Hi,

Hope someone can help me here with my unbalanced cells. 

Since I set up my DIY LifePO4 battery from lithiumbatteriessa and attached the Daly BMS about a year ago, I have had one pair of cells that has never charged as high as the rest (I have a 2p8s - 24V battery bank). There are differences also between the other cells, but this pair stands out. When it starts charging from the sun this pair is initially above the rest of the batteries, but then the others overtake it later in the day and it ends up 0.2V lower when the BMS seems to stop charging. Once the sun has set and there is an ongoing load of 150-250W they do seem to settle down with very little difference between cells, usually less than 0.01V. I know that the cells do start to differ more as the state of charge gets very low, especially under load.

It seems, based on shunts and BMS Ah counting, that I am about 10% short on the full 5.1kWh capacity.

 

I saw a Youtube video which David Poz uploaded this week where he also had some lower voltage cells and didn't get the full capacity out of a new battery. He opened it up and attached a resistor to the highest voltage cells to bring them down and force the BMS to start charging again. Eventually he got the cells better balanced and the full capacity from the battery. I'm not sure how long it would stay in balance after this, but I'd like to give it a try if it might help. I don't feel like the Daly BMS has done anything towards balancing these cells in the last year, maybe because they were so far out to start with and it can only balance in the mA range and is turning off when some cells get to a high voltage.

I definitely don't want to unscrew all my batteries and put them in parallel, partially because I don't have a CC/CV charger and also because of the hassle and fear of stripping the threads on the aluminium posts on these batteries which seem very soft.

I realise I will have to wait until the pack is fully charged during the day and then sit there and attach the resistor to the highest cells and wait for the lower voltage cells to charge. Would a resistor like this be suitable?  I think if I bought a 20 ohm / 50W one, I should be able to put it over several cells at a time rather than buy multiple resistors to put one on every other cell? 

RB50 10R - Communica [Part No: RB50 10R]

Thanks in advance!

 

1st off, are you sure that the cell measurement/balance lead is not somehow having a bad contact? You can measure with an accurate multimeter

Balancing shoud start when cells reach 3.46V not below

2nd Daly seems to only balance while there is a charge current flowing. Best way I found to balance my cells would be to charge from solar till 100% or close (cells showing 3.45 - 3.5V while accepting 10A). I guess you have half the cells (16S2P) so slowly charge the cells with 5A and let the balancer do its thing. Now keep on charging with 5A and give it time to work( migth take days if very unbalanced). 

I would not take the pack apart, unless there is something seriously wrong.

yea the resistor does the same job as the bulb in my case

 

naturally the bulb can't go over too many cells bulb would pop

how many cells is in your battery

 

the balance current on most bms's are realy low , only suitable for maintenance ballance ie to keep it balanced

 

and naturally i did not want to spend money or waste is more the word on a one time device

 

once balanced they stay in balance very nicely

 

bms balancing is to keep cells balanced not to balance them you will wait a year for it to complete

 

now naturally i don't have a bench powersupply and could not do it the right way, and wasn't willing to spend that much money one time devices

would rather spend it on extra cells

 

conneting them parralel to balance doesn't work if the voltage is low

you gotta ballance on at least 3.4 or higher which requires the bench psu to push the voltage while paralel 

 

found this graph on a blog that mentioned the bulb idea

 

yea and the daly only balances while charging 

 

making it even worse stretching the time

 

 

imo you dont have to wait till its fully charged

 

ie the cells that have a higher voltage especially if there is a big difference , will always run away and cause the bms to cut of on over voltage protection

so evey time you pull down the ones with the resistor/bulb they will help the lower voltage ones to take a bit more charge next time 

 

it feels like it will take forever the moment you cross 3.3-3.35 v  per cell it starts to feel quicker hang in there

 

 

I = V / R   So that means with a single cell 3.3V / 20 Ohm = 0.165A ... pretty much the same or less than the daly BMS.

You will burn off P = V * I 

3.3V * 0.165A = 0.544 Watts so connected for 2 hours you will remove about 1Wh from a single cell

@iiznh I don't think it's a bad connection, but you are right, I should measure them with my multimeter just in case.

Assuming that's not the issue, I'm not sure how I should limit the amps when charging since I don't have anything other than solar to charge the battery bank with....l guess I could limit the charging amps within the settings of my solar charger/inverter? 

I've also read up that maybe I should set the charge voltage from my solar charger/inverter to a lower voltage. At the moment I have it around 28.0 absorb and 27.6 float (I'll have to check that though). Lower voltage apparently would mean the highest voltage cells wouldn't hit their maximum voltage so soon and the BMS might carry on charging. But I still think that it would not be able to balance such a big difference and I'll need to do the resistor/bulb thing. 

Regarding the calculation, I was going to connect the resistor over 3 or 4 cells, so the voltage would be higher, but it would still be small load per cell, so could reduce the resistance down closer to 5 ohms and/or use them in parallel (they are only rated 50W)

@Leondavibe I'm only 8s, so I might be doing 3 on the one side of the cell and 4 on the other. 4 cells might be just ok on a 12v bulb. I'm not actually sure what the resistance of a 12v bulb is offhand, and how many amps it will burn off. Thanks for the diagram, you are right, it seems like it could take quite a while!

I log my individual cell voltages to a database, so it will be interesting to see how long they stay balanced after doing all this!

yea lower voltage can extend the balancing time but since the daly balances so slow its a pointless excercise sacrificing some solar power

50w /3.2v would mean about 15a or on 4 cells ie 50w/12.8 = would be 3.9a

i would not leave anything attached to the cells bypassing the bms protections unattended

 

ie imagine accidently making a calculation error to come back and find a ceel discarged too far damged

 

the reason why i connected my charger over 3 was just to save time ie the amps drop the closer you get to full

by just charging 3 in a go the amps would be higher because the charger would charge as if the 12v battery is realy low

 

and charging in 3 made sense to me since mine is 15s

as yours is 8s it devides nicely into 4

 

if you dont have a car charger you can just charge it during the day via solar

 

discarge the offending cells and let it charge via solar net day

 

each night just discharging via resistor the cells thats highest voltage down

 

just get them closer to the lowest cells 

and next time it charges the lowest voltage cells will reach a ighr SOC though its hard to measure progress once you go over 3.35-3.4 yopu can see the progress

and naturally you will be able to see the progress by going a bit longer evrery time on the battery during load shedding

 

 

though on a singlke cell you dont get the fool oomph as the element doen't get that hot

if you put a resistance on some cells whille charging from solar they will just charge slower

 

ie i found that two 50w bulbs on a single cell would cause it not to charge at all while the others would keep charging

ie its voltage stayed stuck on 3.5 while others would still lift

the problem for me was the wires of the downlighter was a bit thin on the fitting that connected to the bulbs to handle more than 2 cells

 

ie you can add more resistance until you see the voltage doesnt lift that is essentially what the bms tries to do but since it doesnt have enough resisyance to stop the cell from running away before the lowest cells gets full its voltage either pushes total volatge to cut-off or individual cel over voltage protection

you can even add to much resistence so its voltage goes down as long as you remove the bulb before it goes below the lowest voltage cells 

 

i did this once  removing too late which just changes which cell is the bottom cell ie farming a bit backwards

If you can also try get a decent active balancer and use that every other month to keep them nicely top balanced

i cant recasll i think i was charging at 5-6 amps and the two bulbs would keep the 1 cell stagnant at high voltage ie 3.5-3.6

and at lower voltage 3.35-3.4 one was fine

 

was rather playing safe instead of accidently discarging a cell to make the new low voltage cell

you could also just swop cells ie take one cell at the end off put it in the low cells place 

and place low cell at the end

 

ie now you have your high cells in a row and just need one resistor or concentrate on one instead of two on either side

or remove the cell and place at end and move all sense wires

 

 

2 minutes ago, WannabeSolarSparky said:

If you can also try get a decent active balancer and use that every other month to keep them nicely top balanced

if you have a balancer

i did the bulbs cause i dont have one and just blew my budget on the extra cells

2 minutes ago, Leondavibe said:

just blew my budget on the extra cells

Hahaha I would have prioritised that way too

and these cells don't go out of balance that easily

pretty sure you would go months

before you see an imbalance

 

wanne buy the jk bms that has active balance and higher balance current

1 minute ago, Leondavibe said:

wanne buy the jk bms that has active balance and higher balance current

I have 2 on the way from china gonna rip out my 2 batteries builtin bmses they are pretty useless.
They are awesome and will keep your batteries healthy and in spec when correctly configured

Using a element/globe/resistor bypasses the BMS, there will be no protection.

Since it is once cell that is low, I would rather opt to charge the once cell. Than discharging the rest. There is no way to bring them down to the same level since lithium voltages mean nothing.

Charge the battery till full. Disconnect the BMS by plugging out the connector. We do not want to damage the BMS with a higher than spec voltage.

You can effectively take any charger that is regulated to have a max current. Eg cellphone charger that outputs 5V 2A. Connect the 5V directly to the low cell and charge it. Keep a multimeter connected and make sure you do not pass 3.6V. Babysit it till done. Once done connect the BMS again and see if your battery is in balanced.

2 hours ago, iiznh said:

Using a element/globe/resistor bypasses the BMS, there will be no protection.

Since it is once cell that is low, I would rather opt to charge the once cell. Than discharging the rest. There is no way to bring them down to the same level since lithium voltages mean nothing.

Charge the battery till full. Disconnect the BMS by plugging out the connector. We do not want to damage the BMS with a higher than spec voltage.

You can effectively take any charger that is regulated to have a max current. Eg cellphone charger that outputs 5V 2A. Connect the 5V directly to the low cell and charge it. Keep a multimeter connected and make sure you do not pass 3.6V. Babysit it till done. Once done connect the BMS again and see if your battery is in balanced.

I like the idea off @iiznh

When into DIY one can even build a CV adjustable regulator using a LM338 that cannot provide more than 5A. Adjust the voltage to get a current below 5A and charge the lower cell. At this rate a lot of time is saved than to discharge other cells.

The same LM338 can be used as a CC regulator where you can use a power supply over 7V and charge the lower cell. This way you once again prevent more than 5A charging current.

Here is another idea for charging the cell without the BMS connected.

Connect a single solar panel directly to the low cell and place in the sun, since it is essentially a short circuit it will pull the voltage down to the cell voltage but you will get close to the short circuit current flowing to charge the battery (depending on the light intensity). It should be a current limited charger, but you will have to babysit the process to make sure that your voltage does not rise above 3.6V and you overcharge the cell.

Voltage represents a potential, when you pull enough current the potential will drop, so do not worry about the fact that your panel is 48V. For a solar panel that is Isc. (Add a blocking diode if you panel is going to be shaded)

1 hour ago, stok said:

you do get 1s lifepo4 battery protection circuit boards in a variety of amp sizes. and they are dirt cheap less than R50

now if you attach such a 1s board to the cell and the solar panel to the board, you now have overcharge protection, go and enjoy your life!

yea i was making do with what i had on hand

a 1s charging board is probably the best option and least effort, 

as you can leave it unattended 

i don;'tknow so nicely about the protection boards though

as they work on the same pricniple as the balance function of the bms, ie adding resistance to the cell that is peaking

 

they don't however cut the charging altogether as the bms does

 

ie most likely they would be in the 30-50ma range and would only prevent overcharging if you were charging with same 30-50ma

 

please correct away if i have this wrong

 

you need voltage limitation too, hence why balancing is normally done with a bench power supply

if however you have a 1s charging board that will switch off charging once correct voltage is reached , bob's your uncle

or like someone suggested use any dc power supply that is close to the cell voltage and baby sit the one cell

 

my imbalance was in the order of 30% soc and 90%soc and and low output options was just gonna take to long

we were in stage 6 and i wanted it done

 

 

connecting parralel wasnt gonna cut it flat voltage curve

i would have had to

 

parralel leave to balance 

connect series to charge to have some cells peak

parralel - flat part of curve again

series charge again then parralel

riinse repeat until done

 

was considering buying a bench power supply (would be one time use, and wait for it to arrive delay -opted out)

 

 

1 hour ago, stok said:

I use a 1.2 Ohm 50w resistor and a 3.9Ohm 50w for this purposes(in bad cases I also use a small cc/cv buck boost module capable of 5Amps to bring a low cell up to pack average) you can get a good top balance if you have a accurate DMM which can at least display 3 decimal numbers, so that you can have a one millivolt resolution, its a boring process. the best way will always be to parallel all the cells and keep them at 3.5v for a day or two to ensure complete saturation of the cells.

 

yea i like the buck converter idea as that limits volatge anything that helps reduce the risk

15 hours ago, iiznh said:

I = V / R   So that means with a single cell 3.3V / 20 Ohm = 0.165A ... pretty much the same or less than the daly BMS.

You will burn off P = V * I 

3.3V * 0.165A = 0.544 Watts so connected for 2 hours you will remove about 1Wh from a single cell

a daly bms does 0.030a

 

i have no idea what the rsistance of a 50w halgen bulb is 

we can deduce that if it pulls 50w at 12v it is using 4.16ah  there is a reason we stopped using these they burn a lot of power for the light they produce

50(w) /12(v) = 4.16a ieyou can shed 4% of your 100ah cell in 1hour