent lecture on thundersky batteries:

http://media.ev-tv.me/gastankbox-1280.mov

I have thundersky batteries 100 ah for my car.
http://www.youtube.com/watch?v=phvm0z1BUpY

I charge to 3.65v per cell. I really don't see this as a problem. I believe it's closer to 90 ~ 95% charged at this voltage. The specifications say to charge to 4.2v but if you do this, you run the risk of overcharging the batteries. Jack (lecturer) says there's a few miles extra range (in a car) between 3.65 and 4.2 v. It's really not worth the risk of damaging your batteries.
 
In answer to this same question from me a few months ago:
Dave Kois at EV Components said:
On the TS cells you never want to charge them over 3.8v per cell except for the very first charge. The SE cells should only charge to 3.6v per cell. You should base any calculation you make for pack size on 3.2v nominal for either cell because once a load is put to them they fall to 3.2v and stay there for the majority of the discharge curve.
 
Let's understand what happen in two cases ( 3.7 or 4.2V)

What i understand about that « FAST » charge is that TS suggest to cahrge to 4.2V keeping full current during the complete charge process to “bypass” the last step of charging process that take usually more time due to the current that don’t decrease when reaching 3.7V..

I mean normally in CC-CV to 3.7V, the last step of charging process is when you top at 3.7V and leave the current decreasing naturally to around C/20 to end the charge process.. but that last step is longer because the charging current is decreasing.

So I understand that they may want to keep current to the max during ALL the charging process, including the last step of charging and doing that all the way up to 4.2V.

Usually in CC-CV, that last step represent 85% of the capacity and around 25% of the total charging time. So by charging up to 4.2V at FULL current, that 25% time would decrease to 15% … making you saving 10% time

That FAST charging mode save you 10% time… and probably damage the cell for more than 10% I guess…

We know that the internal resistance of the cell increase at the end of charge.. that’s why current naturally decrease for the constant voltage step…

I am understanding that to compensate that internal resistance increase and to keep the current high, allowing voltage to 4.2V compensate and keep current higher.

But what that mean?...

RI increase… Higher voltage(4.2V)… high charge current (100%)…

MORE HEAT !!!


And more heat for 10% time of every charge may certainly affect the cell life.


That’s why I would recommend the old good “smart” charging method of CC-CV to 3.7V

Doc
 
I totally agree with you Doc, I think they are simply trying to cut out a step and save some time. The first BMI/LiFeBatt VMS unit did not have any sort of balancing, all it would do is wait until a cell got to 4.0V, and then cut the charger off. What they eventually found out is that the cells will go out-of-balance quicker, and that it is not good for the cell longevity to not do a proper CC/CV charge mode, with balancing shunts, and to set the CV point at 3.65V per cell.

-- Gary
 
We should discuss about a real charge curve and see how long it takes from 3.7V to 4.2V.

I charged a 4s headwaypack without BMS and one cell climbed up to 4 volt in zero time, while the others where on 3.7V.
Maybe it's better for the TS, because the duration above 3.3V is significantly shorter than with CC-CV method?

Any volunteers? ;)
-Olaf
 
Why not just testing that with one A123 cell like this:


Methode A

1a: taking 100% discharged cell to 2.5V and charge it to 3.7V CC-CV at 2.3A until current reach C/20 (115mA)

2b: note the final cell temperature immediatly when the cell reached C/20 current

3a: note the charged capacity

4a: discharge it to 2.5V at let say 2.3Ah ( 1C)

5a: note the discharged capacity


Wait until cell is back to the ambient temp


Methode B:

1b: taking 100% discharged cell to 2.5V and charge it to 4.2V at 2.3A and Stop charging process when reach 4.2V

2b: note the final cell temperature immediatly when the cell stop charging at 4.2V

3b: note the charged capacity

4b: discharge it to 2.5V at let say 2.3Ah ( 1C)

5b: note the discharged capacity




6: compare both discharged and charged capacity and temperature of each method A and B


I guess that with method B, the cell would heat a little bit more during the ramp of 3.7 to 4.2V duration.. than with the method A during the step of Constant Voltage


Doc
 
Spec sheet typo. Nothing more.
 
My interpretation of it is the cells will be damaged if they go over 4.2v.
Routinely charging them to 4.2v seems like a bad idea. Bad chinese translation?
Experience indicates that 3.65v or 3.7v is high enough to get near full capacity.
The basic chemistry is exactly the same for A123 and Thundersky.
 
I have at my job i have a pdf spec sheet wrote by A123 system that specify that for fast charging the A123 cell allow MAX 4.2V... i'll post it tomorrow.

The max Voltage range of a A123 cell is 2.0V to 4.2V

but the recommanded Voltage range is 2.5v to 3.6V

I still think that the 4.2V is for pople that want to shorten by 10% the total charging time.

Doc
 
Got interesting test data for you guys! :wink:

test conditions:

-Cell tested: (new green A123 cell from ebay sampled from a 1000 cell batch of my friend's project)
-Ambient temp: 22 celsius stable
-Charge current; 1C= 2.3A
-Discharge current 2C= 4.6A


Method 1: ( conventional LiFePO4 method)
charge method 1: CC-CV to 3.700V until C/20 is reached
discharge method 1: down to 2.500V

Results:
charge capacity was :2209mAh
charge time was: 1h 08min 13sec

Discharge capacity was: 2241mAh
Dscharge time was: 29min 18sec


Method 2: ( suggested thundersky or FAST method)
charge method 2: CC until 4.200V is reached
discharge method 1: down to 2.500V

Results:
charge capacity was :2249mAh
charge time was: 58min 45sec
Note:
3.7V was reached at 56min 34sec
2209mAh was reached at 57min 42sec ( same capacity as method 1)

Discharge capacity was:2244 mAh
Discharge time was: 29min 21sec



highlight:
method 2 shortened the charge time by 10min 31sec for a 1C charge to get the same charge capacity

That's 4093sec against 3462sec for the same 2209mAh dumped into the cell

Method 2 = 15.4% less time to get the same 2209mAh

And if we examin what happen for the total charging time up to 4.2V using method 2:

-total capacity charged to the cell: 2249mAh.. it's 1.8% more capacity
-total time to fully complete method 2 to 4.200V: 3525sec = 13.8% less time than method 1 and 1.8% more capacity


Conclusion:
Charging LiFePo4 cell to 4.2V and cut immediatly current, reduce the charge time by 13.8% at 1C

Charging LiFepo4 cell to the same capacity than normal CC-CV do by leaving voltage going over 3.7v reduce charging time by 15.4%...

I am surprized to see that time difference!... the cell got ( overcharged above 3.700V during the last 2min 11sec

The discharge capacity data are pretty close.. near the same for the method 1 or 2
Both method resulted in a 2241 and 2244mAh wich is unsignificant difference

Charging with method 2 using 4.2V that reduce by 13.8% total charge time STILL GIVE THE SAME DISCHARGED CAPACITY

But i dont know the long term possible damaging effect of using method 2 for the cycle life of a LiFePo4

Doc
 
Very interesting, did you do any temperature measurements on the cell during each phase?
 
olaf-lampe said:
Nice results DOC :D

Did you use an all new cell or was it already cycled several times ? IMHO a used cell would eliminate brake-in-effects in your capacity results.
I wonder how a 4c fastcharging would look like :shock:
-Olaf

Thay was a brand new cell ( sampled from a 1000 cell batch coming from Ebay for a friend project)
I could cycle one and retest for sure

I could try the same test for a 4.3C (10A) fast charging ( max allowed charge current from A123 system)

Doc
 
patrickza said:
Very interesting, did you do any temperature measurements on the cell during each phase?

No, i had not my IR thermometer on my hand but at 1C the cell never got hot or warm.. I just ensured that the ambient temp was always 22 celsius.

I may try again and note that temp.

The entire test took over 4 hours including thedelay between each step to let the cell stabilize

Doc
 
I don't think Thunder Sky's numbers are as mysterious as some want to make them out to be.

Thunder Sky simply provides the complete voltage range for their cells - they don't 'derate' them in their datasheets like other companies choose to do.

Yes - they CAN be charged to 4.3V in some conditions, but it's not necessarily something a user should do.

Start here: Thunder Sky Battery Manual

Page 15:
View attachment TS_15.jpg

Page 16:
"The discharge voltage of LFP battery can be 2.0V - 2.5V, which is not harmful to LFP battery."
"If keep the charging voltage not higher than 4.3V and discharging voltage not lower than 2.5V at normal temperature , the cycle life of LFP battery should be more than 2000 times."

Page 27 and 28 give voltage ranges at 25ºC (MAX 4.25V MIN 2.5V) and minus 35ºC (MAX 4.3V MIN 1.5V)

I think it's telling that Thunder Sky sets their own chargers to complete the bulk charging phase at 3.62V per cell.

As with other cells - 100% capacity is there to use, but one will get much longer life by using only 70% or 80% of capacity.
 
Thanks for testing that Doc.

It pretty much confirms what we thought before. I'm willing to wait 15% longer and not push my luck with the cells.

At very high charge rates, there would be the opposite of voltage sag, which may allow for going over 3.7v during fast charge.
 
Cool info. Nice data Doc! Now if you could do that 1000 times and post the AH capacity of each cell please. Thanks. :wink:

Probably if you charge CC/CV to 3.65 most of the time, but those few times when you are stuck with a dead pack and are in a hurry, the 4.2 limit can help in a pinch.
 
http://wikileaks.org/leak/toyota-prius-a123-car-fire-investigation-report-2008.pdf

Appendix D of this PHEV fire investigation report is Battery Pack Design Safety Guidelines (DRAFT) from A123-Systems. They list recommended and absolute voltages for their ANR 26650 cells. You'll recall that their public datasheet recommends charging to 3.6V. Their graphs still show 2.0-2.1V LVC, but they also (recommend or allow?) a 1.6V pulse cut-off on the low voltage side.

Compare this info with the numbers posted in the appendix:

Code:
                    Recommended   Absolute
Max Voltage        3.85              4.20
Min Voltage        1.60              0.50

So we have two different 'recommended' voltages - 3.6 and 3.85. Why?

One may suppose that what Thunder Sky is saying that it's 'OK' to routinely charge to 4.3V. But I think it's clear that they are simply publishing the ABSOLUTE voltages for their cells - which appear to be fairly well in-step with other manufacturer's numbers.

Keep in mind that for warranty, they want users to use THEIR chargers - not invent their own. And their TSL series chargers do automatically reduce current as charging progresses - and they shut down at about 3.62V per cell.

I wonder if we're trying to read too much into this.
 
AndyH said:
http://wikileaks.org/leak/toyota-prius-a123-car-fire-investigation-report-2008.pdf

Appendix D of this PHEV fire investigation report is Battery Pack Design Safety Guidelines (DRAFT) from A123-Systems. They list recommended and absolute voltages for their ANR 26650 cells. You'll recall that their public datasheet recommends charging to 3.6V. Their graphs still show 2.0-2.1V LVC, but they also (recommend or allow?) a 1.6V pulse cut-off on the low voltage side.

Compare this info with the numbers posted in the appendix:

Code:
                    Recommended   Absolute
Max Voltage        3.85              4.20
Min Voltage        1.60              0.50

So we have two different 'recommended' voltages - 3.6 and 3.85. Why?

One may suppose that what Thunder Sky is saying that it's 'OK' to routinely charge to 4.3V. But I think it's clear that they are simply publishing the ABSOLUTE voltages for their cells - which appear to be fairly well in-step with other manufacturer's numbers.

Keep in mind that for warranty, they want users to use THEIR chargers - not invent their own. And their TSL series chargers do automatically reduce current as charging progresses - and they shut down at about 3.62V per cell.

I wonder if we're trying to read too much into this.

Wow. Wikileaks ebike relevancy. :p

For what it is worth, in long parallel strings of "old" or "spurious" A123 18650's I am charging to 3.6v to 3.7v to guarantee full charge and string voltage stability.

It seems to work better than the standard 3.6v norm.

Thanks.
 
I havent read this whole thread yet, but I have read info online, by some techs that claim if you charge a lifepo3 cell to 4-4.2 volt range , by accident, it will not neccassirily ruin the cell because lifepo4 has the ability to " bleed off" the extra voltage it cannot take in , which means the cell wont always overheat or be internally destroyed , while a lipo cell does not have this same ability and is much more susceptable to being ruined by overcharging .


The one article I read, basically claimed that the lifepo4 cell, may show it accepted a charge of 4.2 volts, but thats just a meaningless measurement and when you take it off the charger, it will float back down to the 3.7 volt area quickly....so the lifepo4 cell did not really take in 4.2 volts....but a lipo cell will accept being overcharged allowing it to be ruined.

Now maybe if you constanlty charged a lifepo4 cell to 4.2 volts, it would eventually ruin it .
 
I am using TS 20ah. cells on my 2 trikes (24 cells on one and 20 cells on the other one) and use those TS balancers on each cell and also use TS chargers and they all light up when the voltage reaches 3.7 volts and starts the bleed down also when all the bleed leds come on the charger stops and does not start up again.
 
Back
Top