[dtbaker]

I am attempting to re-balance a pack after a 'low-voltage event' which totally killed one cell in the pack. I have bypassed the completely dead cell, and am attempting manual re-balance before using pack charger.

I have been using a 'revolectrix powerlab' 1000watt single cell charger with luck in the past to bring cells up to a set voltage, and terminate charge automatically... which got the cells all pretty close initially, needing just a little fine tuning with a resistive element when the cells were new.

HOWEVER, this time around, there are a number of cells that just don't seem to hold the charge nearly as well, and even with repeat charges to the set voltage, they 'settle' to a lower voltage than other cells... and when I try a pack charge the cells are obviously out of balance with a number of cells heading up to 4v while a few don't seem to keep up.

I suspect that some cells are damaged internally, perhaps having vastly higher resistance, or 'dendrites', or plating or something....

My question to the group is:
when a number of cells become 'unbalanced' with others in the pack, and do not remain balanced after repeat single cell charging/discharging to a set voltage... is there any way to recover them, or are they dead?

 

[john61ct]

If the "low voltage event" was dead flat for the whole pack, they're likely not worth messing with further.

For large expensive banks, multiple layers of protective electronics are worth investing in to prevent such disasters.

With LFP, 2.95V per cell is my drop dead cutoff, with plenty of alarms and load cutoffs before then.

 

[MattsAwesomeStuff]

Tangent: Not sure why you created 2 threads 2 minutes apart for the same issue..

when a number of cells become 'unbalanced' with others in the pack, and do not remain balanced after repeat single cell charging/discharging to a set voltage... is there any way to recover them, or are they dead?

You are using the word "unbalanced" but that is not really what "balance" means. Yeah, technically they are also out of balance, but that's not the problem.

"Balance" refers to have equal states of charge from, presumably identical cells.

The problem you have is that your cells are no longer capacity-matched, due to damage, and thus will always become quickly out of balance when charged or discharged in series.

It's like lining up an adult with a toddler and telling them to each take 3 steps forward. You notice the toddler is not keeping up. So you line them back up again, and again after 3 steps the toddler is not keeping up. The problem isn't that you're not lining them up good enough, the problem is that your baby is ugly. Err, that your toddler has shorter legs.

here are a number of cells that just don't seem to hold the charge nearly as well, and even with repeat charges to the set voltage, they 'settle' to a lower voltage than other cells... and when I try a pack charge the cells are obviously out of balance with a number of cells heading up to 4v while a few don't seem to keep up.

Hard to parse your grammar here, but, I think this could be a couple different things.

1 - Some cells have lower capacity (due to damage). That means they fill quicker and empty quicker because there's less capacity in them.

2 - Some cells have higher internal resistance (due to damage). This means that they fill slower and discharge slower, and get hotter when they do. When charging, you may be giving it 3.65 volts, and it may be showing 3.65 volts, but it's not adding all that as energy. Chargers usually "terminate" when the charging current drops to a certain amount, so if your resistance is high the charging current will be low especially at the top end, and it will stop before it was done filling.

3 - Some cells may have dendrite damage internally, small shorts that constantly drain the battery. So they self-discharge.

In the case of the 3rd, they're probably garbage, in the case of the 2nd your charger isn't likely to use them effectively and you'll have to rig up some halfassed solution. In the case of the 1st you could add some capacity to the cell, maybe purchase some 18650 LiFes of approximately the missing capacity, hook them up in parallel, and hope you're closer to matched capacity that way.

 

[dtbaker]

1 - Some cells have lower capacity (due to damage). That means they fill quicker and empty quicker because there's less capacity in them.

2 - Some cells have higher internal resistance (due to damage). This means that they fill slower and discharge slower, and get hotter when they do. When charging, you may be giving it 3.65 volts, and it may be showing 3.65 volts, but it's not adding all that as energy. Chargers usually "terminate" when the charging current drops to a certain amount, so if your resistance is high the charging current will be low especially at the top end, and it will stop before it was done filling.

3 - Some cells may have dendrite damage internally, small shorts that constantly drain the battery. So they self-discharge.

I like your explanations, and differing internal resistance may well explain the behavior I am seeing while I am attempting a manual top-balance before reverting to normal use and pack charging. I have brought all cells up to a 'final voltage' cell by cell, but they just don't seem to stay balanced the way they did 18,000 miles ago.

What I am looking for now is a way to identify each cell's IR, and some guideline on how much the IR can vary cell to cell to 'stay in balance' with repeat charge cycles in series. I'd like to limp along for a while if there are just a few cells out of spec, or consider replacing just a few cells rather than the entire pack.

what equipment would I need to establish a cell IR?
method for measuring?
acceptable range of difference to minimize 'drift' with repeat charge cycles?

 

[MattsAwesomeStuff]

What I am looking for now is a way to identify each cell's IR

The bad news is that, generally, high internal resistance goes hand in hand with lowered capacity.

I think you just have lowered capacity, and nothing can help that other than boosting the low cells with some smaller add-on cells like 18650s.

I'm not sure you grasp what's going on, you seem to keep using the same terms for a variety of things. Isolate what you're measuring.

Charge all your cells up to their max voltage independently. Then let them sit, then measure them. If you have self-discharge then that's a separate issue.

If they maintain their voltage, but some cells drop faster under use, then those cells are lower capacity, "balancing" is not going to fix them.

 

[steveob]

internal resistance for batteries, usually you test it under different loads, and the change in voltage divided by the change in current is the battery resistance. Though it moves around a bit due to temp and state of charge, and you want to be quick in taking your measurements, and take them on the flat part of the discharge curve.

capacity, just discharge it fully and measure how many amp hours it takes to charge it fully (or charge it fully and see how many amp hour you get while discharging it fully). A bit time consuming, but that is how it is done.

 

[dtbaker]

internal resistance for batteries, usually you test it under different loads, and the change in voltage divided by the change in current is the battery resistance. Though it moves around a bit due to temp and state of charge, and you want to be quick in taking your measurements, and take them on the flat part of the discharge curve.

capacity, just discharge it fully and measure how many amp hours it takes to charge it fully (or charge it fully and see how many amp hour you get while discharging it fully). A bit time consuming, but that is how it is done.

thanks Steve,that was exactly the type of answer I was looking for.... How to quantify the state of the individual cells so I can decide how many are 'bad' so I can decide whether the replace them.

It sounds like measuring internal resistance may be difficult with typical garage equipment; not having a precise way to apply different loads and measure small difference in current. I may have to resort to comparing voltages after cell charge, after settling overnight, and after use/recharge cycles.

 

[dtbaker]

The bad news is that, generally, high internal resistance goes hand in hand with lowered capacity.

I think you just have lowered capacity, and nothing can help that other than boosting the low cells with some smaller add-on cells like 18650s.

I am not as concerned with reduced capacity as I am with inability to maintain a decent top-balance with repeat charge and use cycles. Unbalanced end of charge voltages obviously will damage cells further.

I'm not sure you grasp what's going on, you seem to keep using the same terms for a variety of things. Isolate what you're measuring.

I think I grasp things quite well thank you. Sorry my language is too imprecise for you.

Charge all your cells up to their max voltage independently. Then let them sit, then measure them. If you have self-discharge then that's a separate issue.

obviously.... but I am not seeing self discharge, nor is that what I asked about.

If they maintain their voltage, but some cells drop faster under use, then those cells are lower capacity, "balancing" is not going to fix them.

pretty hard to tell if some cells are dropping faster under use since the middle of the curve is so flat. i.e. at 10%-70% DOD there is hardly any difference in voltage.

 

[steveob]

I was thinking you just use a decent vom (I like my hobby grade vc97+) and a cheap 100 amp shunt (you can make your own shunt too), and a couple loops of wire in a bucket of water, that draw about 50 amps a piece at 3.7v.

It is totally garage doable though, been done many times, but you have to know what you are doing. You don't want to be measuring float or recovery voltages, just load 1, note shunt and battery voltage, add load 2 in quick succession, note shunt and battery voltage, and shut it down, then sort out the shunt and battery voltages to figure out resistance.

 

[MattsAwesomeStuff]

I am not as concerned with reduced capacity as I am with inability to maintain a decent top-balance with repeat charge and use cycles. Unbalanced end of charge voltages obviously will damage cells further.

Okay, so, I think you still don't get it.

"Unbalanced" is not the issue. Reduced capacity of some cells is the issue. You keep calling it a balancing issue as if it can be resolved by better balancing, but I'm trying to tell you that, from your symptoms, it can't.

I think I grasp things quite well thank you. Sorry my language is too imprecise for you.

I don't matter. I'm trying to articulate that you are mixing terms and problems and the advice you're given based on what you're asking might not be useful to you.

obviously.... but I am not seeing self discharge, nor is that what I asked about.

Well, I'm glad I clarified, because it certainly did seem that that was what you were asking about.

Ex) "HOWEVER, this time around, there are a number of cells that just don't seem to hold the charge nearly as well, and even with repeat charges to the set voltage, they 'settle' to a lower voltage than other cells."

When you say they "settle to a lower voltage" it makes me think you're charging them all to 3.7v but then you come back later and some have fallen to 3.5v or whatnot.

That's exactly why I'm trying to point out that it's important to accurately describe your issues, so that when people try to help diagnose them for you they can give you useful answers.

pretty hard to tell if some cells are dropping faster under use since the middle of the curve is so flat. i.e. at 10%-70% DOD there is hardly any difference in voltage.

Okay, here is a procedure:

1 - Charge all cells in series until one of them reaches your max voltage (3.65 or 3.7v or whatever). This is for convenience.

2 - Charge all the remaining cells up to their max voltage independently.

3 - When all cells have the same max voltage (not just the sum of voltages being what you guess to be correct), discharge them in series. Set a timer if you feel like it.

4 - When one cell falls below it's minimum voltage, stop the test. That cell is the weakest cell. You can yank it out of the series chain, then continue the discharge.

5 - When the next cell falls below it's minimum voltage, stop the test. That cell is the next weakest cell. You can remove it from the chain too.

6 - Do this as long as you like, recording times as you go. This will give you a ballpark for how much life the cells have. If your first cells hits your minimum voltage and the others are all significantly higher, you know it's lost a lot of capacity. If a few others are close behind, you know those are all weak. If they all arrive fairly close to each other, your pack is decently capacity-matched.

 

[dtbaker]

6 - Do this as long as you like, recording times as you go. This will give you a ballpark for how much life the cells have. If your first cells hits your minimum voltage and the others are all significantly higher, you know it's lost a lot of capacity. If a few others are close behind, you know those are all weak. If they all arrive fairly close to each other, your pack is decently capacity-matched.

Two problems with capacity testing and matching...

1. time. I have 48x100ah CALB LiFePO4 cells; draining each to 80% DOD and recording the time would take hours per cell unless I come up with a super heavy resistance. And then take another 4 or 5 hours per cell to charge back up with my Revolectrix at 20amps.

2. results and correction. Let's say I go thru all 48 cells (taking 4 -5 hours per cell), and find several that are 10% less capacity than the others. Then I'm faced with the dilemma whether to just bypass them and adjust my charger end-of-charge voltage for a smaller pack.... or, if too many cells are short, trying to mix in some new cells.... or junking the entire pack even though most of the cells would probably be serviceable for a while longer.

This is why I keep coming back to end-of-charge top balancing... to see what ideas anybody has on how to make the best of this pack without the expense of replacing the whole thing. I'm looking for some way to quantify which cells will not be 'top-balancable' anymore, (regardless of what type of internal damage with higher IR or reduced capacity) so I can decide whether I can just bypass a couple, replace a few, or have to replace the whole pack.

My goal is to bypass/replace as few as possible, adjusting charger voltage if need be, so that no cells exceed max voltage at end of charge... and drive the pack as many more miles I can without junking the pack.

 

[MattsAwesomeStuff]

1. time. I have 48x100ah CALB LiFePO4 cells; draining each to 80% DOD and recording the time would take hours per cell unless I come up with a super heavy resistance. And then take another 4 or 5 hours per cell to charge back up with my Revolectrix at 20amps.

No. You're not following the instructions. If you read the procedure I gave you, you will notice that I said to discharge them IN SERIES so you will be discharging them all at the same time.

I specifically wrote a foolproof method for you that saves you as much time as possible.

This is why I keep coming back to end-of-charge top balancing... to see what ideas anybody has on how to make the best of this pack without the expense of replacing the whole thing. I'm looking for some way to quantify which cells will not be 'top-balancable' anymore,

Again...

You keep calling a capacity issue a "balancing" issue.

Balancing will never increase the capacity of a cell. That is not how it works.

You can keep asking but the answer will be the same every time. Balancing does not fix capacity issues. No one will have any "ideas" on how top balancing will fix your capacity issues, you might as well be asking for ideas on which seat position changes the color of your paint job.

The only way to figure out your capacity is to check your capacity.

This will involve charging them and then discharging them.

If you do not have time to charge and then discharge your batteries, and to make occasional measurements as to the voltage along the way, you will never know their capacity or identify bad cells.

You can rephrase the question over and over about how balancing them will tell you this info, but you will never get an answer.

Balancing just ensures that your capacity measuring is accurate. So, yeah, top balance them first, cell-by-cell. Then discharge them and occasionally measure each cell until you have an idea which ones to keep an eye on and roughly when the pack will be bottoming out. That will identify which cells have less capacity.

 

[dtbaker]

No. You're not following the instructions. If you read the procedure I gave you, you will notice that I said to discharge them IN SERIES so you will be discharging them all at the same time.

The only load heavy enough to discharge the series is.... driving. I totally GET this, believe me I am not an idiot. I have been driving EVs for 10 years now, and what I am facing is getting some more miles out of a pack that is showing signs of decay.

I can and have driven around until the pack voltage starts dropping, note the amp-hrs and miles and have a pretty good idea where my floor is for the pack. I try to live within that capacity, and I KNOW there is no way to recover capacity.

I totally understand that balancing will not increase capacity. That's not what I am worried about. What I am worried about is that it appears *some* cells are not STAYING balanced and returning to end-of-charge at the same time after a couple charge/drive cycles... forcing some cells higher than I would prefer before the pack charger finishes. i.e. 'drifting' end of charge voltage is what I have been asking about.

What I was looking for was a way to quantify and test cells to be predictive about which and how many will drift unacceptably after a few charge cycles so I could decide if it would be cost effective to mix in a few new cells, or replace the whole pack.

I think we're going to have to accept that the question I'm asking is not the one you are answering, and leave it at that.

 

[MattsAwesomeStuff]

The only load heavy enough to discharge the series is.... driving

So...

1 - Bulk charge your battery pack.

2 - Top balance each cell.

3 - Go for a drive based on what you already know about your likely range, try to time it to be near the end.

4 - When you get back, check your voltages and see which cells are significantly lower than the rest of the pack. Those are the shitty cells.

5 - Remove/replace/supplement the shitty cells.

I totally understand that balancing will not increase capacity. That's not what I am worried about. What I am worried about is that it appears *some* cells are not STAYING balanced and returning to end-of-charge at the same time after a couple charge/drive cycles

...

I'm trying to narrow in on why you're not getting this.

If you have cells that are not capacity-matched, they will immediately become unbalanced when used.

If you have 5 LiFePO4 cells charged to 3.65v each.. but one has lost 50% of it's capacity, its voltage is going to plummet twice as fast as the cells of double capacity.

At the same time, for the capacity it does have, from the (reduced) surface area of the plates, it's drawing twice the current. So, for example if they are 100AH cells and you're drawing 100 amps... that's 1C. If one cell is reduced to 50AH max, you're still drawing 100 amps from it, that's 2C. So that chemistry is going through double the abuse, double the waste energy, etc.

Ditto on charging, you'll be charging it at double the C-rate of its neighbors, double the waste.

If you waste energy as it's coming out or into the battery, it will have less capacity compared to the other cells.

So that could be one explanation.

Another explanation can be that, if you're draining the pack until it's total voltage is low, then the weak cells are going to be way beyond their safe bottom limit.

For example, 10 cells:

If spec is each cell to a max of 3.65 volts...

3.65 + 3.65 + 3.65 + 3.65 + 3.65 + 3.65 + 3.65 + 3.65 + 3.65 + 3.65 = 36.5 volts.

If you want to cut off at 2.5 volts per cell....

2.5 + 2.5 + 2.5 + 2.5 + 2.5 + 2.5 + 2.5 + 2.5 + 2.5 + 2.5 = 25v.

But if you have no knowledge of the cell-by-cell voltage, only the pack total, what will actually be happening is probably:

2.77 + 2.77 + 2.77 + 2.77 + 0.00v + 2.77 + 2.77 + 2.77 + 2.77 + 2.77 = 25v.

Which means 9 cells not empty, and 1 cell completely blasted to zero, further damaging it every cycle.

So when you go to recharge it, there's less and less and less capacity in that cell every time. Again, not a balance issue, a capacity issue.

In any case, you cannot know these things until you observe and measure them. There is no further shortcut. You have to do what it takes to figure this out. You can't make good decisions with bad information.

 

[john61ct]

I was under the impression that, at least with LFP, even one "draw down to dead flat 0.0V" event

will result in instant death, time to scrap that cell.

Is that not the case?

 

[Duncan]

I was under the impression that, at least with LFP, even one "draw down to dead flat 0.0V" event

will result in instant death, time to scrap that cell.

Is that not the case?

Hi John
I believe that a cell can survive going down quite low - the instant death bit is when the other cells in its string then make it reverse polarity

That is NOT to say a cell will survive going down low!


Reading this thread it seems to me that you guys are talking across each other

DTbaker seems to have NOT a cell imbalance problem but a cell self discharge problem

If he starts with a balanced pack some of the cells slowly diverge from the others

If that simply happens with time then those cells are self discharging

 

[MattsAwesomeStuff]

I was under the impression that, at least with LFP, even one "draw down to dead flat 0.0V" event will result in instant death , time to scrap that cell.Is that not the case?

Yes, sometimes, reverse charging is worse, leaving it that way for a while is worst. But you got distracted by a tangent and are missing the forest through the trees.

I was just throwing numbers at it as simply as possible. My point wasn't "It reaches exactly zero". My point was "One will be low enough to incur damage while the total voltage is still at a level that would fool you to thinking it was okay."

Substitute 0.00v for any number below 2.0v or whatever you want. Maybe it reaches 1.00v, or 1.5v, or whatever. The actual number doesn't matter, I picked something extreme to try to simplify it and illustrate the point.

...

DTbaker seems to have NOT a cell imbalance problem but a cell self discharge problem. If he starts with a balanced pack some of the cells slowly diverge from the others. If that simply happens with time then those cells are self discharging

That's what he was saying earlier, so that's what I was explaining at first, but he later clarified:

"obviously.... but I am not seeing self discharge, nor is that what I asked about"

He is not talking about self-discharge in any way.

He has capacity issues he is calling "balancing" issues, and doesn't want to take the time to put them through even a single charge/discharge cycle to discover any information that would help fix his problem or even find out what it is.

 

[onegreenev]

I like your explanations, and differing internal resistance may well explain the behavior I am seeing while I am attempting a manual top-balance before reverting to normal use and pack charging. I have brought all cells up to a 'final voltage' cell by cell, but they just don't seem to stay balanced the way they did 18,000 miles ago.

What I am looking for now is a way to identify each cell's IR, and some guideline on how much the IR can vary cell to cell to 'stay in balance' with repeat charge cycles in series. I'd like to limp along for a while if there are just a few cells out of spec, or consider replacing just a few cells rather than the entire pack.

what equipment would I need to establish a cell IR?
method for measuring?
acceptable range of difference to minimize 'drift' with repeat charge cycles?

I would do a full discharge to 2.3 volts each on every cell in the pack resting voltage. Then do a charge and watch your pack as it charges and record your voltages and when the first one comes to your max something like 3.8 volts then that is the one with the least capacity. Check the others and see where those voltages are sitting. This will allow you to pull out the low capacity cells and replace them. This is why I always bottom balance. That way if you end up with an extreme discharge the cell are all at the same voltage. Bottom balancing will allow you to use your true capacity from your lowest capacity cell. The rest of the cells will just have to go along for the ride. If a cell is really low capacity you should replace it with ones that are more like the rest in capacity. You can only use the lowest capacity anyway. So if you top balance and drive your vehicle extreme discharge you risk that low imbalance and causing a cell to go empty before the others and kill a cell or two and even damage surrounding cells. Ask me how I know.

My John Deere Mower is now using Leaf Modules. I bottom balanced them and they are very nicely balanced cells. I charge to a lower than max top voltage and stop before I go to low. That will keep my pack equal. If I need to pull off 12 volts I will do so from a DC DC converter that utilizes the full pack voltage and will not imbalance a cell or group of cells. So when my pack goes low they all go low and will never have a reversal event. Monitor the cells. Yes. Top balance. Nope. Bottom only.



Use the revoltx to fully discharge to 2.3 volts constant current constant voltage just like you do when you charge them. That way you know they are all starting out the same state of charge. Low capacity cells will fill first. So if your low cell goes to to 3.8 real early while all the others are still like 3.2 volts then that cell should be removed.

 

[evric]

Hi Dan,
Talking about testing for internal resistance, one of the best ways is to use the CellPro Powerlab 8 with measures the Internal Resistance for you. This unit will tell you the Capacity and IR of each cell as it cycle the cells.
Get one or borrow one to check your cells.
I have one, but I am in South Australia.
Eric.

 

[dtbaker]

Hi Dan,
Talking about testing for internal resistance, one of the best ways is to use the CellPro Powerlab 8 with measures the Internal Resistance for you. This unit will tell you the Capacity and IR of each cell as it cycle the cells.
Get one or borrow one to check your cells.
I have one, but I am in South Australia.
Eric.

I have a powerlab 6.... and have used it with a custom 'charge-only' setting for initial top-balance to bring cells up to 3.66vpc, and stop.

Thanks for the reminder that is has a discharge-charge cycle, which should give me capacity and IR, I have not read thru how to set that up. I was trying to avoid the many hours per cell it would take to drain and recharge.

At this point, I am going to try to squeeze out some more miles and attempt to deal with the 'unmatched' cells. I have bypassed 2 cells that were really bad, and have my eye on one more that shows much less capacity than others. For now, I have adjusted my end-of-charge voltage down on the pack charger, and will keep an eye on the cells that tend to 'drift' higher and higher at end of charge.

When the manual periodic top-balancing gets to be too much of a pain, I'll have to spring for a new pack, or at least most of one after I do a capacity test on the old ones to determine which are worth keeping.