A123 Battery Handling Guidelines

All,

I think you'll agree this is a fantastic reference document for all things A123, and many thing LFP. It answers virtually every FAQ I can think of and then some. I hope this is helpful to the community. Enjoy, and please....don't harass the author or the service group via email.

Thanks,
dh
 
all other sources of information on lifepo4 chemistry indicate that full charged is 3.65v and fully discharged is 2.1V but this document uses 3.6V as fully charged and 2.5V as fully discharged. i know there is very little difference but is there a reason they used these numbers?

is this because it is not a technical document but kinda like a manual for people to read in their safety classes? or does A123 cathodic material behave differently from the other forms of lifepo4? is there a real risk if the cell is charged above 3.7V? has that risk been documented?
 
dnmun said:
all other sources of information on lifepo4 chemistry indicate that full charged is 3.65v and fully discharged is 2.1V but this document uses 3.6V as fully charged and 2.5V as fully discharged. i know there is very little difference but is there a reason they used these numbers?

is this because it is not a technical document but kinda like a manual for people to read in their safety classes? or does A123 cathodic material behave differently from the other forms of lifepo4? is there a real risk if the cell is charged above 3.7V? has that risk been documented?
Click to expand...

Note first that the numbers in this guide are for OCV after 30 minutes rest. An A123 cell that comes to rest higher than 3.60V will plate Li and lose capacity at an accelerated rate as a result. The higher the voltage, the worse it gets until you get above 4.0V and then we start talking about electrolyte breakdown and potential for thermal events. We allow charge pulses up to 3.8V for 10 seconds, but if you come to rest higher than 3.60, you are overcharged. That's the simple answer....

dh
 
i lot of the literature is incorrect if this is true. i have seen more than a few different BMSs that did not even allow the cell to drop to 3.60V before turning the charge back on again after it had shut off at the 3.9V mark. many of them oscillate between the 3.9v and 3.8V reset voltage.

i just wondered if this is specific to A123.
 
dnmun said:
i lot of the literature is incorrect if this is true. i have seen more than a few different BMSs that did not even allow the cell to drop to 3.60V before turning the charge back on again after it had shut off at the 3.9V mark. many of them oscillate between the 3.9v and 3.8V reset voltage.

i just wondered if this is specific to A123.
Click to expand...

Yes, those cheap BMSs will do that.

I think it is better to use a good balance charger and keep a good eye on the charging to make sure that does not happen.
 
dnmun said:
i lot of the literature is incorrect if this is true. i have seen more than a few different BMSs that did not even allow the cell to drop to 3.60V before turning the charge back on again after it had shut off at the 3.9V mark. many of them oscillate between the 3.9v and 3.8V reset voltage.

i just wondered if this is specific to A123.
Click to expand...

I can show you a whole bunch of DIY car guys over on Prius Chat who haven't been able to figure why their LFP (mostly non-a123) cells don't last like they should. I believe the culprit is exactly what you describe. Common cutoff voltage there is 3.8, with a charge rate of .2C (c/5). That is a formula for adolescent cell death, and many have seen it come to fruition.

I can't speak to what other folks' cells will do, but I can tell you with absolute certainty that these numbers are good for A123 cells.

On the LVC side, remember that 2.5V is at rest. Under load, as low as 1.5V should not scare anyone Just make sure you rebound to 2.5, and EVEN THEN it's not a big deal if you don't bounce all the way back--just don't leave it like that for any period of time; charge it back up. Again, this is A123....don't know about others.

dh
 
My A123 20ah seam be be full at around 3.45v and between 3.45v to 3.59v it is off the the races as the voltage goes up so fast. So there can't be to much energy of there. Do bms's need the be set so high for the A123 20ah format cell ?
 
999zip999 said:
My A123 20ah seam be be full at around 3.45v and between 3.45v to 3.59v it is off the the races as the voltage goes up so fast. So there can't be to much energy of there. Do bms's need the be set so high for the A123 20ah format cell ?
Click to expand...

If you can afford the unused overhead, I would top balance at 3.5V or so for max longevity (at rest, this is something like 95% SOC or more). In some applications it makes sense to target a "nominal" SOC that is quite low--it allows maximum opportunism for regen, and extends the life of the cells considerably compared to parking them up high all the time. This is not really an option for any kind of "plug in" options though. The take away there is the lower the avg SOC, the longer the cell's cycle and calendar life. Mind the temperature extremes and voltage extremes and these things will last a very long time, even with very hard use and regular, properly managedcharging to 100% SOC.

dh
 
Thank you for this guide wb9k and arkmundi.

From now on, I will balance charge to 3.5V. and use the BMS only for discharge.
 
it does appear that all the lifepo4 BMS use the 3.60V as the balancing voltage. all the ping packs were charging to 58.4V too and they lasted years and years.

it would help a lot if there were some documentation to understand why this is so far off from the literature. also i wonder if this is also the case with the lipo and that the charging in excess of full charge at 4.2V is harmful too. this is something everyone talks about but there has never been any references to the literature where it was shown that charging to full voltage is deleterious.
 
dnmun said:
it does appear that all the lifepo4 BMS use the 3.60V as the balancing voltage. all the ping packs were charging to 58.4V too and they lasted years and years.

it would help a lot if there were some documentation to understand why this is so far off from the literature. also i wonder if this is also the case with the lipo and that the charging in excess of full charge at 4.2V is harmful too. this is something everyone talks about but there has never been any references to the literature where it was shown that charging to full voltage is deleterious.
Click to expand...

I think if you read around enough you'll find that Li plating at higher voltages is a well-established phenomenon in LFP. I have seen several papers on it not published by A123. Search google for lithium plating LFP and you'll find all kinds of stuff, including photos showing what it looks like at the macro level, all the way down to shots taken with a SEM.

It's critical to understand that terminal voltage rise is proportional to the charging rate. If you're charging at 2, 3, or 4 C, you'll hit 3.8V at the terminals way before you are at 100% SOC. At that point, charge current has to be throttled back so that Voltage is held to 3.6 or just above on the highest cells. Our cells are considered to be at 100% SOC when terminal Voltage is held at 3.60V and charge current has fallen to 1/20C. These numbers aren't some static target on a chart, the numbers' meanings vary greatly if the cell is at rest vs under load or being charged, whether it was being charged or discharged last, and at what rate. Many BMS's have no idea how fast you're charging and are thus unable to interpret those numbers intelligently. It often makes sense to use custom settings with a BMS tailored to the specific cells you are using after consulting with the manufacturer's specs. This is not always easy since specs can't always be had, and when they can they are often ambiguous or confusing. It's one of many problems that tend to affect the whole industry that really needs to be cleared up with batteries coming into use in more and more "critical" applications, I guess you could say. The automotive world definitely demands a higher level of predictability and robustness than consumer products like cell phones, computers, and cameras, especially when you're talking about the drivetrain.

I hope I'm not insulting you with basics here, but Li plating just is not an obscure issue with overcharged LFP.

dh
 
End of charge, that is what your talking about right ? I have found that you can not slam the cells at 10 or 5 amps until they hit 3.65 volts, (headways cells) and call theme charged. Not so according to my crude load testing one cell at at time, tedious as hell ( don't ask ) 80% charged at best. Hold them at 3.65 volt until amps drop to .100 amp as specified and all deliver as expected

I read lot about HVC and LVC, but do not see much about HVC to end of charge ?? I think this is a real cause of imbalance of packs that I read about and experience my self. What do you think.

Wish I could post that bitching E bike! My passion for this is storng!!

Pete
 
that helped, thanks a lot. googled up the plating of lifepo4 and have read one good paper, i mean really good and well worth reading:

www.mdpi.com/1996-1944/6/4/1310/pdf‎

and i read his comments in the prius chat and just wanted to repeat what he said there because it merits respect and praise:

>>> I agree the specs are not generally well written; they tend to be much too vague. There can be many reasons for this, but most common seems to be a lack of knowledge on the part of the manufacturers themselves, or a lack of a sense of responsibility. A123 is unique in that we try to deliver whole systems to customers rather than just supply cells and leaving it to the customer to figure out how best to use them. I would imagine cell makers that don't do this are both less able and less interested in providing details on how to best manage a pack. That's not to say this is OK, or that A123 has done a great job in this area either....the resulting confusion over what is OK to do to a cell or not is one of the main reasons I post here. It's bad for EVERYBODY for this kind of mysterious atmosphere to reign. The industry needs to do a better job of getting this info out there and of conveying the subtle complexities of electrochemistry that come into play when designing a control system.

That said, I see nothing in the RFE spec that says it's OK to charge above 3.65 under any circumstances--so why are guys doing it? The specs also show standard discharge at 1C down to 2.0 Volts, which seems to directly contradict the practice of many here who are afraid to ever dip below 2.5 Volts, even under load. Maybe other docs out there conflict with those provided here by lopez, but it seems to me if users had just been a little more conservative in their approaches and obeyed the same limits as RFE does in their own spec, the apparent Li plating issues folks have had could have been avoided. This isn't damage from top balancing, per se. It's damage from top balancing done improperly.

Read more: http://priuschat.com/threads/discussions-about-lifepo4-cells-and-battery-university.119569/page-4#ixzz2XxkfPgSn
Follow us: @PriusChat on Twitter | PriusChat on Facebook

i have a whole page full of papers from the google search and will post up as much as i can when i can read some more tomorrow:

https://www.google.com/#sclient=psy-ab&q=lithium+plating+of+lifepo4+cells&oq=lithium+plating+of+lifepo4+cells&gs_l=hp.3...1493.11258.0.16784.32.32.0.0.0.0.147.3329.15j17.32.0.eqrwrth...0...1.1.19.psy-ab.t1trq6lhnxM&pbx=1&bav=on.2,or.r_cp.r_qf.&bvm=bv.48705608,d.cGE&fp=922e862274675230&biw=1680&bih=935
 
if you wanna understand why lipo is more dangerous than lifepo4 when overcharged this should help. also very good paper:

http://www.electrochem.org/dl/interface/sum/sum12/sum12_p037_044.pdf
 
The problem with an unbalanced pack is that the cells that are NOT out of balance tend to get held at 3.8 to 3.9v by the BMS until the rest of the pack catches up and each cell is held at 3.65v.

With a 20Ah pack the minimal 40ma balancing power of your average crappy BMS is going to take a LONG time holding the high cells at over 3.8v.

But I do have packs that have been left on the charger for over 2 years now (Headway packs) and still hold full AH discharge as they did when new. The BMS holds each 12S 1P pack at 43.8V which means each cell is at 3.65v permanently. Why do I do this? Because leaving them off the charger for days when not used causes the packs to go out of balance which would mean when charging the pack again some of the cells hit above 3.8v whilst the pack is balanced.

Lesser of the two evils.

Maybe I should set the charger to 42.96v which would be 3.58v a cell. Would this compromise the BMS at all?
 
Spacey said:
The problem with an unbalanced pack is that the cells that are NOT out of balance tend to get held at 3.8 to 3.9v by the BMS until the rest of the pack catches up and each cell is held at 3.65v.

With a 20Ah pack the minimal 40ma balancing power of your average crappy BMS is going to take a LONG time holding the high cells at over 3.8v.

But I do have packs that have been left on the charger for over 2 years now (Headway packs) and still hold full AH discharge as they did when new. The BMS holds each 12S 1P pack at 43.8V which means each cell is at 3.65v permanently. Why do I do this? Because leaving them off the charger for days when not used causes the packs to go out of balance which would mean when charging the pack again some of the cells hit above 3.8v whilst the pack is balanced.

Lesser of the two evils.

Maybe I should set the charger to 42.96v which would be 3.58v a cell. Would this compromise the BMS at all?
Click to expand...

If you're floating for a long time like that, I would back off the HVC to 3.6 or just below. What should happen during top balancing is that the MAX cells get held to 3.6-3.65 while the other cells catch up, and yes, this can take hours under normal circumstances. A severely imbalanced large pack can take weeks to recover if depending on the balancers to do all the work. However, if your pack noticeably gets out of balance after a good top-off charge in just a couple of days, you have VERY uneven SOH (state of health) among your cells, or some cells with very high self-discharge. You shouldn't be able to get that far out of whack that fast.

dh
 
this guy did his thesis on aging mechanisms. this is more approachable paper since it is long winded, as are all graduate degree thesis. except mine which was 22 pages only. hehe.

anyway he has a great schematic of the SEI and a relatively simple introduction to how it functions and the how the aging is so charge rate dependent for lifepo4 cells. he discusses how the anode is made larger than the cathode at the beginning of life and as the cell ages how the anode capacity shrinks from plating and other mechanisms but that the loss of anode capacity is dramatically reduced by high charging rates and low temperatures. at the end of life the anode capacity shrinks dramatically and it may be that this is the best time to be limiting the factors such as charge rate and voltage that lead to the anode degradation.

he thanks his girlfriend for encouraging him to "keep up the nerdy work".

http://komar.bitcheese.net/files/JensGroot.pdf
 
torqueon said:
End of charge, that is what your talking about right ? I have found that you can not slam the cells at 10 or 5 amps until they hit 3.65 volts, (headways cells) and call theme charged. Not so according to my crude load testing one cell at at time, tedious as hell ( don't ask ) 80% charged at best. Hold them at 3.65 volt until amps drop to .100 amp as specified and all deliver as expected

I read lot about HVC and LVC, but do not see much about HVC to end of charge ?? I think this is a real cause of imbalance of packs that I read about and experience my self. What do you think.

Wish I could post that bitching E bike! My passion for this is storng!!

Pete
Click to expand...

Yes, you're on the right track I think. If you read the prius chat thread linked by dnmun, you'll see this idea clarified some more. (Funny to see myself quoted here...thanks for the props!)

Here's another tidbit that you won't find there. Perhaps you've heard to "quick DC charging" for EV's. This method is generally only capable of taking a pack to 80% or so, because to avoid damage, this is the point at which charge current has to be sharply curtailed--you can't charge at the same high rate at a high SOC because it will shorten the life of the cell considerably. Hopefully all of this is putting this concept into perspective.

dh
 
The Hyperion 1420 charges like that. You can watch it on a display and see it drop the current so that the cells don't ever go much over 3.6 volts during charge.

Sometimes during balance mode some cells will go over 3.6v for a short time.

You can adjust the voltage down to 3.55v too.

If the cells are well matched the Hyperion will balance a pack in less than 5 mins.

The kids at Hyperion have been playing with A123 cells for awhile now. :)

from 2008

http://www.youtube.com/watch?v=vkDgoXikI_8
 
here is another instructive document from a worthy at SAFT:

http://www.battcon.com/PapersFinal2008/McDowallPaper2008PROOF_9.pdf

from reading about how damaging is the high rate of charge at final voltage when the cell voltage has reached this 3.6V level and the current is still high as it would be with a high current balancing charger then the plating could be significant. he also discusses why the limn2o4 pack show rapid degradation in their capacity.

with the bulk charging using a BMS that i am familiar with the current tapering off dramatically as the cell voltage approaches full charge so that the current during this final phase where balancing is occurring is very small (60mA/20Ah=3E-3 C) in comparison to the current that the balancing chargers can still push at this top voltage.

maybe it would be good to have the balance chargers or the power supply type chargers to be able to taper the current off dramatically too, just like with the bulk chargers in order to reduce this plating effect which i am now totally conscious of. i kinda understood that high rate charging and charging at low temperatures was damaging but did not understand how.

before i started reading this stuff, i thought there was not a residual effect of the higher final voltages, mainly because there was no evident damage. because the shunt balancing current is so low that it may not have as much effect as the higher currents during charging that the balancing chargers can produce.

also i had always assumed that any time the cell dropped to 2V under load it was toast. but it does seem that the cell will rebound to a much higher voltage as the free mobile lithium ions finally make it to the electrode and soaked up in the cathodic matrix. so that was why they did not all die when i over discharged them to 1.67V. it seems the trick is to get the over discharged cell back on the charger immediately.
 
dnmun said:
here is another instructive document from a worthy at SAFT:

http://www.battcon.com/PapersFinal2008/McDowallPaper2008PROOF_9.pdf

from reading about how damaging is the high rate of charge at final voltage when the cell voltage has reached this 3.6V level and the current is still high as it would be with a high current balancing charger then the plating could be significant.

with the bulk charging using a BMS that i am familiar with the current tapering off dramatically as the cell voltage approaches full charge so that the current during this final phase where balancing is occurring is very small (60mA/20Ah=3E-3 C) in comparison to the current that the balancing chargers can still push at this top voltage.

maybe it would be good to have the balance chargers or the power supply type chargers to be able to taper the current off dramatically too, just like with the bulk chargers in order to reduce this plating effect which i am now totally conscious of. i kinda understood that high rate charging and charging at low temperatures was damaging but did not understand how.

before i started reading this stuff, i thought there was not a residual effect of the higher final voltages, mainly because there was no evident damage. because the shunt balancing current is so low that it may not have as much effect as the higher currents during charging that the balancing chargers can produce.

also i had always assumed that any time the cell dropped to 2V under load it was toast. but it does seem that the cell will rebound to a much higher voltage as the free mobile lithium ions finally make it to the electrode and soaked up in the cathodic matrix. so that was why they did not all die when i over discharged them to 1.67V. it seems the trick is to get the over discharged cell back on the charger immediately.
Click to expand...

All rc balancing chargers I have ever seen do cut the current way down when the charge is near the end.

You can also hook a b6 charger to a display and watch this happen.

http://www.hobbyking.com/hobbyking/store/__7348__iMAX_Software_Kit_.html?gclid=CLvVq97ok7gCFWVo7AodxSAAlw
 
The fastest balance methods actually pulse low current at the end of charge, giving the balancers time to catch up between pulses. On a graph, the cell voltages look like sawtooth waves that eventually converge at 100% SOC.

dh
 
You must log in or register to reply here.

Similar threads

M
DIY ligo modular battery with Hoverboard packs
Replies
20
Views
1,648
wturber
wturber
docw009
Lishen LFP 2170EH cell
Replies
4
Views
240
docw009
docw009
H
What’s the best lifepo4 AVAILABLE?
Replies
15
Views
532
Hillhater
H
fill
What unreasonable demands are the US making to 'resolve' tariffs for your country
Replies
11
Views
433
gromike
G
Z
Solar Powered Supercharger
Replies
8
Views
777
harrisonpatm
harrisonpatm
Back
Top