48v 16 cell.. do i need to charge individual cells before I put in series and charge?

[beowulf]

Hey all. Got my battery kit ready to go.. but I read that all cells must be top charged and balanced before you assemble the 16 3.2v in to a 48v setup and then use your daily charger on it?

My cells arrived with 3.02v charge. One or two have 3.01 and 3.03. Do I need to individually charge each batter to 3.5 or so.. and if so.. how bad is it if they are off by .01 to .05 or so? The kit has a BMS with it, and I think there is a balance feature there as well. It's Jehu Gracia's 48v 74ah per cell battery kit. So I dont want to mess this up.

I'll be receiving my 48/3000 Victron MultiPlus 2 in a couple days, so hoping I can add the AC input cord to that, and use AC wall power to charge the battery through the inverter/charger during off peak hours.

Thanks.

 

[BentleyJ]

3.0V is almost fully discharged but the cells have such a flat voltage curve its difficult to determine exact SoC based on voltage. If the cells are all at 2.5V or 3.5V then they are balanced but anywhere in the middle, voltage is not as certain.

Its highly recommended to place the cells in parallel and charge at no more than 0.2C (15A) until you reach 3.5V and the cells are taking very little to no current.

 

[beowulf]

3.0V is almost fully discharged but the cells have such a flat voltage curve its difficult to determine exact SoC based on voltage. If the cells are all at 2.5V or 3.5V then they are balanced but anywhere in the middle, voltage is not as certain.

Its highly recommended to place the cells in parallel and charge at no more than 0.2C (15A) until you reach 3.5V and the cells are taking very little to no current.

So I have a 30v/10a bench charger I bought for drone batteries.. not sure it works with LifePO4 but someone said it would work. It's a small unit ($100 from GetFPV called ToolkitRC) vs the ones I see others use that cost like $50 on amazon.. which are much larger but seem to also put out up to 30v/10a.. not sure there is a concern with my unit being so small or if the big ones are just cheap china junk that uses larger crap parts or what. This has banana plugs for output with an XT60 connector.

So I should put all 16 cells in parallel.. e.g. wire up all 16 neg to one neg lead and all 16 pos to one pos lead, then attach to this, and set it to 3.5v at 10a and let it go? Each cell is 74ah.

IF that is what to do.. great.. but then how do I monitor the actual batteries.. I dont want the expanding, overcharged, etc. They are in the rack mount with bars around them to keep them from expanding, but worried if one cell is not doing well.. how do you tell? I don't have an external BMS.. not opposed to buying one as I plan on building one more of these (have all but the batteries to do so).. and a few REPT/CATL 320ah battery packs eventually. But not sure if I have to have that or do I just let this go.. and how do I know when it's fully charged?

I have a nice Fluke V multimeter.. not sure if there is a way to get better reading on voltage than my cheap one.. do I need more than .00 places of accuracy to ensure they are all balanced?

These battereis been sitting brand new for about 2 months since I got them. Hence I am hoping at 3.02 or so they aren't ruined.

I'd also like to try to keep them at about 80% tops. I think I read that keeping between 20% and 80% would give them 6000+ charge cycles vs going to full which could yield about 3000 charge cycles (0 to 100 DOD?).

What sort of gauge wire should I use as the "bus bar" that each of the 16 leads on neg and pos attach to that then plugs in to the bench charger? As well does it matter each of the 16 leads (for both neg and pos) gauge?


Thanks.

 

[BentleyJ]

So I should put all 16 cells in parallel.. e.g. wire up all 16 neg to one neg lead and all 16 pos to one pos lead, then attach to this, and set it to 3.5v at 10a and let it go? Each cell is 74ah.

If the power supply as described is Constant Current & Constant Voltage controlled then Yes, this is correct. The power supply should ramp up to 10A and hold there while the voltage slowly increases to 3.5V. The batteries will absorb at 3.5V and the current will naturally start to fall off, called tail current. When the current drops to 0.5A or less all the cells should be fully charged.

I dont want the expanding, overcharged, etc. They are in the rack mount with bars around them to keep them from expanding, but worried if one cell is not doing well.. how do you tell?

If you don't abuse the cells with high rates of charge or discharge they shouldn't expand. A good BMS will monitor individual cells and calculate wire resistance. If one starts to show high resistance or develops a high rate of internal discharge (if that cell always seems to need more balancing than the others)

I don't have an external BMS.. not opposed to buying one as I plan on building one more of these (have all but the batteries to do so).. and a few REPT/CATL 320ah battery packs eventually. But not sure if I have to have that or do I just let this go.. and how do I know when it's fully charged?

If the BMS that came with the kit doesn't have a Bluetooth or other means of communicating and displaying cell voltages I wouldn't use it at all. Get a JK-BMS with 1A cell balancing built in.

I have a nice Fluke V multimeter.. not sure if there is a way to get better reading on voltage than my cheap one.. do I need more than .00 places of accuracy to ensure they are all balanced?

Fluke is fine, that's what I use. If you get the JK-BMS you can calibrate the pack voltage using the Fluke. Most BMS read out to 0.001V but are realistically only accurate to about 0.003V in the best case.

These battereis been sitting brand new for about 2 months since I got them. Hence I am hoping at 3.02 or so they aren't ruined.

2.5V is the lower safe limit. If they are all still at 3.0V or above they are good. If not, the cell was probably already bad when it was delivered.

I'd also like to try to keep them at about 80% tops. I think I read that keeping between 20% and 80% would give them 6000+ charge cycles vs going to full which could yield about 3000 charge cycles (0 to 100 DOD?).

I typically charge with very low current at 3.45V per cell. That will get them to 90%+. So far after almost 5 years there is no swelling or issues. I believe if you only charge to 80% you are wasting capacity.

What sort of gauge wire should I use as the "bus bar" that each of the 16 leads on neg and pos attach to that then plugs in to the bench charger?

For 10A, 14ga is recommended. Probably could get away with 16ga. Just don't use alligator clips. Crimp a good terminal on the wire and tighten it down on the cell post or screw.

 

[beowulf]

This is the BMS I have for the battery kits: https://www.aliexpress.us/item/1005...in.404.26331802QyqIVl&gatewayAdapt=4itemAdapt

 

[beowulf]

OK.. so if I understand everything.. I will wire up 32 x 14awg wires of equal length with round connectors on the ends that get attached to each pole and bolted on. I then twist tie (or attach) 16 pos to one single thicker wire and same for the 16 neg, to put all 16 cells in parallel. Those two leads go to my 30v/10a bench supply, and I set that to 3.6v to get near the top, and to 10a. I let it run to charge all 16 cells in parallel. Is that correct so far? Does it matter that I am not using that diagonal lead style that Prowse shows in his video with the bus bar for multiple batteries given that I am attaching 16 leads (of each pos and neg) to a larger single lead and not a long bus bar?

Do I use my Fluke in DC mode to touch the two leads and verify that it is at 3.6v with almost no amp draw to know when to stop charging?
Do I then touch pos/neg for each cell to verify they are all 3.6v?
How do I know they are all balanced? Is there some variation like if one is 3.61v and another is 3.59.. then I need to manually charge the one to 3.61? Would I still charge that at 10a or lower that to like 1a?

These are all 74ah 600a cells.

Thank you.

 

[LakeHouse]

OK.. so if I understand everything.. I will wire up 32 x 14awg wires of equal length with round connectors on the ends that get attached to each pole and bolted on. I then twist tie (or attach) 16 pos to one single thicker wire and same for the 16 neg, to put all 16 cells in parallel. Those two leads go to my 30v/10a bench supply, and I set that to 3.6v to get near the top, and to 10a. I let it run to charge all 16 cells in parallel. Is that correct so far?

This will work but take a long time. If your cells are nearly empty, it will take 74ah * 16 cells / 10a = 118 hours, or about 5 days.
If you want to speed it up, you can assemble into the final pack (with BMS!!) and charge with the Multiplus until the BMS cuts off charging, then disassemble and top balance as you note.
There is a very good resource on top balancing here.

 

[beowulf]

Ah.. ok.. so I do have it all assembled.. just took off the PCB that puts them all in series to turn it in to a 48v battery pack. I thought it was basically best to first top balance before you turn in to a 48v battery? It wont damage/degrade the cells using the built in BMS to charge first? Also.. why can't it handle the top charge with the BMS? If I can charge it all the way up using the BMS in 48v mode.. why do I need to then run them in parallel at 3.2v to charge them up just a tiny bit more? Why wouldn't the 48v mode finish the job and balance them? I believe the JMD BMS has a balance charger in it? Could be wrong.

 

[Brett V]

If you want to avoid problems in the future, a proper top balance is essential.

 

[LakeHouse]

I thought it was basically best to first top balance before you turn in to a 48v battery?

You should top balance before you use it as a battery, but think of this as more like a part of the assembly process.

It wont damage/degrade the cells using the built in BMS to charge first?

No, the BMS will disconnect when one cell reaches over-voltage protection.

Also.. why can't it handle the top charge with the BMS? If I can charge it all the way up using the BMS in 48v mode.. why do I need to then run them in parallel at 3.2v to charge them up just a tiny bit more?

Because this won't balance the cells. You need each individual cell to be charged to 100% for them to be balanced. The BMS will stop charging when exactly one cell gets to 100%, but the rest will be less than that.

I believe the JMD BMS has a balance charger in it?

I'm not familiar with this BMS. If it has active balancers, this can work. If they are passive balancers, they don't bleed off enough energy to balance the cells.

 

[beowulf]

You should top balance before you use it as a battery, but think of this as more like a part of the assembly process.

No, the BMS will disconnect when one cell reaches over-voltage protection.

Because this won't balance the cells. You need each individual cell to be charged to 100% for them to be balanced. The BMS will stop charging when exactly one cell gets to 100%, but the rest will be less than that.

I'm not familiar with this BMS. If it has active balancers, this can work. If they are passive balancers, they don't bleed off enough energy to balance the cells.

Ah.. didnt know that they stop once one cell gets there. The BMS has 17 wires that plug in to my PCB.. one for each cell. I assumed this was to monitor the cell and/or balance them. Maybe it's just to monitor each cell?

From the video Jehu shares, it looks like the little display that you can use with it shows the 16 cell voltages with 4 numbers of precision. So if all cells are 3010 to 3030.. 3.6v is the max.. but it would ideally show 3600 on the display. Do I need to get to 3600 on all 16 cells, or could it be 3580 to 3600 or so? Just unsure how accurate these things are and where to stop (or the BMS cuts off). Jehu's site says these are 3.5v max so not sure if that is accurate or not. 3.2v nominal. It's a china 100a balancer.. so not sure if it is accurate or not.

So.. charge all in series using BMS to 3.4v maybe, then undo, put in parallel, charge to 3.5 (or 3.6..whatever the max is). But in parallel will it ALSO stop soon as the first one reaches 3.6.. or rather.. without a BMS attached.. just bench power to the leads.. is there any concern about one cell going a little over while the rest catch up.. or does that parallel approach (per Prowse video) slowly stop charging once all equal out at the voltage specified on the bench supply?

 

[LakeHouse]

I assumed this was to monitor the cell and/or balance them. Maybe it's just to monitor each cell?

Definitely to monitor, maybe also to balance but that depends on the BMS.

Do I need to get to 3600 on all 16 cells, or could it be 3580 to 3600 or so?

You want to get all the cells to the same voltage above 3.45V. Below 3.45V the charge curve is flat, so for example they could all be at exactly 3.30V, but range in State of Charge from 30%t to 60% (these numbers might not be exact, but the idea is that SoC and voltage correlate poorly below 3.45V).

So.. charge all in series using BMS to 3.4v maybe, then undo, put in parallel, charge to 3.5 (or 3.6..whatever the max is).

Yes. You could go higher than 3.4, but you have the idea.

But in parallel will it ALSO stop soon as the first one reaches 3.6.. or rather.. without a BMS attached.. just bench power to the leads.. is there any concern about one cell going a little over while the rest catch up.. or does that parallel approach (per Prowse video) slowly stop charging once all equal out at the voltage specified on the bench supply?

When they're connected in parallel you're forcing them all to the same voltage, so they all reach 3.6V at the same time. One cell couldn't possible have higher voltage that the others because the voltage difference would create a current that would discharge the high cell into the lower ones.
Just be aware that there is no BMS protection in parallel, so you must ensure that your power supply is set to not exceed your safe voltage (usually 3.65V for LFP cells, but anything lower than that but above 3.45 is fine).

 

[beowulf]

Hey thank you for this. So I have used my little 30v/10a charger (that I got for drone bench work). I also bought a 2nd one on amazon because figured charging each cell would take a while.

So I am not sure if I am doing something wrong at this point. When I put my drone one (little tiny ToolkitRC P200).. on the battery which is at 3.0v.. the amp jumps to 9.8. I set the V to 3.4. These LEV60 seem to top out at 3.5 if I am reading things right, so I assume I do NOT want to go past 3.5v on the power supply? Like can I do 3.7 to push it a little faster charging until the battery itself is up to about 3.4v, then dial the V back down to 3.5? Or is anything over 3.5 out of the bench supply going to damage the battery?

Anyway, it took about 3 to 4 hours to get one cell from 3.0 to 3.3v. Around 3.3 to 3.35 or so. I was thinking I would get all 16 cells to 3.3v exactly, then build the 48v battery with the BMS, etc. BUT.. you said get them all to 3.45 or so before I do that? So basically I would set the power to 3.5v, and just let it run for about 10 hours or so for each cell... checking every hour or so to see where it is at? I am using my Fluke multimeter which has 4 places of accuracy apparently, so hopefully given the insane expense of that multimeter and the name, it's pretty accurate.

So does that seem right? I think I read/saw that as you near the top of the V (in my case 3.45 or 3.5 maybe?) the amp will go down to < 1 amp. But here is the thing. On the amazon 30v/10a bench supply, it was at <1a when the battery was around 3.25v. In fact, my little tiny TookitRC was charging quite a bit faster than the other one.. so I guess that little thing at about 2x the price is indeed using some good components vs the china $50 one I got.. because it was taking a few hours more to go from 3.25v to 3.3v or so.

So I saw Prowse video on putting all 16 cells in parallel, and he said it should only take a few hours using that cheap bench supply (I got the same one he did) to charge all 16 up.. and yet.. I can't seem to get it to charge a single cell in hours.. so am I doing something wrong, or does parallel charging work about the same speed as a single cell?

My concern is these LEV60 (https://jag35.com/products/high-discharge-600a-lifepo4-1780w-prismatic-cells?variant=40806504398963) are 3.5max.. so is setting the supply to 3.5 ok or is there a risk of damaging them at that setting? More so.. given that I charge each of the 16 cells to 3.3v and they all dropped to 3.28 to 3.30 after a couple hours.. I don't know how the hell I will know if they are actually top balanced.. because if I get one to 3.5 or 3.45.. and then move the bench supply to another, that first one is going to drain a bit.. so how do I get them all accurate together to then assemble and finally put on the inverter charger so that each night that thing can charge them up? I know the 100a BMS that Jehu uses has balancing on it.. I believe on both charge and discharge.

The other concern is.. I did read that from time to time the batteries can somehow get out of balance again. So how the hell do you know this? I am all in on the Victron stuff, and I'd buy one of those $130 or or so battery shunts if that is a good way to be able to know, but that seems like its for the whole 48v pack, not individual cells. So do I have to check the display on the battery like every few days for ever.. or is this something that happens over the course of a year or so? Assuming it does.. do I now have to disassemble the battery each time, pull the cells out, charge them again, etc? Like given they are LiFePo4 and there is a BMS.. I assumed once they are set, that's it.


Lastly (probably again).. but I am still not clear if either these cells (Because they are 3.5 not 3.65.. are they different) can go from 0 to 100 to 0 over and over.. or if I should set my Victron inverter/charger to charge at 20% (or 10%?) and stop at 80% or 90%? I want to maximize the cells.. and given that we seemingly have to top charge(balance) them to like 99% capacity before using.. not sure if that means to keep them all balanced we always have to charge to 99% or so each day so that the BMS can ensure they stay balanced?


Again thank you for all the info. Excited to get this damn battery in use! I have everything (but the cells) to build a 2nd one and want to get that in place soon as well.

 

[LakeHouse]

battery which is at 3.0v.. the amp jumps to 9.8. I set the V to 3.4.

This is the wrong order. Set the power supply voltage with nothing connected, THEN attach a cell and don’t adjust the voltage again. Otherwise you risk exceeding your target voltage and damaging the cell.

I assume I do NOT want to go past 3.5v on the power supply?

Correct.

Like can I do 3.7 to push it a little faster charging until the battery itself is up to about 3.4v, then dial the V back down to 3.5?

No, don’t do this. If you need to charge faster, improve the connection between the power supply and the cell. Ditch the alligator clips and small-gauge wire and connect with 12 or 10 gauge and a ring terminal.

I was thinking I would get all 16 cells to 3.3v exactly

Getting them all to 3.3V won’t balance them. They all need to be at the same voltage, connected in parallel, above 3.45V, at the same time. Otherwise they aren’t balanced.

But here is the thing. On the amazon 30v/10a bench supply, it was at <1a when the battery was around 3.25v. In fact, my little tiny TookitRC was charging quite a bit faster than the other one.. so I guess that little thing at about 2x the price is indeed using some good components vs the china $50 one I got.. because it was taking a few hours more to go from 3.25v to 3.3v or so.

Bad connections. Use your fluke while it’s charging to measure the voltage at the power supply and the voltage at the cell (as in, the actual cell terminals). Then improve the connections as noted above.

does parallel charging work about the same speed as a single cell?

Charging 16 cells in parallel with take 16 times as long as a single cell.

More so.. given that I charge each of the 16 cells to 3.3v and they all dropped to 3.28 to 3.30 after a couple hours.. I don't know how the hell I will know if they are actually top balanced.. because if I get one to 3.5 or 3.45.. and then move the bench supply to another, that first one is going to drain a bit..

Not drain, just settle to a lower voltage. Voltage and SoC are poorly correlated with this chemistry. That’s why you need to get all cells at the same high voltage to balance them. And be aware that as soon as you disconnect power, the voltage will decrease, but SoC will still be 100%

 

[beowulf]

This is the wrong order. Set the power supply voltage with nothing connected, THEN attach a cell and don’t adjust the voltage again. Otherwise you risk exceeding your target voltage and damaging the cell.

Yah.. sorry.. that is what I am doing. Set the voltage, then apply the leads. Yeah the connection from the cheap one is thin wire alligator clips.. so that makes sense. Likely cant push more amps.

SO.. because they settle by about .5v it seems (is that normal..if so how you ever get them all to 3.65 or 3.5 in my case if each one settles when I take the leads off?).. then do I basically want to set my V to 3.5, and let it go until what.. .01amps or something on the amp output? When do you know its at the max (without it damaging the cell)? Does it always take so long? These are 74ah cells too so I assumed much faster than the 180ah or 320ah from various videos. But it sounds like you're saying get them to 3.5, they'll settle to about 3.45 give or take a few .01, .02. Is it ok if some are 3.45, some are 3.46, some are 3.43 and so on.. close enough at that point? I think I saw on these that 3.4 is like 99% and 3.45 is 99% and 3.5 is 100%.. so.. not sure how I would know they are at the highest they can go (per cell).

Once I assemble the kit.. and put the Victron 48/3000 inverter/charger on the battery, how do I avoid or limit the balance disparity over months (or longer hopefully)? Or is there just no way to avoid that and it is required to take apart battery and manually charge them each again to 3.45 or so to top balance them again?

 

[time2roll]

If they have been charged in series and look good the passive balance function of the BMS should keep the cells balanced. Any significant out of balance could take a very long time as the passive system is very slow. If balance is an issue or becomes an issue I recommend adding an active balancer such as NEEY to get them back marching together. I would not do a parallel top balance again once placed in series.

 

[LakeHouse]

then do I basically want to set my V to 3.5, and let it go until what.. .01amps or something on the amp output?

You don't need to go right to 0 Amps, and it might even be slightly detrimental because it’ll take a very long time. You’ll find lots of different answers, but if the current has started to taper off and you’re getting 3.5V at the terminals, you’ve probably gone far enough.

Is it ok if some are 3.45, some are 3.46, some are 3.43 and so on..

Once you’ve top balanced them all to 3.5 at the same time, you’re done. They settle to some lower value, and they might not all be exactly the same, but that’s ok, they’re balanced. Make them into a battery.

Once I assemble the kit.. and put the Victron 48/3000 inverter/charger on the battery, how do I avoid or limit the balance disparity over months (or longer hopefully)? Or is there just no way to avoid that and it is required to take apart battery and manually charge them each again to 3.45 or so to top balance them again?

Once they’re balanced, if you use them and they get fully charged occasionally so the BMS’s balancer can do its thing, you shouldn’t have to rebalance them.

 

[beowulf]

You don't need to go right to 0 Amps, and it might even be slightly detrimental because it’ll take a very long time. You’ll find lots of different answers, but if the current has started to taper off and you’re getting 3.5V at the terminals, you’ve probably gone far enough.

Once you’ve top balanced them all to 3.5 at the same time, you’re done. They settle to some lower value, and they might not all be exactly the same, but that’s ok, they’re balanced. Make them into a battery.

Once they’re balanced, if you use them and they get fully charged occasionally so the BMS’s balancer can do its thing, you shouldn’t have to rebalance them.

OK.. so last question then.. should I put them ALL in parallel (I did buy 40 bus bars to connect them together) and charge it that way? I would prefer not only because I already have the batteries in the rack mount with the compression rods in place.. so was hoping to avoid taking it all out, but if that is the best way to get them to 3.5v together, I'd do that. If not, then you're saying just charge each up till the meter reads about 3.45 to 3.5, then stop charging.. charge the next one..etc.. until they are all settled to around 3.45 or higher. So if some are 3.44, 3.47, 3.45, etc.. that's OK?

 

[LakeHouse]

should I put them ALL in parallel (I did buy 40 bus bars to connect them together) and charge it that way?

Yes, this would be the best way.

If not, then you're saying just charge each up till the meter reads about 3.45 to 3.5, then stop charging..

I didn’t say this. I wouldn’t recommend it, but it could work. The idea would be to get all cells at or very close to 100% SoC and hope for the best.

until they are all settled to around 3.45 or higher. So if some are 3.44, 3.47, 3.45, etc.. that's OK?

Don’t worry about what voltage they settle at. If you got them to 3.5V with minimal current being accepted, they’re at or close to 100% SoC. Whether they settle to the same voltage or not is irrelevant.

 

[LakeHouse]

Don’t worry about what voltage they settle at. If you got them to 3.5V with minimal current being accepted, they’re at or close to 100% SoC. Whether they settle to the same voltage or not is irrelevant.

Maybe just to clarify a little: The goal of a top balance is to get each cell to the same SoC (100%), not the same voltage.