LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[Maine Sail]

Originally Posted by Jd1

I am gradually making progress with my lithium install. The batteries are in the boat and set up for initial testing. I have set up the House Power BMS and the little cell boards. Yesterday I finished a top balancing. When I left, all four cells were in shunting mode indicating they were above 3.65V but below 3.8V. This morning I was expecting to see all batteries around 3.65 or so volts.
Instead, I have the following result:
3.596V
3.576V
3.599V
and
3.383V !!!!!
The first three cells are in the range I was expecting but what is going on with cell #4?
This is not the first time that cell #4 behaved odd - it was doing a similar thing while I was in the process of balancing but I wanted to verify this behaviour after balancing - which I now have.
I am unable to do a full load test at this time as I am still waiting for my order of battery cables to come in.
Another bit of trivia - the cell that is acting up has marks on the case left by allthread rod as if it was previously installed. I was assured by Balqon that the cells are all new but that some cells out of every batch are tested and the marks might be from that. IMHO the marks are wayy too deep for that unless the have some gorilla of a guy strapping them into the test setup ... but that is only my opinion.
Question: Is the behaviour I am seeing an indication of battery age or previous use or should I ignore this issue until I have my cabling set up for a full capacity test ?
Has anyone else run into this kind of issue ?
I will be doing a full capacity test one way or another but I am getting concerned and wanted to see if other people have run into this kind of behaviour.

Thanks

You left these unattended and on charge at 3.65V per cell while actively shunting.......?

How exactly did you top balance?

Were the individual cells charged to 3.XV+ and X.x tail current before wiring in parallel?

If so what were these parameters?

Did you parallel top balance?

What voltage did you push the parallel cells to?

What was the tail current you stopped at during top balance?

What was the max parallel top balance voltage?

How long were cells wired in parallel before pushing the top balance voltage to 3.75 - 3.8V?

Was any baseline capacity done at all before the installation?

Does cell number four also tap out early on LV?

[Jd1]

Originally Posted by Maine Sail

You left these unattended and on charge at 3.65V per cell while actively shunting.......?

No, I charged under supervision until all cells were shunting and then disconnected the charger

How exactly did you top balance?

A combination of single cell charging, some cell draining and some multi cell charging to get them all into the same ballpark for cell voltages. Let them sit for a day to rest, charge the pack and top off individual cells as needed to get all cells into shunting territory. Let sit to finish shunting plus an overnight rest. Top up the next day to get all cells to eventually shunt. Most of this happened at 20A charge current. I did a final topping off at about two amp for about half an hour. Disconnected charger and let the batteries finish shunting and rest.
I have not yet moved the cell level board to another cell to verify that it isn't an issue with that circuit (which I doubt).

Were the individual cells charged to 3.XV+ and X.x tail current before wiring in parallel?

I did not do a tail current charge. The cells are 700 AHr and I figured 20A was insignificant so virtually all charging was based on voltage with shunting and resting. There was a minor bit of reduced rate charge but again that was probably insignificant in the overall picture.
No significant charging was done while the cells were in a parallel configuration.

If so what were these parameters?

The cells were mostly just brought up to the 3.65 to 3.8V range and the shunting took them down to just below 3.65 and resting might have dropped them a bit further here and there hence the multiple charges to shunting and subsequent rests.

Did you parallel top balance?

Initially the cells were set up in parallel for 3 or 4 days but no significant charging was done in that configuration. They were mostly allowed just to equalize their voltages.

What voltage did you push the parallel cells to?

didn't

What was the tail current you stopped at during top balance?

Let's call it 2 amps but it wasn't a proper tail current.

What was the max parallel top balance voltage?

N/A

How long were cells wired in parallel before pushing the top balance voltage to 3.75 - 3.8V?

3 to 4 days

Was any baseline capacity done at all before the installation?

No, I was unable to do that as I had no SOC monitor nor BMS (to automatically turn the load off) nor proper cabling.

Does cell number four also tap out early on LV?

Don't know this until I can do a capacity test when cabling gets here, probably Thursday.

Thanks for your input!

[T1 Terry]

Originally Posted by Jd1

I am gradually making progress with my lithium install. The batteries are in the boat and set up for initial testing. I have set up the House Power BMS and the little cell boards. Yesterday I finished a top balancing. When I left, all four cells were in shunting mode indicating they were above 3.65V but below 3.8V. This morning I was expecting to see all batteries around 3.65 or so volts.
Instead, I have the following result:
3.596V
3.576V
3.599V
and
3.383V !!!!!
The first three cells are in the range I was expecting but what is going on with cell #4?
This is not the first time that cell #4 behaved odd - it was doing a similar thing while I was in the process of balancing but I wanted to verify this behaviour after balancing - which I now have.
I am unable to do a full load test at this time as I am still waiting for my order of battery cables to come in.
Another bit of trivia - the cell that is acting up has marks on the case left by allthread rod as if it was previously installed. I was assured by Balqon that the cells are all new but that some cells out of every batch are tested and the marks might be from that. IMHO the marks are wayy too deep for that unless the have some gorilla of a guy strapping them into the test setup ... but that is only my opinion.
Question: Is the behaviour I am seeing an indication of battery age or previous use or should I ignore this issue until I have my cabling set up for a full capacity test ?
Has anyone else run into this kind of issue ?
I will be doing a full capacity test one way or another but I am getting concerned and wanted to see if other people have run into this kind of behaviour.

Thanks

If you are charging each cell separately, does #4 cell always drop low when left sitting over night? Have you tried removing the shunt board from #4 cell, recharging to 3.6v and letting it sit over night to see if it still drops down as low?
The symptoms you are describing are all too common when using these shunting boards, no matter who makes them, this has lead to quite number of people removing all their shunting boards and simply controlled voltage charging and many have discovered their problems simply went away. Others had already had cells dragged to 0V resulting in cell damage, but the cells still came good once the shunt boards were removed, just down on capacity.

T1 Terry

[Jd1]

Originally Posted by T1 Terry

If you are charging each cell separately, does #4 cell always drop low when left sitting over night? Have you tried removing the shunt board from #4 cell, recharging to 3.6v and letting it sit over night to see if it still drops down as low?
The symptoms you are describing are all too common when using these shunting boards, no matter who makes them, this has lead to quite number of people removing all their shunting boards and simply controlled voltage charging and many have discovered their problems simply went away. Others had already had cells dragged to 0V resulting in cell damage, but the cells still came good once the shunt boards were removed, just down on capacity.

T1 Terry

Yes, the cell has behaved like this earlier.
My next test will be to exchange the cell boards of another cell with cell #4's board to test out exactly that scenario. The cell board does not appear to have an elevated temperature (indicating failure) but one never knows for sure until this is verified.

[Maine Sail]

What T1 said in paragraph #1 plus...

It does not sound like you actually balanced the pack.. Without a tail current for each cell at 3.XX V it is tough to say where that cell is. If parallel balancing they need to be charged in parallel, above resting V, not just left to sit in parallel. The best time to let them sit is after pushing them to the 3.8V top balance point..

Once all cells have been brought to the same point, for Winston I use 3.75V and the same tail for each cell, then wire them in parallel and push to 3.8V and allow the current to taper. Remember this is now a 2800Ah pack so even the difference between 3.75 and 3.8V will take considerably time.. Watch cell temps while doing so..

This is why getting each cell as close to identical as you can before wiring in parallel and pushing to 3.8V is pretty critical. Letting them sit in parallel brings voltage parody but mA level currents are still moving and with a 700Ah pack this can take time due to the voltage....

I highly doubt this is a cell board issue and shunt balancing that wide of a spread, 0.216V, is a loooooong tedious process best left to manual then parallel top balancing. Shunt balancing is intended for minor variances. You have large variances...

It sounds to me like the top balance simply was not adequate for the size bank. It could be a bad cell but you need to charge each cell to a top point, which is exactly the same, then let them all sit, and measure voltage of each cell individually.

My Winston cells shipped to me out of balance and it was as simple as doing a corrective top balance. 500+ cycle later they still in near perfect balance..

[Jd1]

Ok, I will first exchange cell boards and give that a try and once that has been verified I will bring all cells to the same voltage, hook them up in parallel and do another (parallel) balance. I will then gradually reduce current over a 5 hour time period (the longest I have patience for while watching the entire setup), let them sit overnight to rest while still paralleled, separate the cells and let them rest overnight and finally measure cell voltages again.

[Maine Sail]

Originally Posted by Jd1

Ok, I will first exchange cell boards and give that a try and once that has been verified I will bring all cells to the same voltage, hook them up in parallel and do another (parallel) balance. I will then gradually reduce current over a 5 hour time period (the longest I have patience for while watching the entire setup), let them sit overnight to rest while still paralleled, separate the cells and let them rest overnight and finally measure cell voltages again.

The batteries will reduce the current automatically at X volts and X SOC.......... You don't need to stand there are watch it the entire time just set a timer to check every 30 minutes or so and hit them with a temp gun.

Do not use your power supply to determine cell voltage use your DVM at the battery terminals as the determination....

[evm1024]

Just for a FYI. With he "same" 700 AH cells that were top balanced per Maine Sails advice my cells read 3.339 each with one cell dropping down to 3.338 every now and then. These cells are about 70 AH below a charge to 13.9 v @ 20 amps (3.475 per cell). Very nicely balanced.

I only have a few cycles to 60% SOC on them.

I've not done a full capacity test on them yet. Something I need to do. I would love to have the voltage curve at the lower knee for these cells of mine. (both at a 100 A rate and at a 20 A rate - time....)

[nimblemotors]

It isn't complicated, remove the bms junk, put them in parallel, and charge them all with a constant 3.6v power supply until the current drops to zero.
Now they are top balanced. Disconnect them from the charge and each other and let each cell sit for at least a couple days or even a week or two.
Check the voltages, if they have dropped too much, they are not good.

Then discharge each one down to 2.5 volts, and check their capacity, they should be pretty close to rated capacity. Now they are bottom balanced.

Now you know what you have, if they are good and their capacity.

[T1 Terry]

Originally Posted by nimblemotors

It isn't complicated, remove the bms junk, put them in parallel, and charge them all with a constant 3.6v power supply until the current drops to zero.
Now they are top balanced. Disconnect them from the charge and each other and let each cell sit for at least a couple days or even a week or two.
Check the voltages, if they have dropped too much, they are not good.

Then discharge each one down to 2.5 volts, and check their capacity, they should be pretty close to rated capacity. Now they are bottom balanced.

Now you know what you have, if they are good and their capacity.

This sounds like an electric vehicle battery balancing regime, of no value if you don't intend to completely drain you cells each time you use them. Electric vehicle users often drag their cells down to the last bit of capacity, house battery owners do not, or at least should not. The capacity test will only match the manufacturers figures if tested by the same method as the manufacturer used, 0.5C current draw until the cell measures 2.5v under the 0.5 load. For a 700Ah cell that would require a 350 amp load that could maintain that load for 14 hrs, not likely to be something you have on hand is it.
Even at 3.6v you will damage the cells if you hold them at that voltage for very long so don't do it if you want long battery life. If you are using a 20 amp charger that equals a 0.02C load, paralleled as 4 x 700Ah cells equals 2800Ah, charge that pack with a 20 amp rate charger and you will be charging at a 0.007 charge rate, trickle charging to 3.6v kills lithium cells faster than shooting them, don't do it, if you really want to use this method, restrict the top end voltage to 3.5v and charge till the current flow drops, then a short charge to 3.6v on each individual cell and stop as soon as it reaches 3.6v. If the cells all stay above 3.5v for 12 hrs, the cells are fine.
Most definitely remove the shunt modules to do this balance charging or you won't know if it's the cells still accepting charge or the balance modules consuming the charge. Also, you will not know if the cell is not holding charge or if the shunt module is consuming the cell capacity.
As far as cell balance in a 4 cell in series string, you won't see an imbalance until the cells are full, anything above 3.45v, no balance board made can shunt 20 amps so they will be as much use as an ass pocket in a singlet in trying to top balance a battery pack while charging, so why add complication to a system with something that isn't capable of doing the job you think you fitted it for in the first place. 12v Li batteries stay within 100mA of each other all the way till the last bit of charging when the cell goes over 3.45v, with or without balance boards, you don't need to balance within a range this tight anyway, so why try to fix something that isn't broke by using something that can't do the job intended anyway?

T1 Terry

[Jd1]

Originally Posted by T1 Terry

If the cells all stay above 3.5v for 12 hrs, the cells are fine.
T1 Terry

Added about 7 Ah to the cell that is showing strange results which got me to 3.8V. I disconnected the cell board and watched the voltage fall rather rapidly - I will check again tomorrow but I think I was well into the 3.5V territory within half an hour of stopping the charge current. I will verify things tomorrow but it definitely looks like the cell is dropping it's charge. The remaining cells are all above 3.5V which I am quite happy about.
I also disconnected the Cellog 8 just to be certain that there is no possible way that electrons could sneak from one end of the cell to the other end

[T1 Terry]

Originally Posted by Jd1

Added about 7 Ah to the cell that is showing strange results which got me to 3.8V. I disconnected the cell board and watched the voltage fall rather rapidly - I will check again tomorrow but I think I was well into the 3.5V territory within half an hour of stopping the charge current. I will verify things tomorrow but it definitely looks like the cell is dropping it's charge. The remaining cells are all above 3.5V which I am quite happy about.
I also disconnected the Cellog 8 just to be certain that there is no possible way that electrons could sneak from one end of the cell to the other end

Do the recharge a second time if it falls below 3.5v after a few hrs, I suspect that cell has higher internal resistance so the voltage climbs much faster as it reaches the end of charge. If this cell continues to do the same thing once in use, without the balance boards attached I'd looking for a claim

T1 Terry

[nimblemotors]

This sounds like an electric vehicle battery balancing regime, of no value if you don't intend to completely drain you cells each time you use them.

No value to know if the cell is bad, or its capacity, uh huh,
And the 3.6v will damage the cells.

[T1 Terry]

Originally Posted by nimblemotors

No value to know if the cell is bad, or its capacity, uh huh,
And the 3.6v will damage the cells.

This is a quoted post from an admin member of the Australian Electric Vehicle Association forum. These guys eat and sleep this stuff and the combined qualifications of these guys would make your head spin

you can read more from where it came from here Upgrade to Lithiums? - The Australian Electric Vehicle Asn - Page 2

I once used a 0.35A LED constant current power supply to balance out a set of 16 90Ah cells, all with BMS modules on them that shunted at 3.65v
I left it on over a weekend

after the weekend, all cells were 3.65v, all bulged, all lost capacity, all had higher internal resistance.
please don't float them

you already have a single cell charger - just leave it attached to the lowest cell until it is full, then move to the next one

what current is the single cell charger?

A short burst to 3.6v will not harm the cells, holding them at 3.6v or higher will wreck them

T1 Terry

[bill good]

Terry, I am replacing my house batteries with LIFEPO4S. What is a "short burst" @ 3.6V? I note all the CV portions of the charge is approx 2hrs no matter on the capacity/brand & was going to use the overall voltage of 14.6V for manual top equilizing when/if needed. The 2 hr max charging time would always be active both with normal daily 14v max & the periodic 14.6V. Solar system max 30Amp 300 AmpHr CALB cells. Cell shunts from EV-POWER. Aim to operate 14-12.6V range. Any problem seen with this setup??

Regards Bill