As I mentioned I first charged these cells, INDIVIDUALLY, to 3.75VPC and X current taper. Today I would stop at 3.65V. The bench-top power supply allows you to set the voltage to 3.75 and let the cell become "full" at 3.75VPC. I held the voltage at 3.75V and allowed the current to tail off to exactly 20A then stopped charging and moved onto the next cell.
WARNING: It is fairly important that you top each cell up prior to wiring them in parallel, unless you have a large power supply or do them in series using a large inverter/charger. They should ideally be very close in SOC/balance before placing the parallel jumpers in place. This helps to get most of the balancing done.
Within seconds of wiring these in parallel only 0.59A was moving between cells which means my balance to 3.75VPC was pretty close.
It is really a net myth that simply wiring the cells in parallel will serve to get them balanced. While they may be at the same "voltage" they may not be at the same SOC. This is because once voltages become equal or at parity, the current has little to no voltage to move any meaningful current between the cells. You need to actively charge in parallel above the resting voltage in order to get the cells to balance and only when the current has declined to next to nil will you know the cells are all at the same SOC. With a small power supply this can take lots of time.
The resting parallel conundrum:
As the voltages converge, after being wired in parallel, the movement of current between cells slows to a crawl, it is simple Ohm's law. We are talking 0.0001A level current movements and at this rate "balancing" takes forever.
This I why I top balance the cells until my bench top power supply is flickering between 0.00A and 0.1A of current. Once this is achieved I will discharge them to about 50-60% SOC, in parallel, and if I have the time they can sit there for long periods. You don't want to charge to 100% SOC then let them sit there as this can serve to degrade the cells. While this level of degradation is slow it also depends upon the cell temp. If your cells are at 90F in Florida, and you let them sit at 100% SOC, they will degrade faster than sitting at 100% SOC at 50F in Maine. Bottom line, don't let your batteries linger at or near 100% SOC for very long.
TIP: Never trust the volt meter on the bench top power supply as there may be voltage drop or inaccuracies between the supply & actual battery terminals. ALWAYS measure the actual battery terminal voltage using a good quality DVM, when top balancing.
Winston cells should not be top balanced or have the voltage pushed beyond 3.800VPC, this despite what the Chinese manual tells you. In reality, with four years of in-depth experimentation and many hundreds of hours of behavior observation, I personally would not advise pushing them to any more than 3.65V for a top balance as there is no need to do this for a fractional C system.
CALB cells also should not be pushed beyond 3.650VPC, when top balancing. 3.800V is the absolute max voltage you want to push any of these prismatic cells to in a closely monitored & attended top balance. During carefully monitored and attended top balancing is the ONLY time your LFP cells should ever be charged above 14.2V / 3.55VPC.
Remember, when I wired these cells in parallel I now had a 1600Ah 3.2V bank. Even the last 0.05V, from 3.75VPC to 3.80VPC requires many, many hours to get there.
Please note that these are Winston cells and Winston recommended the 3.8V top balance voltage. With CALB or other cell brands 3.6V- 3.65V is often the max you would want to push to. Please consult with the manufacturer of your cells for max top-balance voltages. Even with Winston a 3.65V max should work fine, but Winston insisted on 3.8V. I don't actually agree with it, for fractional "C" use and lower charge voltages, it has however worked very well on this particular bank.