Top Balancing (16), 280AH LiFePO4 cells. Please check my Plan.

[PeteW]

Hi folks,
I need to top balance (16), 280AH Keheng prismatic cells. The voltages range between 3.29V and 3.28V with only one cell at 3.27V I do not have them yet but I intend to use OverKill BMS. This battery bank will end up in an Airstream Travel trailer. I've been kind of flipping back an forth on battery size. I initially planned on building 12V batteries, then began considering 24V batteries because I heard it would be easier on the inverter. But, because I already have (8),12V, 100W Rich Solar PV panels and everything in my Airstream will be either 120VAC or 12VDC I'm back to my initial plan of 12V. All of this indecision is the reason I haven't purchased my BMSs or my MPPT Solar controller. I've had a bad case of paralysis of analysis.

I’d like someone to check my understanding of the top balancing procedure and point out any mistakes in my understanding. The welded studs on my cells are 6mm. I am uncertain of the correct torque specs (Newton meters, Inch-pounds) to torque the nuts to. I would appreciate that information or a source for it, please. I also have some additional questions below that I’d appreciate some advice on as well.

My power supply is a 300W (10A, 30V) Long Wei. My multimeter is a Klein MM600. For the PS I'll be using 10 AWG multi strand tinned copper silicone insulated lead wires with 6mm ring terminals at both ends of the wires, to charge the cells / batteries.

To frame my understanding and my questions, I’m using excerpts from Filterguy’s excellent, “Tutorial: Top Balancing LiFePo4 Cells using a low cost 10A benchtop power supply”.

“3 Pre-charging the cells to get them ‘mostly’ charged

1. Assemble the cells in series and add your BMS to create a full battery.”

Since I intend to compress my cells, am I correct to assume that it’s best to compress the cells for this? I’m planning to use plywood, all thread rods, washers and nuts to contain and compress the cells. Is there a source for what the torque would be on the nuts that tension the all thread rods and plywood?

“2. Set the BMS to shut off charge when any cell reaches 3.65V. (3.6V for 3.6v top balance)”

Understood. With the BMS connected to the battery, configure it to shut off the charge when any cell reaches 3.65V.

“3. Set the power supply constant current to .2 C or less.
A. While the power supply is disconnected from the batteries and turned off, short the leads together.
B. While the power supply is off, Set the voltage arbitrarily high and the current as low as it will go.
C. Turn on the power supply and slowly turn up the current till you hit the target current.
D. Disconnect the leads
Note: For most cells used in solar systems, .2 C is a lot more than 10 Amps. (.2C of 100Ah cells is 20A). Consequently, in most cases you can just max out the current setting on the supply.

4. With the power supply disconnected, set the voltage to your target pack voltage.
For a 12 V system: 3.65 X 4=14.6 Volts. (Or 14.4 V for 3.6v top balance)...”

I’m going to paraphrase this to be sure I have the correct understanding.Please correct me if I’ve misinterpreted the above instructions 3-3 (A-D) and 3-4 (12V system).
I’m building (4), 12V batteries (4 cells in series).
With the PS off, turn the course Voltage knob to high, but leave the Amp knob off. Then short the positive and negative leads together. Turn the PS on and turn up the current until it maxes at 9.8A -10A (as high as it will go). Then disconnect the lead ends. With the lead ends disconnected adjust the voltage to 14.6V (4 x 3.65). Turn off the PS. Then attach the positive lead to the positive battery post, and the negative lead to negative battery post and turn on the PS. Monitor and disconnect the PS when the battery reaches 14.6 or a single cell reaches 3.65 (BMS should shut down charging also). This will need to be done to each set of 4 cells (in series). Realistically I’m guessing this will take a couple of weeks.

“4 Top Balancing your cells

... Note 2: The procedure below uses 3.65V as the target voltage for top balancing. …
1. Wire all your cells in parallel.
Note: If your cells are more than about .15V different, it is best to charge the lower cells to about the same as the higher cell(s) before paralleling them. To do this, parallel the lower voltage cells and use this top balance procedure to get them close to the voltage of the higher cells and then top balance them all)”

After pre-charging all of the cells. If some of the cells are more than .15V lower than the higher cells use steps 4-5 and 4-6 to parallel charge the lower cells before top balancing all of the cells together.

Would it make sense to charge my 2.27 cell to 2.9 before doing anything else, even before pre-charging the 4 cell batteries?

"2. Before hooking the power supply to the cells, set the power supply to .2C or less of the combined AH of the cells.
This will usually turn out to be the max current setting of the power supply.
3. Set the voltage on the power supply to 3.65V."

Set up the PS for Top Balancing. Same as the procedure for charging the "in series" 12V batteries.

"4. Hook the negative lead to the negative pole of one end of the series of cells and the positive lead to the positive
pole at the other end of the series of cells as shown in the image on the right (Be sure to get the polarity
correct). The diagonal hook-up helps balance the resistance between the power supply and the individual cells"

Connect the cells in parallel for top balancing. Connect the positive lead from the PS to the positive end of the parallel string of cells and the negative lead from the PS to diagonal negative post at the opposite end of the parallel string of cells.

"5. …NOTE: Be patient. Depending on the state of charge when the top-balancing started, this can take a long time. Perhaps multiple days. It may sit at the same voltage and not change for a long time. This is normal. Do not give up and do not try to increase the voltage to speed things up.

Because the voltage clamp on the low-cost power supplies often has a small drift (Typically ~0.02V) I do not leave cells unattended if the Power Supply is set at 3.65V. If I need to leave the cells unattended, I will set the PS to 3.6V or even 3.55V and let them top charge unattended. When I have time to ‘babysit’ the cells I will then re-do the balance at 3.65V. Since the cells are already top balanced to above 3.5V, the final top balance is very quick (Even at 3.55V I do not leave them unattended for extended periods. I do not want them to sit at even 3.55 volts and no current for more than overnight).”

Even based on how long it will take to charge all 16 cells (as 4 cell batteries), I’m guessing that once the last set of 4 are charged I’ll be ready to top balance them, correct? Or, will they be at such a high state of charge that the Top Balancing would go pretty fast? At this point I would be connecting all 16 cells in parallel, correct?

So, depending on their SOC, my understanding of 4-5 is that this process, using the 10A PS may still take some time, possibly a couple of days. It’s unrealistic to think that I could monitor them closely continuously for 24 to 48 hours. So, based on my understanding I should probably set the voltage to 3.55V at the most (maybe only 3.5V), checking on them frequently except overnight, until I can remain with them as they are topping off. At that time I could safely increase the voltage to 3.65V (4-6 notwithstanding). Disconnecting them from the PS when they current drops to zero.

“6. Once you see the voltage at the cells reach ~3.5 volts, the voltage will start going up much faster. At this point start watching the voltage at the power supply closely. I do not trust the constant voltage circuit on the power supply to hold the voltage exactly at 3.65V, so at the end of the charging I use my multimeter and adjust the power supply to make sure the voltage at the power supply does not go above 3.65V. Do not worry if the voltage at the cell is lower than the Power Supply…they will catch up. (I find that the cells take 1-4 hours to get to 3.65 V with 200AH 24V banks. Higher capacity banks of cells will take longer)”

I am not sure I understand the proper way to use a multimeter to adjust the power supply toward the end of the charging cycle. Would I simply put the positive and negative multimeter probes on the positive and negative connections on the PS to check the PS voltage reading against the multimeter? Then trust the multimeter over the PS?

 

[sunshine_eggo]

To frame my understanding and my questions, I’m using excerpts from Filterguy’s excellent, “Tutorial: Top Balancing LiFePo4 Cells using a low cost 10A benchtop power supply”.

Since I intend to compress my cells, am I correct to assume that it’s best to compress the cells for this? I’m planning to use plywood, all thread rods, washers and nuts to contain and compress the cells. Is there a source for what the torque would be on the nuts that tension the all thread rods and plywood?

Compressing during top balancing is not necessary, but it's fine to do.

I’m going to paraphrase this to be sure I have the correct understanding.Please correct me if I’ve misinterpreted the above instructions 3-3 (A-D) and 3-4 (12V system).
I’m building (4), 12V batteries (4 cells in series).
With the PS off, turn the course Voltage knob to high, but leave the Amp knob off. Then short the positive and negative leads together. Turn the PS on and turn up the current until it maxes at 9.8A -10A (as high as it will go). Then disconnect the lead ends. With the lead ends disconnected adjust the voltage to 14.6V (4 x 3.65). Turn off the PS. Then attach the positive lead to the positive battery post, and the negative lead to negative battery post and turn on the PS. Monitor and disconnect the PS when the battery reaches 14.6 or a single cell reaches 3.65 (BMS should shut down charging also). This will need to be done to each set of 4 cells (in series). Realistically I’m guessing this will take a couple of weeks.

Simplified:
Power supply on, disconnected from the battery.
Set voltage to 14.6V
Connect to battery.
Crank amps to max

You could build a 4P4S battery w/BMS and charge them all at once.

Most cells come at around 30%-50% SoC. If you do it 4P4S, it will take up to 78 hours to fully charge from 30%.

After pre-charging all of the cells. If some of the cells are more than .15V lower than the higher cells use steps 4-5 and 4-6 to parallel charge the lower cells before top balancing all of the cells together.

This has been demonstrated to be unnecessary. The concern was that there would be excessive current between high and low, but the current is well within the 140A limit. It is also more and less important at different times. If cells are at 3.35 and 3.65, the 3.65V cells will be pulled down very rapidly because it's in the upper leg of of the voltage curve. Loaded cells rapidly drop into the working range and narrow the voltage gap.

Cells at empty (less than 2.9V resting) and cells at full (>3.5V resting) can see some pretty substantial currents for extended periods.

One can also mitigate the current. A bus bar tightened down to the proper torque spec will conduct max current. A bus bar secured simply with finger tight fasteners will conduct far less current.

Would it make sense to charge my 2.27 cell to 2.9 before doing anything else, even before pre-charging the 4 cell batteries?

I assume you mean 3.27 to 3.29V.

No need to pre charge low cell.

Set up the PS for Top Balancing. Same as the procedure for charging the "in series" 12V batteries.

Yep. Can use simplified steps above.

Connect the cells in parallel for top balancing. Connect the positive lead from the PS to the positive end of the parallel string of cells and the negative lead from the PS to diagonal negative post at the opposite end of the parallel string of cells.

Yep.

Even based on how long it will take to charge all 16 cells (as 4 cell batteries), I’m guessing that once the last set of 4 are charged I’ll be ready to top balance them, correct? Or, will they be at such a high state of charge that the Top Balancing would go pretty fast? At this point I would be connecting all 16 cells in parallel, correct?

Assuming you need to replenish 10% of the total battery capacity in this phase, 44.8 hours.

So, depending on their SOC, my understanding of 4-5 is that this process, using the 10A PS may still take some time, possibly a couple of days. It’s unrealistic to think that I could monitor them closely continuously for 24 to 48 hours. So, based on my understanding I should probably set the voltage to 3.55V at the most (maybe only 3.5V), checking on them frequently except overnight, until I can remain with them as they are topping off. At that time I could safely increase the voltage to 3.65V (4-6 notwithstanding). Disconnecting them from the PS when they current drops to zero.

This is fine.

I am not sure I understand the proper way to use a multimeter to adjust the power supply toward the end of the charging cycle. Would I simply put the positive and negative multimeter probes on the positive and negative connections on the PS to check the PS voltage reading against the multimeter? Then trust the multimeter over the PS?

The PS will report the voltage it sees after wiring losses. The multimeter will see the direct open circuit voltage and will be more accurate.

Basically, if you see the voltage > 3.65 on the multimeter, adjust the power supply to a lower voltage.

Thanks for reading the guide! Many just skim it and go do their version of it.

 

[Substrate]

So, based on my understanding I should probably set the voltage to 3.55V at the most (maybe only 3.5V), checking on them frequently except overnight, until I can remain with them as they are topping off. At that time I could safely increase the voltage to 3.65V (4-6 notwithstanding). Disconnecting them from the PS when they current drops to zero.

As long as "safely increase the voltage to 3.65v" means *disconnecting*, and adjusting your voltage following the multimeter, not the power supply as a reference, and reattach like you did earlier in the process initially.

This is how the guys who do it in steps do it - DISCONNECT - adjust voltage a little higher with multimeter - reconnect and never crank on the voltage knob while connected. Your time limitation kind of mimics those who do this stepped version by choice.

 

[PeteW]

sunshine-eggo

Thank you for taking your precious time to help me!

Compressing during top balancing is not necessary, but it's fine to do.

Since I would be creating a battery, even if just for the pre-charge before top balancing I thought that compression would be needed. I don’t want to be doing things just for the sake of doing. I would like to understand why. I thought compression was to prevent expansion due to charging and discharging. Now I’m wondering if I understand the reason for compression. Where can I learn more?

You could build a 4P4S battery w/BMS and charge them all at once.

Due to my newbieness I had not really considered going this route (4P4S).
Would this be a good way to go for my finished battery?
Would I just use a single 12V BMS for this?
Would a single BMS be adequate protection?
Would this present any Solar Controller issues or problems?

Most cells come at around 30%-50% SoC. If you do it 4P4S, it will take up to 78 hours to fully charge from 30%.

This is embarrassing, but I received these cells nearly one year ago. The cell voltage measurements are nearly the same as when I received them, but I’m sure they have been losing power just sitting.

One can also mitigate the current. A bus bar tightened down to the proper torque spec will conduct max current. A bus bar secured simply with finger tight fasteners will conduct far less current.

Understood, I wrote to Keheng a couple of times and asked about torque with regards to using nuts on their welded 6mm studs. I never received an answer. Do you know what the proper torque is, or where I can look it up.?

I assume you mean 3.27 to 3.29V.
No need to pre charge low cell.

Yes , I meant 3.27 - 3.29V. I understand about no need to pre-charge the low cell.

The PS will report the voltage it sees after wiring losses. The multimeter will see the direct open circuit voltage and will be more accurate.

Basically, if you see the voltage > 3.65 on the multimeter, adjust the power supply to a lower voltage.

I understand about the possible voltage loss. Do I understand correctly that I check the voltage at the Positive and Negative connections on the PS where the leads originate?

Thanks for reading the guide! Many just skim it and go do their version of it.

Hey, in another thread I'd posted you told me to read it. I took you at your word.

 

[PeteW]

As long as "safely increase the voltage to 3.65v" means *disconnecting*, and adjusting your voltage following the multimeter, not the power supply as a reference, and reattach like you did earlier in the process initially.

This is how the guys who do it in steps do it - DISCONNECT - adjust voltage a little higher with multimeter - reconnect and never crank on the voltage knob while connected. Your time limitation kind of mimics those who do this stepped version by choice.

Understood. Yes that is what I meant, except for your tip about using the multimeter. Thank you.

 

[sunshine_eggo]

sunshine-eggo

Since I would be creating a battery, even if just for the pre-charge before top balancing I thought that compression would be needed. I don’t want to be doing things just for the sake of doing. I would like to understand why. I thought compression was to prevent expansion due to charging and discharging. Now I’m wondering if I understand the reason for compression. Where can I learn more?

All cells expand on charging. The amount is very small. Compressing them improves cycle life. Conducting the top balance without compression won't have a measurable effect on cycle life.

Due to my newbieness I had not really considered going this route (4P4S).
Would this be a good way to go for my finished battery?

With 4S4P (4 separate 12V), you have redundancy and individual cell monitoring.

Would I just use a single 12V BMS for this?

Yes, but you would be BMS current limited, i.e., your big 4P4S battery would still be limited by the BMS current rating. With 4S4P, you have almost 4X the BMS current capability.

Would a single BMS be adequate protection?

Yes.

Would this present any Solar Controller issues or problems?

Only the current limitation.

This is embarrassing, but I received these cells nearly one year ago. The cell voltage measurements are nearly the same as when I received them, but I’m sure they have been losing power just sitting.

That's a very good sign.

Understood, I wrote to Keheng a couple of times and asked about torque with regards to using nuts on their welded 6mm studs. I never received an answer. Do you know what the proper torque is, or where I can look it up.?

4-6 N-m.

I understand about the possible voltage loss. Do I understand correctly that I check the voltage at the Positive and Negative connections on the PS where the leads originate?

No. On the battery terminals to which the PS is attached.

Hey, in another thread I'd posted you told me to read it. I took you at your word.

 

[PeteW]

Thanks again for all of this help!

 

[PeteW]

Spoiler: Doing some Top Balancing 4 cells at a time

sunshine_eggo

diysolarforum.com

Substrate

diysolarforum.com

sunshine_eggo

diysolarforum.com

Substrate

diysolarforum.com

No BMS as yet, so I'm doing some Top Balancing 4 cells at a time (As opposed to building 12V batteries and pre-charging). If you see this I'd be interested in your opinions on this. I do not have a BMS yet. More on this below.
I began doing some top balancing 4 cells at a time yesterday. I only did them for an hour per set of 4 cells. I did two sets yesterday and two sets today, 4 cells per set, 16 cells altogether. I did take notes about cell voltages before, during and after. I took notes on PS and multimeter voltages during and PS amperage reading at the beginning and end of each top balance session. I was pretty conservative. I turned the amperage up to 10A approximately, but kept the voltage at 3.55V. I numbered each cell when I first received them, so I've kept a log. I checked the voltage of each cell before starting so I could have something to reference. I won't get into a long explanation, preferring to wait for questions. I mostly want to make sure I'm not doing any harm. This is mostly a learning exercise for me.

Regarding BMSs:
I've been strongly considering the Over Kill BMS. Mostly because I know very little about BMSs, and the Overkill BMS is recommended by Will and others in the forums. At the same time I began leaning toward the Overkill BMS I was also considering building a 24V battery bank. Now, I'm pretty sure a 12 V battery bank will be a more practical way to go since it is for my Airstream trailer and at this point all of the low voltage lighting, appliances, etc. will be 12V. A DC to DC converter just adds another layer of complexity and something else to go wrong. Just this past weekend someone suggested I consider a Victron BMS. I am considering a Victron Energy Hybrid 3000W inverter and solar controllers, shunt, etc. Nothing, is set in stone. All I have at this point are 16, 280AH LiFePO4 cells, 8, Rich Solar (RS-M100, VOC 22.8V) 100W Monocrystalline PV panels. I've purposely waited to buy solar controllers, BMS/s, and components while I learn more about them. For me it's still all a bit over whelming, but little by little I'm digesting it. I'd appreciate some feed back on this topic, particularly about BMS brands. I do know that I want temperature sensing capability and Bluetooth capability. The Victron Energy 12/200 claims to monitor and control energy coming from the 12V alternator. In my case this would be from the tow vehicle. I haven't read enough yet to understand all of this completely.

 

[willbarrow]

This thread and those referenced have been extremely helpful to me as I am building a very similar pack and am new to all of this—3 4S packs with 12 new 280AH CATL cells and 3 Overkill BMSs. Thank you for confirming your process here for others to follow along.

I began doing some top balancing 4 cells at a time yesterday.

I've been unable to find an answer to the question of top balancing cells together or separately when they will be used in separate 4S packs but combined in parallel for one battery bank. I can understand the need to top balance 16 cells together if they'll be part of a single 48v pack, but is there any reason to top balance all 12 of my cells together if they will be in separate packs each with their own BMS?

If indeed it is best to top balance all cells together, that raises another question of capacity testing packs together or separately. I imagine that testing 3 12v 4S packs one by one will be more efficient than trying to discharge my whole battery bank to cuttoff voltage in a single test and weed out any bad cells; however, after discharging and fully recharging separately, won't that potentially put each of my packs at a different SOC making them less efficient when combined into a single bank? Perhaps I'm overthinking this or not understanding how the BMSs work to equalize across packs. I'd really appreciate any insight anyone can share on how this works.

One more thing. Prior to top balancing I am charging all of my cells in a 3P4S configuration with a single BMS at 14.6v and a cell cuttoff voltage of 3.65v. I'm thinking I should reduce the cell cutoff voltage to prevent possibly over charging any of my parallel cells since the BMS sees 3 cells as one in this configuration. How much should I reduce the voltage if at all? Thanks in advance.

 

[PeteW]

This thread and those referenced have been extremely helpful to me as I am building a very similar pack and am new to all of this—3 4S packs with 12 new 280AH CATL cells and 3 Overkill BMSs. Thank you for confirming your process here for others to follow along.


I've been unable to find an answer to the question of top balancing cells together or separately when they will be used in separate 4S packs but combined in parallel for one battery bank. I can understand the need to top balance 16 cells together if they'll be part of a single 48v pack, but is there any reason to top balance all 12 of my cells together if they will be in separate packs each with their own BMS?

If indeed it is best to top balance all cells together, that raises another question of capacity testing packs together or separately. I imagine that testing 3 12v 4S packs one by one will be more efficient than trying to discharge my whole battery bank to cuttoff voltage in a single test and weed out any bad cells; however, after discharging and fully recharging separately, won't that potentially put each of my packs at a different SOC making them less efficient when combined into a single bank? Perhaps I'm overthinking this or not understanding how the BMSs work to equalize across packs. I'd really appreciate any insight anyone can share on how this works.

One more thing. Prior to top balancing I am charging all of my cells in a 3P4S configuration with a single BMS at 14.6v and a cell cuttoff voltage of 3.65v. I'm thinking I should reduce the cell cutoff voltage to prevent possibly over charging any of my parallel cells since the BMS sees 3 cells as one in this configuration. How much should I reduce the voltage if at all? Thanks in advance.

Hey there Will,

I'm also very new at all of this, but you are asking all of the right questions and your instincts off separating cells are correct. Hopefully you are also using the wiki and resource library, where there are many fine articles, to get some "good lern'n" .

I've been completely side tracked building some furniture and most recently taking care of my wife who had hip surgery last week and is now in a brace for the next 6 weeks. Fun - Fun! Anyway, before I got myself into trouble doing too many difficult projects at once, the link below was my latest thread on my top balancing quest and the help I was receiving. Possibly it will offer some help to you. I will be getting back to it, but probably after the holidays at this point.

https://diysolarforum.com/threads/testing-top-balancing-experiment-4-cells-at-a-time.48797/page-2

I recommend starting your own thread and posting questions, it's the best way to get noticed. Someone will take you under their wing and provide some really helpful guidance. If you do it, send me a link so I can follow it.
Pete

 

[willbarrow]

Hi Pete, Thanks for your response and for sharing that link. There is definitely a lot of great info here and quite the community. All the best to you with your distractions, and happy holidays!

 

[PeteW]

Hi Pete, Thanks for your response and for sharing that link. There is definitely a lot of great info here and quite the community. All the best to you with your distractions, and happy holidays!

Thanks, same to you.

 

[Eoin]

Just joined the forum.
Excellent thread and thanks for the concise information and time taken to describe in detail the explanations.