Top Balancing "How to"

[Just John]

Cells from the same batch and MFG will often be within a few percent of each other. However I have seen cells that are 10-20% different in SOC, and some folks have reported receiving cells at 90% and 50% in the same order. That is not something that passive balancing will remedy. Maybe if you only charge to say 13.7-13.8V, and don't mind only using 50% of the packs capacity?

Its terribly unusual for B grade cells which fail the highest batch group testing to have the test suspended when equipment is tied up, and they are sold as is to the grey market resellers.

I can tell you from personal experience, they do ship at different state of charge. Always verify.

I had multiple cells arrive above 95% SOC, and others at 50%.

I also think that sometimes the balance is counterproductive on a BMS, but don't have enough experience to say conclusively.
I know the Daly BMS does 35 milliamps of resistive balancing, and then doesn't do it continuously. It could easily take a year to balance cells like that (or even more, depends on how often it gets above 3.4v where I set the balance to start).

A top balance is just a sure way to get the SOC in cells quickly in sync. You are welcome to do as you wish with your cells, but I recommend it.

 

[noenegdod]

Cells from the same batch and MFG will often be within a few percent of each other. However I have seen cells that are 10-20% different in SOC, and some folks have reported receiving cells at 90% and 50% in the same order. That is not something that passive balancing will remedy. Maybe if you only charge to say 13.7-13.8V, and don't mind only using 50% of the packs capacity?

Its terribly unusual for B grade cells which fail the highest batch group testing to have the test suspended when equipment is tied up, and they are sold as is to the grey market resellers.

I just realized I misread that last post I quoted. I am going to parallel charge them to about an 80% SOC and call that my top balance. I dont think I am going to take them all the way up to 3.650 but hold them for a day or so at 3.500 until they stop accepting any current and call that done. Then let any balancing from the BMS do its thing. I dont think that just throwing cells together into a battery and letting the BMS look after it from an unknown SOC is a productive idea.

 

[Hedges]

I just realized I misread that last post I quoted. I am going to parallel charge them to about an 80% SOC and call that my top balance. I dont think I am going to take them all the way up to 3.650 but hold them for a day or so at 3.500 until they stop accepting any current and call that done. Then let any balancing from the BMS do its thing. I dont think that just throwing cells together into a battery and letting the BMS look after it from an unknown SOC is a productive idea.

The curves I see indicate that 80% isn't far enough up the curve to achieve much balancing. Looks to me like 95% would be a much steeper part of the curve.

LiFePO4 balancing / voltage limiting

So I just replaced my old AGM bank with a bank of 16 180ah CALB cells. I bottom balanced them; paralleled them all and brought them down to 2.70 volts.

forum.solar-electric.com

I could see charging them as a series pack with BMS attached until it disconnects, then using a CC/CV power supply set for 3.65V to top off each cell individually. But if cells do start out at widely varying SoC, that will take a while. (no longer than parallel charging except requiring more frequent intervention to move to next cell.)

 

[John Frum]

I could see charging them as a series pack with BMS attached until it disconnects, then using a CC/CV power supply set for 3.65V to top off each cell individually. But if cells do start out at widely varying SoC, that will take a while. (no longer than parallel charging except requiring more frequent intervention to move to next cell.)

Agree but...
The big blue cells that are popular here are usually close in capacity and charged to 3.29ish volts.
I estimate that 280ah cells would range from 272-280ah.
Even if the cells come bottom balanced(which I think they effectively are), the cells will be 8ah difference max typically.

 

[noenegdod]

The curves I see indicate that 80% isn't far enough up the curve to achieve much balancing. Looks to me like 95% would be a much steeper part of the curve.

Just so I have a clearer understanding:

Is this because of individual cell differences? 3.375 could be 80% in one cell but could be 70% in another?

 

[Luthj]

The voltage indicates the relative density of lithium ions on each electrode. So a 100AH battery at 3.375V will have half the capacity of a 200AH battery at 3.375V, all else equal. In theory the same would apply between a 280 and 272AH cell. During factory testing the cells are discharged to the cutoff, then recharged to a storage SOC level. This is effectively bottom balancing. Though how accurate it is, I cannot say. Since most of us operate in the top regime of the SOC curve, the difference in total cell capacity, plus and difference in SOC, will be the amount that needs to be balanced between the cells.

So if 2 cells vary by 5AH in capacity, and they are at the same SOC (lets say 50%). Then the top referenced imbalance would be 2.5AH. However if the equipment at the factory counts AH in vs a simple voltage cutoff, then the top reference imbalance would be 5AH.

 

[Hedges]

Just so I have a clearer understanding:

Is this because of individual cell differences? 3.375 could be 80% in one cell but could be 70% in another?

Yes. Difference between 75% and 85% on the graph part way down the page I linked looks like about 0.025V
But 0.025V is the difference between 94% and 95%, 10x the sensitivity.

When people top balance they sometimes report significant voltage drop across the busbar, which leaves some cells lagging behind.
Charging to a higher voltage means the cells start to shoot up in SoC, but voltage is held back until other cells are close behind.

So good electrical contact, connecting power supply to opposite ends of positive/negative bus (and maybe connecting multiple places with multiple wires for large banks or breaking into multiple 12V banks with matched wiring) and charging to a higher voltage would all get the cells to closer SoC condition.

 

[noenegdod]

The voltage indicates the relative density of lithium ions on each electrode. So a 100AH battery at 3.375V will have half the capacity of a 200AH battery at 3.375V, all else equal. In theory the same would apply between a 280 and 272AH cell. During factory testing the cells are discharged to the cutoff, then recharged to a storage SOC level. This is effectively bottom balancing. Though how accurate it is, I cannot say. Since most of us operate in the top regime of the SOC curve, the difference in total cell capacity, plus and difference in SOC, will be the amount that needs to be balanced between the cells.

So if 2 cells vary by 5AH in capacity, and they are at the same SOC (lets say 50%). Then the top referenced imbalance would be 2.5AH. However if the equipment at the factory counts AH in vs a simple voltage cutoff, then the top reference imbalance would be 5AH.

I understand that. That is pretty simple. I was just trying to get my head around why you could not leave 4 cells sitting at 3.375 until they were not drawing any more current and call them balanced. Hedges dumbed it down enough for me

One more question though. If you paralleled all the cells for a month, two months, 6 months at 3.295 (or some other "storage" voltage), which all my cells were delivered at, would they be balanced at some point or is there something else at play that I am ignorant to?

 

[Hedges]

People say that cells sitting wired in parallel does much less than charging in parallel. Maybe because that requires one to discharge current (pulling cell down in voltage) in order to charge another (pulling that up in voltage.)

 

[Luthj]

One more question though. If you paralleled all the cells for a month, two months, 6 months at 3.295 (or some other "storage" voltage), which all my cells were delivered at, would they be balanced at some point or is there something else at play that I am ignorant to?

There is a minimum potential voltage required to move the ions. That's around 10mv above resting. This does vary with SOC and temperature. So unfortunately passive balancing with parallel cells like that won't really work on any reasonable time scale. Though its possible self discharge could get them within 10% (over a year?), but that's conjecture.

You can test this yourself by parallel cells and measuring the current. When I did this with some 280AH cells which were 1-2 mv different, I didn't measure any current on my 1mA resolution multimeter. That 2mv ended up being 30Ah between the highest/lowest cell when I eventually balanced them. I series charged them and used some resistors to bleed off the high cells. It ended up taking 12 hours of constant attention. I should have just used my bench supply to top off each cell to 3.55-3.6V before assembly.

 

[noenegdod]

There is a minimum potential voltage required to move the ions. That's around 10mv above resting. This does vary with SOC and temperature. So unfortunately passive balancing with parallel cells like that won't really work on any reasonable time scale. Though its possible self discharge could get them within 10% (over a year?), but that's conjecture.

You can test this yourself by parallel cells and measuring the current. When I did this with some 280AH cells which were 1-2 mv different, I didn't measure any current on my 1mA resolution multimeter. That 2mv ended up being 30Ah between the highest/lowest cell when I eventually balanced them.

There is the missing link. Thank you! Everything makes sense now.

A little smarter every day thanks to you all

 

[ArthurEld]

To top balance with a balancer you have to go into the knee until one cell almost hits 3.65V Then set the charge controller to float a little below that until the cells are balanced. Repeat until all cells reach 3.65.

edit: It's best not to float for longer then necessary above 3.4V So, I let it float at 3.4V until it is balanced to within .003V. At that point the cells are very close to fully charged.

My voltages are for cells. I am talking about serial so multiply by # of cells.

 

[Just John]

There is a minimum potential voltage required to move the ions. That's around 10mv above resting. This does vary with SOC and temperature. So unfortunately passive balancing with parallel cells like that won't really work on any reasonable time scale. Though its possible self discharge could get them within 10% (over a year?), but that's conjecture.

You can test this yourself by parallel cells and measuring the current. When I did this with some 280AH cells which were 1-2 mv different, I didn't measure any current on my 1mA resolution multimeter. That 2mv ended up being 30Ah between the highest/lowest cell when I eventually balanced them. I series charged them and used some resistors to bleed off the high cells. It ended up taking 12 hours of constant attention. I should have just used my bench supply to top off each cell to 3.55-3.6V before assembly.

This is correct. I speak from experience.
Now, probably somewhere between 3.40 and 3.45 would work as well, but you really need them above 3.4v for it to work as a balance. From 3.4 to 3.65 is at most 5 amp hours, might as well spend the extra hour doing it to 3.65v.

 

[benja.min47]

Hi,
I just got 4 lifepo4 100ah cells. The 4 tested at 2.99v. For the moment they are connected in parallel, at rest. I would like to do the first parallel charge at 3.65v as will in his video. But I only have that as material:


Can I do it?

Thanks

 

[John Frum]

Hi,
I just got 4 lifepo4 100ah cells. The 4 tested at 2.99v. For the moment they are connected in parallel, at rest. I would like to do the first parallel charge at 3.65v as will in his video. But I only have that as material:
View attachment 39938

Can I do it?

Thanks

Yes but it will be very slow.
Do you have a bms and a "12 volt" charger?

 

[benja.min47]

Yes but it will be very slow.
Do you have a bms and a "12 volt" charger?

Yes I have a BMS and a 12v charger. As recommended, I wanted to balance before final assembly. Then I will charge with the 12v charger under the control of the BMS!

 

[John Frum]

Yes I have a BMS and a 12v charger. As recommended, I wanted to balance before final assembly. Then I will charge with the 12v charger under the control of the BMS!

Its way faster to do the bulk of the charge as a serial pack with bms.

This document makes it easy https://diysolarforum.com/resources...ls-using-a-low-cost-benchtop-power-supply.65/

 

[John Frum]

Just to make it clear how long it could take.
Most cells come it 3.29ish volts which is ~50% charged.
100 amp hours * 4 cells * .5 state of charge / 3 amps = 66.6 hours

 

[benja.min47]

Just to make it clear how long it could take.
Most cells come it 3.29ish volts which is ~50% charged.
100 amp hours * 4 cells * .5 state of charge / 3 amps = 66.6 hours

Thank you for the link. I had consulted the method. On the other hand for the video of will, as I do not speak English I did not understand everything !!!
So let's imagine that I have more than 66 hours to lose, like will, I connect my power supply to the batteries connected in parallel, I set to 3.65v and I wait ?? The power will turn off by itself at the end ?? Can I put additional protections ?? thank you,

 

[HRTKD]

Thank you for the link. I had consulted the method. On the other hand for the video of will, as I do not speak English I did not understand everything !!!
So let's imagine that I have more than 66 hours to lose, like will, I connect my power supply to the batteries connected in parallel, I set to 3.65v and I wait ?? The power will turn off by itself at the end ?? Can I put additional protections ?? thank you,

Yes, you can wait. Patience is good. That's how I top balanced mine. If you're unsure about how the process will work, set the voltage target for a lower number, like 3.5v and see what it does when it gets to that point. The amps should drop to zero, or close to it, and the volts will not rise above 3.5. Then you can go to stage two and set the volts at 3.65.