This question arrises more and more often and i didn't find anything on my first search (maybe i missed something).
There is some discussion about how to initially balance new (or old stock) LiFePO-cells the 'right' way. I found the following suggestions:
1. parallel top balancing
Connecting all cells in paralell (first make sure the voltage difference of the cells is low enough - say less than 50 mV) and charge them up to 3.45V (or 3.5 or even 3.65V) until the load currents falls below a certain threshold (say 1/20 C). Some say 3.45V is high enough (= 100% SoC) others tend to go up to 3.65 V, which may be already too high when current is low so that the cells starts suffering. There are also some shortcomings here when doing parallel balancing - so that it is not really recommended.
2. seriel top balancing
connect all cells in series, add a suitable BMS and charge the whole pack up to 100% SoC (i.e. 3.45V/cell or whatever you set as your top charge level in the BMS). At the end of charge manually test each single cells voltage with a good and precise DMM (at least 3 decimals, better 4). Make sure all cells reach the same voltage level by trickle charging each cell separately if necessary).
3. bottom balancing (in series or parallel)
discharge all cells (in parallel or series) to their low-voltage cutoff (say 2.5V or 2.7 V will do the same job). Start with a high load (high discharge current of up to 1C) and repeat (after a little recovery rest) the procedure with a lower load (say 1/20C) to finally discharge to 2.5 or 2.7V/cell. Check each cells voltage with a good DMM and trim each cell down to the same level.
Discussion:
If found the opinion that procedure 2. (seriel top balancing) should be the preferred method.
Method 1. in most cases will take much longer as it needs very high current or will take many hours/days to fully charge all cells. It also needs reassembling the cells for final use in series and you have the effort of compressing the cells twice.
Method 3. is said to be less accurate as the lower cell voltages dont represent the SoC as exactly as the upper level will do. I am not sure about that point as the voltage/SoC-curve in both areas has a kind of knee - so both areas could be used fo leveling/balancing cells to a certain SoC-niveau.
Method 2. has the opportunity to put the pack in its final state for use and you just have to compress it once. Its also much faster than the parallel balancing, as it needs less current and can be done with most better laboritory power supplies capable of supplying the voltage you need for the pack.
A separate point maybe the correct way of compressing the cells for their initial use/charge and the 'right' pressure. Some say you need to apply around 300 kg of pressure to the pack +/- 10%. Too much is bad but too less is also bad. Some don't compress their pack at all with no visible problems, some get bad bulging cells.
Eager to hear your opinion and experience here
There is some discussion about how to initially balance new (or old stock) LiFePO-cells the 'right' way. I found the following suggestions:
1. parallel top balancing
Connecting all cells in paralell (first make sure the voltage difference of the cells is low enough - say less than 50 mV) and charge them up to 3.45V (or 3.5 or even 3.65V) until the load currents falls below a certain threshold (say 1/20 C). Some say 3.45V is high enough (= 100% SoC) others tend to go up to 3.65 V, which may be already too high when current is low so that the cells starts suffering. There are also some shortcomings here when doing parallel balancing - so that it is not really recommended.
2. seriel top balancing
connect all cells in series, add a suitable BMS and charge the whole pack up to 100% SoC (i.e. 3.45V/cell or whatever you set as your top charge level in the BMS). At the end of charge manually test each single cells voltage with a good and precise DMM (at least 3 decimals, better 4). Make sure all cells reach the same voltage level by trickle charging each cell separately if necessary).
3. bottom balancing (in series or parallel)
discharge all cells (in parallel or series) to their low-voltage cutoff (say 2.5V or 2.7 V will do the same job). Start with a high load (high discharge current of up to 1C) and repeat (after a little recovery rest) the procedure with a lower load (say 1/20C) to finally discharge to 2.5 or 2.7V/cell. Check each cells voltage with a good DMM and trim each cell down to the same level.
Discussion:
If found the opinion that procedure 2. (seriel top balancing) should be the preferred method.
Method 1. in most cases will take much longer as it needs very high current or will take many hours/days to fully charge all cells. It also needs reassembling the cells for final use in series and you have the effort of compressing the cells twice.
Method 3. is said to be less accurate as the lower cell voltages dont represent the SoC as exactly as the upper level will do. I am not sure about that point as the voltage/SoC-curve in both areas has a kind of knee - so both areas could be used fo leveling/balancing cells to a certain SoC-niveau.
Method 2. has the opportunity to put the pack in its final state for use and you just have to compress it once. Its also much faster than the parallel balancing, as it needs less current and can be done with most better laboritory power supplies capable of supplying the voltage you need for the pack.
A separate point maybe the correct way of compressing the cells for their initial use/charge and the 'right' pressure. Some say you need to apply around 300 kg of pressure to the pack +/- 10%. Too much is bad but too less is also bad. Some don't compress their pack at all with no visible problems, some get bad bulging cells.
Eager to hear your opinion and experience here