Help Please! How do I test all 48 LiFePo4 cells on my my boat? (12v system, 3.2v 100Ah CALB)

[TommyHolly]

Hello,
Maybe someone can help me please? I suspect that some of my cells are bad. My main goal right now is to test all 48 cells but I'm not sure what is the best way? Wouldn't a capacity test show me a better, more definitive result? (Or does a capacity test on 48 cells take over a month like I was told?)
MY QUESTIONS FOR YOUR GUYS ARE AT THE BOTTOM.... thanks!

SYSTEM SPECS:
- I bought a 2012 Lagoon 450F sailboat boat down in the country of Panama which had a 12v, 1,200Ah LifePo4 system consisting of 48 CALB-brand cells, (3.2v 100Ah cells).
- The batteries are now about 6 years old, and the former owner, who was not an electrician, designed and installed the system himself.
- The original owner connected them first in groups of 4S (4 cells in series), to make twelve, 12v 100Ah batteries. Then he hooked up all twelve batteries in parallel to form a giant 1,200Ah system. Yes, I know that is not how you should connect them.
- Each group of 4s batteries had something called a "QNBBM 4S Lithium Battery Balancer" connecting each group of four. But nothing else.
- There was no BMS!! The previous owner bragged those 4S Lithium Battery Balancers were the BMS...ugh
- There was no Smart Shunt or any kind of way to monitor the system when I bought the boat. He used the Xantrex Inverter/Charger monitor which only displays voltage like a volt meter.
- The system is charged by 2,060W of solar panels, through three Outback Solar Charge Controllers (80A + 80A + 60A).
- The system is also charged through shore power through the Lagoon (Lagoon boat manufacturer) French 220v 60A Cristec battery chargers.
- There was a XANTREX 3,000w FSW3012 Inverter/Charger being used ONLY as an Inverter and not as a charger.

I'm pretty sure some of the batteries are bad but I've been getting bad advice from the "Lithium experts" here in the marina. I want to test these to make sure which ones work.

ORIGINAL PROBLEM:
- The original problem was that the entire system would shut off after turning the inverter on for a few minutes with any kind of load 300w to 1,000w
- The Xantrex 3,000w inverter/charger only had the inverter portion connected, the charger wasn't being used. Instead the shore power connects to the 220v 60A Cristec battery chargers.
- The voltage displayed on the Xantrex Monitor or with a voltmeter directly on the batteries would rapidly go down from 13.4v to about 12.7v to 12.9v and then the Xantrex would shut off. This would happen after only 60-120 seconds of operation. Measuring individual cell voltages on the batteries during this time, I would see it start about 3.4v and watch it go down to about 3.1v before it shut off.

THINGS I'VE FIXED AND NOW KNOW FOR SURE:
- I checked the CALB website and see the chart showing that these batteries drop off completely at around 3.15v for a cell. So they were rapidly discharging down to their lower limit of 12.7v.
- I also know now that LiFePo4 batteries shouldn't show much of a voltage drop, they should hold relatively steady and only drop off at the end when they are empty.
- I know that you need a BMS which is connected to all the batteries and not just the individual cells consisting of each 4S battery. So I think those QNBBM 4S Lithium Battery Balancers were useless.
- I know that batteries need to be first connected in Parallel, and then afterwards, connect them in series to make a 12v system. So eventually I'll be making a 12-cell 3.2v 1,200Ah battery (12P), and then connecting 4 of these together in series (4S), to make a 12P 4S 12v 1,200h system
- I purchased and installed a Victron 500A Smart Shunt, it works great on the temporary Lead Acid batteries I installed
- I purchased a DALY 500A LiFePo4 12v 4S Smart BMS and plan on using that when I put back the Lithium Batteries again. I will also hook them up first as four large 3.2v batteries in Parallel, and then connect those four in series to make them 12v.
- One of the Outback 80A solar controllers was bad, I've since replaced it and luckily they replaced it even though it was 2 months out of warranty. The solar system works great now.
- I found quite a few slightly loose connections, as well as a few connectors that weren't properly crimped on or corroded. When I re-install the system, I purchased a 1/4" thick copper bar to connect the cells to. And some new 1/0 AWG "welding wire" cable with new tinned-copper connectors and professional hydraulic crimper to make my own cables with heat shrinks.
- Also there was a leaky hatch directly above the batteries. The previous owner was an ass and tried to use caulk to stop the leak but that didn't work. The batteries were sitting in over 2" inches of water. (Luckily that was too low to reach the connections). But you could tell that water was dripping on top of them because many battery metal washers had signs of light rust or slight green corrosion. None of it was too bad.
- In the meantime, I needed my 12v lights, fans, refrigerators and freezer to work... so after removing the lithium cells, I was forced to purchase eight 160A Deep Cycle Marine lead acid batteries as a temporary measure until I can work out the problems. If the lithium cells are bad, I'll likely purchase two more of these lead acid batteries and give up with the lithium fix.

ADVICE I WAS GIVEN BY THE "EXPERTS" HERE AT THE MARINA:
- I was told to remove all the cells from the boat, connect all 48 of them in parallel, and place them on a Variable Voltage Charger exactly at 3.65v. I purchased two 10A variable voltage chargers and charged all 48 cells over six days exactly to 3.65v until the chargers displayed 0.0000A.
- To test the them I was told to disconnect all the cells, and record the voltage drop over approximately 2 weeks or whenever they stop dropping voltage, taking readings a few times a day.
- I watched the video by Will Prose showing a capacity tester, so I purchased one of those for $50 and had it shipped here to Panama, but both the "experts" at the Marina here said not to use that capacity tester because it's not needed. They said that the voltage drop readings were all I needed.
- I've been recording voltage levels on each individual cell now for six days. They all started at around 3.64v six days ago, and have been slowly dropping down due to internal resistance each day. Out of the 48 cells and six days of testing, about 12 cells are at around 3.621v, most cells (about 24 of them) are at 3.612v, and the worst 12 of them are at 3.602v
- The Marina "experts" said that these batteries are good. Maybe they are? I still don't trust them because they rapidly discharged before so quickly and I don't think that was from only slightly loose connectors. So I'll continue testing them for another week.

MY QUESTIONS FOR YOUR GUYS:
1. Should I continue testing voltage drop readings on individual cells for another week? The best cells have slowed to only dropping 0.002v a day and the worst cells drop 0.004v a day. The best cell went from 3.64v to 3.621v and the worst cell went from 3.64v to 3.602v. (Not that much difference at all.)

2. Should I perform a capacity test on each individual cell (after fully charging them back up to 3.65v of course). I watched Will's video and it took 5-6 hours for each test. So this would take weeks...

3. How else can I tell if these cells are good? My goal is to try and reinstall only the good cells and form a system comprising only of good ones as long as I have at least 600Ah. I don't want to be in the middle of the ocean and have my batteries go out because my navigational gear and autopilot rely on them, so this is a HUGE safety concern for me.

4. Do you guys think the problem is the batteries at all? It seemed like that to me since they discharged so quickly with such a light 1,000w load on them. 13.4v down to 12.7v was a drastic drop. If not, what else can it be? (The Xantrex Inverter is new, the Solar Controllers are new, and I tightened all the connectors before I disassembled the Lithium cells.) It maybe could be that some of the connectors were so loose they were falling off... It could be that the light rust on the washers was not that good of a connection. But would those problems show voltage dropping like that?

Please help because my life is at stake if this system fails on the open water on a long passage.

 

[MisterSandals]

Then he hooked up all twelve batteries in parallel to form a giant 1,200Ah system. Yes, I know that is not how you should connect them.

I dunno, connecting 12 (4S) batteries in parallel does not sound unusual or wrong. Wiring them in a balanced manner is the key to whether it will work or not.

Out of the 48 cells and six days of testing, about 12 cells are at around 3.621v, most cells (about 24 of them) are at 3.612v, and the worst 12 of them are at 3.602v

These all sound terrific!

Do you guys think the problem is the batteries at all?

My gut feel is no, not the batteries but loose connections and maybe poor connections of connectors and corroded surfaces.

It seemed like that to me since they discharged so quickly with such a light 1,000w load on them. 13.4v down to 12.7v was a drastic drop.

Is this with the new tightened connections? If so, it sounds bad. I would check the individual cell voltages and see which ones are significantly lower than the others and remove them and form 4S batteries with he best ones. Retest with same charge and 1000W discharge.

Test the cell voltages and clean up their connections to rule that out. Label your cells (with letters A, B, C...) and take notes on which ones are suspect and how they charge/discharge.

You may have a few bad connections fix things and you may have a few bad cells. I'd be optimistic you can get this to work as good as new and be a reliable battery bank.

I'll look over the pix and comment if i see anything noteworthy.

 

[mrzed001]

- The original owner connected them first in groups of 4S (4 cells in series), to make twelve, 12v 100Ah batteries. Then he hooked up all twelve batteries in parallel to form a giant 1,200Ah system. Yes, I know that is not how you should connect them.

Wow. That is a big 12V system. For that amount of battery at least a 24V or better a 48V system would be ideal.

- Each group of 4s batteries had something called a "QNBBM 4S Lithium Battery Balancer" connecting each group of four. But nothing else.
- There was no BMS!! The previous owner bragged those 4S Lithium Battery Balancers were the BMS...ugh

No BMS is always red flag.

- There was no Smart Shunt or any kind of way to monitor the system when I bought the boat. He used the Xantrex Inverter/Charger monitor which only displays voltage like a volt meter.

Volt means (almost) nothing in LiFePO4 cells.

- The system is charged by 2,060W of solar panels, through three Outback Solar Charge Controllers (80A + 80A + 60A).
- The system is also charged through shore power through the Lagoon (Lagoon boat manufacturer) French 220v 60A Cristec battery chargers.
- There was a XANTREX 3,000w FSW3012 Inverter/Charger being used ONLY as an Inverter and not as a charger.

At least not a big charger. Slow charge, cells had time to balance.

- The voltage displayed on the Xantrex Monitor or with a voltmeter directly on the batteries would rapidly go down from 13.4v to about 12.7v to 12.9v and then the Xantrex would shut off. This would happen after only 60-120 seconds of operation. Measuring individual cell voltages on the batteries during this time, I would see it start about 3.4v and watch it go down to about 3.1v before it shut off.

That is not such a big drop. Cutoff voltage set to 2,8V is enough.

- I know that batteries need to be first connected in Parallel, and then afterwards, connect them in series to make a 12v system. So eventually I'll be making a 12-cell 3.2v 1,200Ah battery (12P), and then connecting 4 of these together in series (4S), to make a 12P 4S 12v 1,200h system

Not a good idea. It is better in more parallel packs.

- Also there was a leaky hatch directly above the batteries. The previous owner was an ass and tried to use caulk to stop the leak but that didn't work. The batteries were sitting in over 2" inches of water. (Luckily that was too low to reach the connections). But you could tell that water was dripping on top of them because many battery metal washers had signs of light rust or slight green corrosion. None of it was too bad.

Ouch.

ADVICE I WAS GIVEN BY THE "EXPERTS" HERE AT THE MARINA:
- I was told to remove all the cells from the boat, connect all 48 of them in parallel, and place them on a Variable Voltage Charger exactly at 3.65v. I purchased two 10A variable voltage chargers and charged all 48 cells over six days exactly to 3.65v until the chargers displayed 0.0000A.

Good.

- To test the them I was told to disconnect all the cells, and record the voltage drop over approximately 2 weeks or whenever they stop dropping voltage, taking readings a few times a day.

To test them you need to give them constant charge and monitor the cell Voltages. Which fall hard below 2,8V are bad.

- I watched the video by Will Prose showing a capacity tester, so I purchased one of those for $50 and had it shipped here to Panama, but both the "experts" at the Marina here said not to use that capacity tester because it's not needed. They said that the voltage drop readings were all I needed.

Tester good. "Experts" not so good

- I've been recording voltage levels on each individual cell now for six days. They all started at around 3.64v six days ago, and have been slowly dropping down due to internal resistance each day. Out of the 48 cells and six days of testing, about 12 cells are at around 3.621v, most cells (about 24 of them) are at 3.612v, and the worst 12 of them are at 3.602v

Use the tester, give them load

2. Should I perform a capacity test on each individual cell (after fully charging them back up to 3.65v of course). I watched Will's video and it took 5-6 hours for each test. So this would take weeks...

If the tester can do 12V then make 12V packs and test that.

 

[MisterSandals]

Can you provide a rough diagram how the 12 batteries are connected together? I see several larger wires going to a connecting terminal at the top but its hard to see if the batteries are paralleled properly (balanced).

That is a big 12V system. For that amount of battery at least a 24V or better a 48V system would be ideal.

This is a good point. What is this battery bank used for? Is it mostly 12V lights and electronics? Or does this power a bow thruster (12V too?).

Stepping back and looking at your use case for these batteries (understanding the goals) might be helpful before proceeding with selecting a solution.

 

[TommyHolly]

I dunno, connecting 12 (4S) batteries in parallel does not sound unusual or wrong. Wiring them in a balanced manner is the key to whether it will work or not.

These all sound terrific!


My gut feel is no, not the batteries but loose connections and maybe poor connections of connectors and corroded surfaces.

Is this with the new tightened connections? If so, it sounds bad. I would check the individual cell voltages and see which ones are significantly lower than the others and remove them and form 4S batteries with he best ones. Retest with same charge and 1000W discharge.

Test the cell voltages and clean up their connections to rule that out. Label your cells (with letters A, B, C...) and take notes on which ones are suspect and how they charge/discharge.

You may have a few bad connections fix things and you may have a few bad cells. I'd be optimistic you can get this to work as good as new and be a reliable battery bank.

I'll look over the pix and comment if i see anything noteworthy.

I was told by DALY BMS manufacturer as well as a few videos I watched and other online groups, that I shouldn’t make small battery packs. That I should connect them all in parallel first and make 4 groups. Then take those 4 groups and wife them in series to make 12v. That’s the way the BMS can manage it.

Hopefully this new copper bar that I bought makes a better connection then those old washers and loose connectors on wires.

My re-test was after I tightened down everything and looked for any loose connections. It still failed after that. So I removed all the batteries and now I’m troubleshooting. I’ll label all my cells and I’ve taken detailed readings of each one on this resting test.

Yes, hopefully I can get this to work. The resting test worked well I think.

 

[TommyHolly]

Wow. That is a big 12V system. For that amount of battery at least a 24V or better a 48V system would be ideal.


No BMS is always red flag.

Volt means (almost) nothing in LiFePO4 cells.

At least not a big charger. Slow charge, cells had time to balance.

That is not such a big drop. Cutoff voltage set to 2,8V is enough.


Not a good idea. It is better in more parallel packs.

Ouch.

Good.

To test them you need to give them constant charge and monitor the cell Voltages. Which fall hard below 2,8V are bad.

Tester good. "Experts" not so good

Use the tester, give them load

If the tester can do 12V then make 12V packs and test that.

Most boats have a 12v system that everything runs off of. Why would I try to build a 24v or 48v system? Is need to covert down to 12v again anyway.

i’m not sure what the cutoff voltage was in the Xantrex inverter. I know I never reached that voltage or got close to it. The battery charts show that 3.1v is the final drop off for these batteries. That makes sense because the system would shut down around 12.7v which is about 3.15v per cell, the very bottom lower limit before it drops off a cliff on the chart.

ok, I’ll use the capacity tester after I finish this resting test. It’s the same exact one that will had in his video with the tiny fan on it and it looks like an open circuit board.

 

[TommyHolly]

Can you provide a rough diagram how the 12 batteries are connected together? I see several larger wires going to a connecting terminal at the top but its hard to see if the batteries are paralleled properly (balanced).


This is a good point. What is this battery bank used for? Is it mostly 12V lights and electronics? Or does this power a bow thruster (12V too?).

Stepping back and looking at your use case for these batteries (understanding the goals) might be helpful before proceeding with selecting a solution.

The way the system was set up before I removed everything was that four 3.2v 100Ah cells were combined In series to make a single 12v 100Ah battery. This was repeated twelve times so I had 12 batteries. And those 12 batteries were connected in parallel to make 1,200Ah. The wiring was a little goofy though, he had random batteries jumping off in parallel with other random packs. So it wasn’t in a long row, just mixed and connected wherever as long as it made a parallel circuit.

The battery bank is mainly used for lights, fans, fridge and freezer, auto pilot and navigational gear and a 1,000w 10A Portable Air Conditioner sometimes. No bow thrusters. 3,000w Xantrex is more than I need.

 

[MisterSandals]

That I should connect them all in parallel first and make 4 groups. Then take those 4 groups and wife them in series to make 12v. That’s the way the BMS can manage it.

Someone usually points out that this is the way car batteries are made: many in parallel and then series connected. I am not saying it doesn't work.

I actually have my main RV bank as 2P4S and it works great. I did spend a LOT of time building battery packs and testing to determine and rank the cells. I started with 13, took the best 8 to make my big bank. I paired best with worst, second best with second worst... This gave me my best shot at having all 4 pairs being equal capacities. I think this is an effective strategy.

I did not like the idea of managing 2 batteries with 2 BMSs so that was a big contributor to my 2P4S decision. All by other batteries are 4S so i don't have any further experience... you'd be out on the deep end of the pool.

 

[MisterSandals]

And those 12 batteries were connected in parallel to make 1,200Ah. The wiring was a little goofy though, he had random batteries jumping off in parallel with other random packs.

This would have been key to keeping the batteries equal to each other. It sounds like it wasn't the worst case but I suspect it may have lead to imbalances between batteries.

The battery bank is mainly used for lights, fans, fridge and freezer, auto pilot and navigational gear and a 1,000w 10A Portable Air Conditioner sometimes. No bow thrusters. 3,000w Xantrex is more than I need.

Yea that sounds reasonable, i don't see a compelling reason to switch to 24V or 48V battery. The 3000W inverter is on the raged edge of what can be wired and fuse (250A). Hopefully you wire and fuse it for the limits of the inverter so the next guy to own or work on the system isn't setup for catastrophe.

 

[TommyHolly]

Someone usually points out that this is the way car batteries are made: many in parallel and then series connected. I am not saying it doesn't work.

I actually have my main RV bank as 2P4S and it works great. I did spend a LOT of time building battery packs and testing to determine and rank the cells. I started with 13, took the best 8 to make my big bank. I paired best with worst, second best with second worst... This gave me my best shot at having all 4 pairs being equal capacities. I think this is an effective strategy.

I did not like the idea of managing 2 batteries with 2 BMSs so that was a big contributor to my 2P4S decision. All by other batteries are 4S so i don't have any further experience... you'd be out on the deep end of the pool.

You answered a follow up question. What do I do when I’ve finally retested all the batteries and rank them from best to worst. You are saying pair them up so that all the batteries are as equal as possible in capacity.

So cells 1 and 48 go together in parallel. Cells 2 and 47 parallel in the next battery. 3 and 46…. And keep stacking until I have 12 cells in parallel to make 4 batteries. Then connect them in series.
*Note: Because my DALY BMS only has 4 wires and is a 4S, that’s the only way I can connect it.

I spoke with one of the “experts” last night. I told him that these resting voltage tests aren’t providing much data and all the batteries are relatively the same reading.

I plan on starting a capacity test Individually for each cell on Monday. Or would it be more effective if I grouped the cells together by rank from the resting voltage test, and perform a capacity test that way, 4 cells at a time forming a battery?

 

[schmism]

When the batteries are in parallel they are forced to be at the same voltage.
Ensuring all the batteries in the pack stay as closely balanced as possible means you get the best performance out of the pack.
Therefor, having the most batteries in parallel as possible means by default the the most batteries will be the same voltage, for this reason its typically recomend to have as many in parallel BEFORE you put those sets in series.

You started with 4s 12p. This means each 4s pack needs its own BMS. (12 BMS required for best, safest performance)
Instead moving to a 12P 4S (12 batteries in parallel, and 4 of those groups in series) means you need one large BMS to control the entire pack and "most" of the batteries are in parallel.

Pros and cons.

1) 4s 12P - many small BMS mean they are inexpensive and have lots of redundancy, if one fails you simply remove 4 batteries controlled by that BMS from the pack.
2) 12 bms = each bms only need to be sized to handle 1/12 of the total load as they are all in parallel and can share the load. (typically people plan for some overhead ie if 1 or 2 fail the remaining can take the total load. (3000w/12v=250a/10bms = 25a bms instead of 20a ones.
3) 12P 4S - the most batteries in parallel for best balance - one bms
4) 1 bms - it has to be 250a (for 3kw load) this can be a single point of failure for the pack.
5) 12p4s - a single large pack. This would typically be all contained in the same physical location which can be difficult to do on a boat so haveing several smaller collections of cells spread out over the boat could be better.

Per the last item, you also have the option of doing things between the 2 extremes. 2 sets of 6p4s in parallel. This is basicly 2 large batteries that you could split and put in different sections of the boat. again because you should have a BMS for each pack it means each one could be sized for 125a understanding that if one failed you would limited to 125a of load instead of the full 250.

My personal take on configurations from best to worst (however thats mostly due to on paper design and not considering IRL space considerations or BMS price/size considerations)
1 - 12p 4s
2 - 6p 4s
4 - 3p 4s
12 - 1p-4s (or more commonly writen) 4s 12p

Technically 6 - 2p-4s is also an option however i consider this to be basicly the same as the last setup. My only reason for doing this was due to BMS size/quantity/price limitations i might have access to.

 

[MisterSandals]

What do I do when I’ve finally retested all the batteries and rank them from best to worst. You are saying pair them up so that all the batteries are as equal as possible in capacity.

What i did for my 2P4S was to make the 2P pairs (for of them) equal to each other. So i paired best and worst, second best with second worst... all within a battery.

You're a level of difficulty further in figuring out how to make batteries (a lot of them) work together in parallel.
I think the most important detail will be wiring them as balanced as possible.

I really like the arguments that schmism made in the previous post.

If i could, i would experiment with different configurations to see if your batteries are matched enough for 12P4S or somewhere along the spectrum towards 4P12S. If you come up with any test results or new info, we're here for you to adjust your strategy(s).

 

[TommyHolly]

When the batteries are in parallel they are forced to be at the same voltage.
Ensuring all the batteries in the pack stay as closely balanced as possible means you get the best performance out of the pack.
Therefor, having the most batteries in parallel as possible means by default the the most batteries will be the same voltage, for this reason its typically recomend to have as many in parallel BEFORE you put those sets in series.

You started with 4s 12p. This means each 4s pack needs its own BMS. (12 BMS required for best, safest performance)
Instead moving to a 12P 4S (12 batteries in parallel, and 4 of those groups in series) means you need one large BMS to control the entire pack and "most" of the batteries are in parallel.

Pros and cons.

1) 4s 12P - many small BMS mean they are inexpensive and have lots of redundancy, if one fails you simply remove 4 batteries controlled by that BMS from the pack.
2) 12 bms = each bms only need to be sized to handle 1/12 of the total load as they are all in parallel and can share the load. (typically people plan for some overhead ie if 1 or 2 fail the remaining can take the total load. (3000w/12v=250a/10bms = 25a bms instead of 20a ones.
3) 12P 4S - the most batteries in parallel for best balance - one bms
4) 1 bms - it has to be 250a (for 3kw load) this can be a single point of failure for the pack.
5) 12p4s - a single large pack. This would typically be all contained in the same physical location which can be difficult to do on a boat so haveing several smaller collections of cells spread out over the boat could be better.

Per the last item, you also have the option of doing things between the 2 extremes. 2 sets of 6p4s in parallel. This is basicly 2 large batteries that you could split and put in different sections of the boat. again because you should have a BMS for each pack it means each one could be sized for 125a understanding that if one failed you would limited to 125a of load instead of the full 250.

My personal take on configurations from best to worst (however thats mostly due to on paper design and not considering IRL space considerations or BMS price/size considerations)
1 - 12p 4s
2 - 6p 4s
4 - 3p 4s
12 - 1p-4s (or more commonly writen) 4s 12p

Technically 6 - 2p-4s is also an option however i consider this to be basicly the same as the last setup. My only reason for doing this was due to BMS size/quantity/price limitations i might have access to.

Hey thanks for that reply.
The original configuration the previous owner had was 4S 12P. Each battery consisting of 4 cells had that QNBBM Lithium Battery module attached to each cell for a total of 4 QNBBM’s. Those would supposedly balance out the cells in each 4S series set. I’d see them work because the red discharge light and green charge lights would go on. (See attached picture.)

I have no idea how to properly test these, nor do I know how to find which ones are the problem to make the system crash like that dropping from 13.4 to 12.7 within 2 minutes. So far, the resting voltage tests show these are all within 0.02v of each other after top balancing over a week ago. I hope my capacity tests next week discover the bad ones. (If any of these are truly bad?)

I bought a DALY 500A Smart BMS 4S LiFePo4 so that should be enough or at least I was told to handle a 1,200Ah 12P 4S system.

I’m wondering if I should also use those old QNBBM Lithium Balancers in conjunction with the DALY 500A BMS? Im guessing I would connect one balancer to each pack of 12P batteries? I read through some other threads talking about how a BMS doesn’t do a good job of balancing? I’ll probably start a new thread when it comes closer to the time when I put the Lithium back in.

 

[TommyHolly]

What i did for my 2P4S was to make the 2P pairs (for of them) equal to each other. So i paired best and worst, second best with second worst... all within a battery.

You're a level of difficulty further in figuring out how to make batteries (a lot of them) work together in parallel.
I think the most important detail will be wiring them as balanced as possible.

I really like the arguments that schmism made in the previous post.

If i could, i would experiment with different configurations to see if your batteries are matched enough for 12P4S or somewhere along the spectrum towards 4P12S. If you come up with any test results or new info, we're here for you to adjust your strategy(s).

When you say matched enough, what kind of tolerance are you talking about here?
Matched by resting voltage?
Matched by capacity?

For example as of a few hours ago and a week after top balancing all 48 cells in parallel, my best battery is 3.618v and my worst is 3.598v…
A difference of 0.02v or 0.55%

 

[MisterSandals]

When you say matched enough, what kind of tolerance are you talking about here?
Matched by resting voltage?
Matched by capacity?

My understanding it that capacity is the most important. Its when one cell fills or empties faster than the others, and this is very prevalent at the knees (very high and very low voltages).

My "testing" (really using daily and monitoring which cells were the outliers at the high and low states of charge) was with 4S (and 2P4S) batteries. I kept good notes of which ones were the outliers, rearranged them many times until i had enough data (actually when we moved plants out of the nursery and the daily use stopped).

So not really scientific but the number of observations were significant.

 

[wholybee]

Other than the lack of proper BMSs, the setup doesn't seem too far off. It did last 6 years, so it isn't all wrong.

For testing, I would test each 4s pack by itself, and only go to a cell level of testing if I found a problem. Get at least 1 BMS rated for 100A or more. A Daly or overkill will work. (eventually you will want 12 BMSs)

Connect the BMS to a 4s pack. Then charge it to 3.65V, or until the BMS disconnects. If the BMS disconnects early, note the voltage of all the 4 cells. Balance them, and try again. Connect the pack to your inverter. Then run a large load. Your goal is to draw 100A. A 1500W space heater is too much, but on the low setting might be about perfect. Note the current draw, and how long it takes the battery to fully discharge. If it lasts about an hour with a 100A load, consider that pack good. Charge it up again, and move on to the next.

1200Ah of Lithium on a boat is gawd awful huge. All said and done, if 4 packs (400Ah) test good, then you are still in good shape. If the chargers were setup correctly so they didn't overcharge the batteries, there is a good chance most or all are ok. I doubt he ever ran a battery that size dead flat.

Also, a fully charged cell is ~ 3.4V after it comes to a resting voltage. Since all of your cells are over that, you seem in good shape. That 0.02 difference in voltage at over 3.5V is much less than 1% of capacity. They are still in pretty good balance. If you want better, you probably need a much much larger power supply than a pair of 10A to balance 48 cells. But, you could balance them in groups of 4. It'll take a long time, but I think it will work better than 20A into 48 cells.

 

[HRTKD]

I'll be the voice of dissent on your cell configuration. I like redundancy. A 12p4s configuration has no redundancy. 4s12p is a lot of BMS to keep track of, but it gives you the highest granularity of measuring each cell. With this configuration, if you have a bad cell, you're going to know it. This configuration would also let you take one or more batteries offline for maintenance without losing the entire battery bank.

A configuration somewhere in the middle might be more palatable? 2p4s6p?

 

[David S]

I'll be the voice of dissent on your cell configuration. I like redundancy. A 12p4s configuration has no redundancy. 4s12p is a lot of BMS to keep track of, but it gives you the highest granularity of measuring each cell. With this configuration, if you have a bad cell, you're going to know it. This configuration would also let you take one or more batteries offline for maintenance without losing the entire battery bank.

A configuration somewhere in the middle might be more palatable? 2p4s6p?

In this scenario, I tend to agree with HRTKD. If you lose up to 11 x 4s packs, you still have 12v for critical loads. On a boat at sea, that might be critical. Totally a back of the napkin opinion, but…. Other thing that does is limit your reliance on being able to move huge amperage through a single BMS. Kinda like installing a bunch of battleborns in parallel. Downside is, you need 12 bms

 

[TommyHolly]

My understanding it that capacity is the most important. Its when one cell fills or empties faster than the others, and this is very prevalent at the knees (very high and very low voltages).

My "testing" (really using daily and monitoring which cells were the outliers at the high and low states of charge) was with 4S (and 2P4S) batteries. I kept good notes of which ones were the outliers, rearranged them many times until i had enough data (actually when we moved plants out of the nursery and the daily use stopped).

So not really scientific but the number of observations were significant.

I tried my first capacity test using that cheap $40 capacity tester that will had on his channel. For some reason I can’t get it to work, but it might be because of the battery?? Maybe I discovered my first bad battery?

- I too balanced the battery at 3.65v until the charger showed 0 amps going in.
- When I hooked up the battery to the tester, it showed 3.52v
- I watched the video again and set up the tester so it shuts off at 2.60v as a Low Cutoff Voltage like he recommended.
- Then I adjusted down the current to 20A, which on my battery is 2C, just like in the video.

**Almost immediately, the Low Voltage alarm went off and the test stopped immediately. The screen flashed “2.60!” I couldn’t resume the test because the battery would instantly drop down under 2.6v during the test??

I grabbed a multimeter and it was actually displaying 3.17v directly on the battery. So maybe the capacity tester is bad??

I was forced to turn down the current to 15A and the voltage on the Capacity Tester says 2.7v which is dangerously low to shutting off the test again. The multimeter directly on the cell showed 3.102v. So maybe I have to turn it down all the way to 1C 10A???

Im thinking either the cheap tester is bad or the battery is bad. I may make a new thread about this tester thing to see if anyone has experience using it?

 

[mrzed001]

I tried my first capacity test using that cheap $40 capacity tester that will had on his channel. For some reason I can’t get it to work, but it might be because of the battery?? Maybe I discovered my first bad battery?

- I too balanced the battery at 3.65v until the charger showed 0 amps going in.
- When I hooked up the battery to the tester, it showed 3.52v
- I watched the video again and set up the tester so it shuts off at 2.60v as a Low Cutoff Voltage like he recommended.
- Then I adjusted down the current to 20A, which on my battery is 2C, just like in the video.

**Almost immediately, the Low Voltage alarm went off and the test stopped immediately. The screen flashed “2.60!” I couldn’t resume the test because the battery would instantly drop down under 2.6v during the test??

I grabbed a multimeter and it was actually displaying 3.17v directly on the battery. So maybe the capacity tester is bad??

I was forced to turn down the current to 15A and the voltage on the Capacity Tester says 2.7v which is dangerously low to shutting off the test again. The multimeter directly on the cell showed 3.102v. So maybe I have to turn it down all the way to 1C 10A???

Im thinking either the cheap tester is bad or the battery is bad. I may make a new thread about this tester thing to see if anyone has experience using it?

I think the tester V meter is wrong, maybe needs to be calibrated ?