diy solar

diy solar

Top balancing, recommendations (fastest or best practices)

Eric of NordkynDesign said:
Avoid charging the cells individually, or in batches; the whole process would take just as long, but would also result in some fully charged cells lying around for several days.

A point that should at least be considered before doing the single cell method:

A fully charged cell is not going to lose much SOC if sitting around for awhile. What we have are experts claiming top paralleling balancing is the better method and other experts claiming topping of cells individually is the best method. And then we have the step method thrown in. They all have valid points and I will not claim or suggest either way is right or wrong. However, I will follow the instructions I feel most comfortable with and that's charging each cell individually for the reasons I gave above and the link I posted supports it.

The recent bloated cell I believe is an example of this. We have no idea what happened there but my guess is that cell had a much higher state of charge than the rest or possibly a much higher IR. As I recall the parallel balancing went well. Then the pack was fully charged and a capacity test was done. After the test the cells were taken apart and that's when it was discovered one of the cells was severely bloated. It's possible had that cell been individually charged the cell would not have bloated.

I do love these discussions and I am learning so much from them. And I thank everyone for sharing and contributing. (y)
 
We have no idea what happened there but my guess is that cell had a much higher state of charge than the rest or possibly a much higher IR. As I recall the parallel balancing went well.
We may be talking about a different situation or remember the facts differently. What is clear to me from my understanding of physics, is that when in parallel it is not possible for one cell to get to a higher voltage than any of the other cells regardless of its state of charge or its internal resistance. That is the advantage of parallel top balancing.
My guess is that in many of these top balancing fiascos the voltage setting on the power supply was set too high but that was not obvious because the cell voltage is all that some of those power supplies show. Hours later the whole parallel pack went above the recommended voltage and that is when the bloating was observed. Individual charging is just more time consuming but if a mistake occurs with voltage setting it only affects one cell.
 
There isn't any significant advantage to individual top balancing instead of batch. The disadvantages are more chances for user error, and longer time. As mentioned, if the cells are properly paralleled with good connections, they will all have the same voltage.
 
There isn't any significant advantage to individual top balancing instead of batch. The disadvantages are more chances for user error, and longer time. As mentioned, if the cells are properly paralleled with good connections, they will all have the same voltage.

Have you or @Ampster or anyone else read the material from the link I posted?


Please read this before commenting and if you have read it I would like to know your reasons for disputing it. Thanks...:)
 
What we have are experts claiming top paralleling balancing is the better method

Do we?

I've seen the parallel step-balancing method advocated (Nordkyn, MarineHowTo, and apparently the Battery Manufacture Engineer Ghostwriter heard about it from through her work).
And the individual-cell method advocated (GWL)-- (I suggest you look closely at the voltages they are talking about though, its different from our current model and may not be representative).

But I have not found anywhere that explicitly recommends the single step balancing method over the other methods.

edit: in partial answer to my own question, I found an updated (I think) guide from GWL. It seems they now are okay with the parallel balance method to 3.55 or 3.6 vpc. [see here (3 minute video) and here (more extensive lifepo4 guide)] @upnorthandpersonal you should probably take a peak at this too.

However, I will follow the instructions I feel most comfortable with and that's charging each cell individually for the reasons I gave above and the link I posted supports it.

I'm in the same boat, just currently more comfortable with a different method than you. To a degree, for a one time balance, any of the three methods is probably fine. But I do see the logic of both the step-method, and the single cell method, above that of the simple parallel method.

I think both the step-method and the single-cell method are trying to solve the same problem in different ways.

The recent bloated cell I believe is an example of this. We have no idea what happened there but my guess is that cell had a much higher state of charge than the rest or possibly a much higher IR. As I recall the parallel balancing went well. Then the pack was fully charged and a capacity test was done. After the test the cells were taken apart and that's when it was discovered one of the cells was severely bloated. It's possible had that cell been individually charged the cell would not have bloated.

I do love these discussions and I am learning so much from them. And I thank everyone for sharing and contributing. (y)

That case is still way too weird, and way too nebulous for me to draw any conclusions from. There are still a lot of unanswered questions and a few parts of that saga that dont really add up, I've kinda lost faith in getting clear answers to.

But I think I agree with your overall concern, and its a concern I share. In the last 6ish months, I've noticed an uptick in the number of people having issues with their cells (mostly the 280's), mostly during initial balancing + testing. Now it could just be that this is the inevitable result of so many people buying grey market cells at this price point (quality control issues, and the occasional weak or out of spec cell), or it could be that the larger form factor loosely matched cells + low current AliX chargers are leading to a situation where some people are damaging cells during the initial balancing and testing, or it could just be plain and simple user error.

In any case the rise in reported problems during initial pack assembly/testing, is what has renewed my interest/concern in balancing best practices. I think the step method minimizes the chances of user error, and damage due to 'over charging' and minimizes time spent charging in the knees.
 
Last edited:
I read the article. All the graphs are showing 4.0 to 4.2v?

It doesn't mention balancing to only 3.65 volts (way below the 4.2 in that article).

Did i miss something?

I think the instructions are from the olden days of lifepo4 when it was common to charge to higher voltages, it does make me question the relevance to current practices though. Holding for a matter of hours or days @ 4.0 could be very difference than holding for the same length of time at 3.6-3.7
 
Well, I'm definitely not a battery chemist, but I am pretty familiar with electrical theory. I did read the article, and I don't quite follow the premise. It's showing different starting voltages, which should not be possible with all the cells in parallel (as long as they have been given time to equalize). As they charge, the should theoretically all push and pull through the cycle balancing each other out. Electricity will flow from the higher voltage to the lower voltage. It also looked like they were charging at a higher voltage than the top recommended voltage for the cells. Again, I don't know jack about battery chemistry, but that seems like a bad idea regardless. It would speed up the process, but risk over volting the cells.

Curious what the pros have to say.
 
Have you or @Ampster or anyone else read the material from the link I posted?

That document shows voltages above the absolute max in some cases, and it also says:
The bad results of this kind of overcharging may not be visible at first. However after repeated periods of such overcharging, the cell will gradually start loosing its capacity and will have worse performance than the other properly charged cells.

So a few hours at 3.6V isn't going to kill the cells. In fact several drop in MFGs have regular charge voltages at 3.6Vpc, and still get decent cycle life.

If a person is really concerned, they can charge to 3.5V, thats pretty good for balancing, and less harsh on the cell. If parallel balancing is expected to take a long time, and cannot easily be monitored, set the charger to 3.45V. When you have time to monitor, charge to 3.5-3.6V, which shouldn't take very long at all. When it hits the final voltage, disconnect, add a resistor to draw down to 3.4 (shouldn't take long).


Do not ever apply a power supply voltage over the MFGs recommended spec. It won't speed balancing if you have good leads. The internal resistance of the cells is so low, that there isn't any benefit to pushing higher (such as with lead).
 
  • Like
Reactions: Dzl
So a few hours at 3.6V isn't going to kill the cells. In fact several drop in MFGs have regular charge voltages at 3.6Vpc, and still get decent cycle life.

I agree with most of what you said, but to clarify on this particular point. The issue being identified isn't charging to 3.6 or even 3.65 vpc, its forcing the cells to hold at this voltage for an extended period of time. Whether or not that is an issue in the time frames we are talking about is beyond me, but I want to be clear on the issue.

That said, I think your last paragraph is a good approach:
If a person is really concerned, they can charge to 3.5V, thats pretty good for balancing, and less harsh on the cell. If parallel balancing is expected to take a long time, and cannot easily be monitored, set the charger to 3.45V. When you have time to monitor, charge to 3.5-3.6V, which shouldn't take very long at all. When it hits the final voltage, disconnect, add a resistor to draw down to 3.4 (shouldn't take long).
 
The issue being identified isn't charging to 3.6 or even 3.65 vpc, its forcing the cells to hold at this voltage for an extended period of time. Whether or not that is an issue in the time frames we are talking about is beyond me, but I want to be clear on the issue.

Yup...we are not talking about a few hours here but rather several days. My concern is more pronounced if one cell has a higher SOC of charge or a higher IR.

I did look at the graphs in the link I posted. I believe the graphs can apply to current day charging recommendations of 3.65 volts. The graphs are probably from the older days. I should have mentioned that in the first place...lol.
 
Yup...we are not talking about a few hours here but rather several days. My concern is more pronounced if one cell has a higher SOC of charge or a higher IR.

I think the stepped method is one way to minimize the amount of time above ~3.45-3.55. The pack should already be close to 100% by the time you get to the last voltage step, due to the long absorption time in the previous step or two. Or at least that's the impression I get from the instructions.

I did look at the graphs in the link I posted. I believe the graphs can apply to current day charging recommendations of 3.65 volts. The graphs are probably from the older days. I should have mentioned that in the first place...lol.

Just checked the date on the PDF, its from 2012
 
edit: in partial answer to my own question, I found an updated (I think) guide from GWL. It seems they now are okay with the parallel balance method to 3.55 or 3.6 vpc. [see here (3 minute video) and here (more extensive lifepo4 guide)] @upnorthandpersonal you should probably take a peak at this too.

Oops...I just caught your edit. Thanks. I will watch it in a little bit.

Just checked the date on the PDF, its from 2012

Yeah that's old.
 
Yeah, exactly, so if you want to stay in spec then you are probably looking at a lower balance voltage.

My personal application is as a UPS, about 3.395v per cell, which seems safe, but I suspect even that exceeds the battery specs.
 
I read the article and it hasn't changed my view. It did shed some light on the theory of using your BMS to balance a pack. The point there was that it meant many cells would spent time at 3.65 being shunted. However with an adjustable BMS that could turn on balancing at 3.4 volts the issue is mute.
It certainly didn't change my mind about my preference for parallel top balancing as a one time ocurrance. I am not sure I saw any argument against parallel top balancing. I do not distinguish between the two forms of parallel top balancing. One being in one step or the other being in two or more steps. The only difference might be a few hours longer at 3.60 to 365 volts but in the long run I think it is insignificant.
I think individual top balancing is more time consuming and adds more chance for error.
 
I am not sure I saw any argument against parallel top balancing.

The body of the document doesn't address it directly, but if you look at the tips on the second to last page, it is explicitly stated.

1) The cells should be balanced to full one by one before the assembly of the battery pack. This way the will have the same capacity and they will need only very little balancing.
 
It is inaccurate because there is no guaranty that they will have the same capacity. They will have the same voltage. Physics tells us that cells in parallel if given enough current or time will end up at the same voltage.

That is fine if someone wants to follow that one document which is printed by a manufacturer of acessories. I disagree with the conclusion that physics doesn't work and the only way to get them to the same voltage is to individually charge them one by one.
NOTE: Susequently a later version from GWL apparently clarifies that parallel top balancing os okay.
 
Last edited:
It is inaccurate because there is no guaranty that they will have the same capacity.

That's true and there is no guaranty parallel balancing cells will have the same capacity. That's a known. I agree.

That is fine if someone wants to follow that one document which is printed by a manufacturer of acessories.

That someone must be Gazoo,,,,lol.

I disagree with the conclusion that physics doesn't work and the only way to get them to the same voltage is to individually charge them one by one.

While I agree with you I am not so sure that the author of the document would disagree with you. I think the concern was holding one of the cells at the full SOC for an extended period of time while the others catch up.

NOTE: Susequently a later version from GWL apparently clarifies that parallel top balancing os okay.

Yes I noticed. I watched the video.

So my concern has not been addressed. My concern is more pronounced if one cell has a higher SOC of charge or a higher IR. Let's leave out the high IR for now and stick with SOC.

Let's assume in a 16 cell pack one of the cells has a 80% SOC and all the other ones have a 50% SOC. The one with 80% SOC is obviously going to reach 3.65 volts before the other ones and that cell is going to have 3.65 volts applied to it for several days during the balancing process. Is it possible that cell could be damaged from having 3.65 volts applied for several days? That's basically all I am asking.

I have become convinced if parallel balancing using the step method would be the safest way. Agree...disagree? I know someone mentioned it as a way of preventing overcharge due to user error and that's the reason for the step method. Towards the end of the charge is when one really needs to keep a close eye on the voltage until done.
 
Let's assume in a 16 cell pack one of the cells has a 80% SOC and all the other ones have a 50% SOC. The one with 80% SOC is obviously going to reach 3.65 volts before the other ones and that cell is going to have 3.65 volts applied to it for several days during the balancing process. Is it possible that cell could be damaged from having 3.65 volts applied for several days? That's basically all I am asking

You mean while parallel balancing, or using the BMS balancer when the pack is already assembled in its 16S configuration?

If it's the former (parallel balancing), it can't be at 3.65v while the others are at a different voltage. They'll all be at the same voltage.

If it's the latter, then, yeah, I don't think that's good - basically super mismatched cells, and you're trying to use a puny balancer.

Better to just use partial capacity for the several days/weeks.
 
No I meant while parallel balancing.

I should have said the cell will be at full capacity for several days during the parallel balancing process, but will still have 3.65 volts applied to it.

In other words is it possible to damage a cell that is charged to full capacity while keeping 3.65 volts applied to it. I understand if done with a charger the charger will cut off. But a power supply will not cut off and will hold the voltage at full capacity. I thought I read somewhere on this forum it could be a problem. But that was awhile ago and I can't find it.
 
Back
Top