diy solar

diy solar

LFP cell behavior

bdserv

New Member
Joined
Apr 5, 2021
Messages
70
I fabricated a 48 volt 16S LiFePO4 battery pack for my golf cart using Coslight 150 AH cells. I carefully top balanced the cells before final assembly using the parallel method and bringing them all slowly up to 3.65 volts [all 16 cells at the same time]. I actually did this twice; once after receiving the cells and again after a short discharge test. I stopped the top balance when the total bank current had dropped below 1 amp.

After the top balance I noticed that the cells had varying self discharge rates and within a few days or so the highest/lowest cells were 50 mV or more apart - some were still over 3.500 volts and some were in the 3.450 or less range.

During a normal charge they do not stay together - the highest cell will reach 3.55 volts [my BMS over voltage trip point] with other cells still in the 3.46 volt range [charging at about 0.1C].

During normal discharge in the golf cart they will quickly "come together", and when they are below 3.4 volts per cell or so they are all within less than 10 mV of each other. They will remain this close together even when the cart has not been used for a day or more.

Working with these cells is new to me, but my thoughts on this are that it might be normal for these cells to charge and self-discharge at vastly different rates when near the top of charge curve, and since the stored energy above 3.4 volts or so is minimal there isn't really anything to be worried about here.

I would appreciate hearing from others with experience with these types of cells on this.
 
Your cell differences don't sound bad. What BMS are you using?
It is an inexpensive Chinese BMS from LLT. However, I did check the cell voltages reported by the bms against a Fluke dvm and they were within a mV or two. The bms controls charge directly, and discharge using a relay configuration.
 
I would think bottom balancing would be better in a golf cart style use. I'm guessing you charge by a plugin.
 
It is an inexpensive Chinese BMS from LLT. However, I did check the cell voltages reported by the bms against a Fluke dvm and they were within a mV or two. The bms controls charge directly, and discharge using a relay configuration.

Does the BMS perform cell balancing?
 
Above 3.43v is surface capacitance charge range. It is normal.

It is like a super cap mostly caused by negative graphite anode layer when cell is fully charged to voltage above 3.5v.

For LFP, the maximum surface charge energy is only equivalent to about 0.01% of cell total capacity when fully charged to low taper off absorb current at 3.65v. On its own it can take a couple of days to bleed off, but like self leakage, its rate of bleed is variable to given cell. You can bleed it off in a minute or two of discharge. 0.01% of 150 AH is only 15 mA/H or 900 mA/minutes. It is exponential decay and takes a little time for charge to diffuse out. A power resistor can be used to bleed off surface charge in about a minute or two. 1 to 3 ohm load resistor.

If you have a BMS with operating balancer of 100 mA per cell it should fully bleed off surface charge in less then an hour after charging completes.

After you bleed off most of surface charge you should have about 3.45v open circuit voltage for fully charged cell.
 
Last edited:
I fabricated a 48 volt 16S LiFePO4 battery pack for my golf cart using Coslight 150 AH cells. I carefully top balanced the cells before final assembly using the parallel method and bringing them all slowly up to 3.65 volts [all 16 cells at the same time]. I actually did this twice; once after receiving the cells and again after a short discharge test. I stopped the top balance when the total bank current had dropped below 1 amp.

After the top balance I noticed that the cells had varying self discharge rates and within a few days or so the highest/lowest cells were 50 mV or more apart - some were still over 3.500 volts and some were in the 3.450 or less range.

During a normal charge they do not stay together - the highest cell will reach 3.55 volts [my BMS over voltage trip point] with other cells still in the 3.46 volt range [charging at about 0.1C].

During normal discharge in the golf cart they will quickly "come together", and when they are below 3.4 volts per cell or so they are all within less than 10 mV of each other. They will remain this close together even when the cart has not been used for a day or more.

Working with these cells is new to me, but my thoughts on this are that it might be normal for these cells to charge and self-discharge at vastly different rates when near the top of charge curve, and since the stored energy above 3.4 volts or so is minimal there isn't really anything to be worried about here.

I would appreciate hearing from others with experience with these types of cells on this.
Your results sound pretty good to me. The "weakest cells" will charge to a higher voltage more quickly than others, as it becomes nearly full - and your BMS is behaving correctly, slowly balancing the "bigger" cells upwards to match that endpoint Voltage reading. On discharge, with the same discharge current taken from each cell (with roughly equal power loss as a result), the "weakest cells" will fall off a bit more quickly.

I would not, however, prefer to charge my cells to their absolute maximum on every single night. If you were to set your BMS maximum cell Voltage to only 3.5V, or maybe even less, you might enjoy slightly better total lifespan. Maybe once per month, reset your BMS voltage limit to 3.6 per cell and top them off, then drop it back down. Unlike Lead-Acid batteries (which always last longest when stored at 100% SOC) LFP batteries last longer if you don't keep them near 100% all the time.
 
I would not, however, prefer to charge my cells to their absolute maximum on every single night. If you were to set your BMS maximum cell Voltage to only 3.5V, or maybe even less, you might enjoy slightly better total lifespan. Maybe once per month, reset your BMS voltage limit to 3.6 per cell and top them off, then drop it back down. Unlike Lead-Acid batteries (which always last longest when stored at 100% SOC) LFP batteries last longer if you don't keep them near 100% all the time.

Yes, this is what I am doing - minimizing charge cycles for the reasons you mentioned. Looks like I can get around 4 18-hole rounds before I need to recharge.
 
If you were to set your BMS maximum cell Voltage to only 3.5V, or maybe even less, you might enjoy slightly better total lifespan.
the highest cell will reach 3.55 volts [my BMS over voltage trip point]

A BMS should not be used as a charge controller, it is designed to be a safety disconnect.
You should set the BMS hvd to at least 3.6V (approaching the safety limits of the cell) so your charge controller (get it? it controls the charging) can charge to your settings without continually tripping the BMS safety mechanism.
 
A BMS should not be used as a charge controller, it is designed to be a safety disconnect.
You should set the BMS hvd to at least 3.6V (approaching the safety limits of the cell) so your charge controller (get it? it controls the charging) can charge to your settings without continually tripping the BMS safety mechanism.

I have heard this before - is your reason because the BMS would be a single-point failure when relied upon as a charge controller?

When using the charger as the primary "charge controller" how do you set your output voltage when the individual cell voltages can differ from one charge cycle to another? In other words, how do you guarantee that the charger always shuts down before any individual cell trips the BMS hvd [especially as the cells age and characteristics change]?
 
With regard to managing individual cell voltages, only the BMS can do that. But you can set the charging devices to charge to a lower voltage than the BMS cuts off at. That assumes that your charging devices are programmable. Not all of them are. My converter is not, but I don't use it that much so I'm not worried about it.

Set your solar charge controller to charge to 14.2v and the BMS cutoff at 14.6v. That's plenty of headroom for one cell to be higher than the others and still allow the other cells to get a full charge.
 
Still waiting on what they're charging with?

Also, waiting for when you come back with, I was out riding and drained the the cells below 2.5v. Think electric vehicle where bottom balancing is more important than top balancing. ?
 
Still waiting on what they're charging with?

Also, waiting for when you come back with, I was out riding and drained the the cells below 2.5v. Think electric vehicle where bottom balancing is more important than top balancing. ?

Presently charging with a 48 volt DPI golf cart charger in LFP mode [it has several modes for FLA as well as LFP]. However, the LFP mode does not allow for output voltage adjustment below 58.4 volts [3.65 per cell in my 16S configuration] as far as I know, so if using the BMS as the charge limiting device becomes a concern I will need to address that.

The BMS will protect against individual cell under voltage so I am not worried about going below 2.5 volts, or even getting anywhere near it.

However, I am closely monitoring the cells as I get some running time on them, both in charge and in discharge. It will be interesting to see how close they stay together as I get closer to the bottom of their SOC, but I also intend to be very conservative as to how far I plan to push that.
 
Presently charging with a 48 volt DPI golf cart charger in LFP mode [it has several modes for FLA as well as LFP]. However, the LFP mode does not allow for output voltage adjustment below 58.4 volts [3.65 per cell in my 16S configuration] as far as I know, so if using the BMS as the charge limiting device becomes a concern I will need to address that.

The BMS will protect against individual cell under voltage so I am not worried about going below 2.5 volts, or even getting anywhere near it.

However, I am closely monitoring the cells as I get some running time on them, both in charge and in discharge. It will be interesting to see how close they stay together as I get closer to the bottom of their SOC, but I also intend to be very conservative as to how far I plan to push that.
Ok, but from many years of reading....what possibly could go wrong with us humans as a protection level and the famous words of "I don't plan on pushing it that far", nobody does until....;)

Just becareful.
 
A BMS should not be used as a charge controller, it is designed to be a safety disconnect.
You should set the BMS hvd to at least 3.6V (approaching the safety limits of the cell) so your charge controller (get it? it controls the charging) can charge to your settings without continually tripping the BMS safety mechanism.
I disagree with your statement. My BMS is designed to control all charge and discharge based on "cell level" voltage.
 
I disagree with your statement. My BMS is designed to control all charge and discharge based on "cell level" voltage.
That's true. However @MisterSandals is saying is correct, that the BMS is not a charge controller. The BMS is a last-ditch defense against something going really bad. All that the BMS is intended to do is stop charging if an individual cell level gets to high, or stop discharging if an individual cell gets too low (and a number of other things, but those two are the primary). That isn't "managing" the charging of cells. A good charger should manage constant voltage up to a set point, then switch to constant current through the absorption phase. If it is an automated charger (like a solar charge controller) it should then lower the voltage to whatever has been set for float. None of this is done by the BMS.
 
That's true. However @MisterSandals is saying is correct, that the BMS is not a charge controller. The BMS is a last-ditch defense against something going really bad. All that the BMS is intended to do is stop charging if an individual cell level gets to high, or stop discharging if an individual cell gets too low (and a number of other things, but those two are the primary). That isn't "managing" the charging of cells. A good charger should manage constant voltage up to a set point, then switch to constant current through the absorption phase. If it is an automated charger (like a solar charge controller) it should then lower the voltage to whatever has been set for float. None of this is done by the BMS.
The thing is it has been a blanket absolute statement made here many times. This information could confuse people that are starting on there learning curve. Besides as long as he uses a BMS that uses relays instead of FET control. And as long as the relays are appropriate, I see no issue with this.

"It is an inexpensive Chinese BMS from LLT. However, I did check the cell voltages reported by the bms against a Fluke dvm and they were within a mV or two. The bms controls charge directly, and discharge using a relay configuration."


I would be more worried about "It is an inexpensive Chinese BMS" You mileage may vary .
 
Back
Top