fatty wrote:if you top-balance at 4.2V and discharge to 4.1V, the cells will necessarily be closer than if you top-balance at 4.1V and discharge to 4.0V. Not practically relevant, just refuting the claim that top-balancing at 4.2V would "throw the pack out of balance"
But it does do that!
In my case irrelevant, since I would not want the packs to ever get that high anyway.
> Sure, middle-balancing minimizes how far
out of balance the pack gets at each end of SoC
By avoiding both shoulder areas, you may never see any significantly imbalanced state at all. . .
> by setting LVC and HVC by the lowest capacity cell, the other cells will necessarily be kept in a healthier range, since only the lowest capacity cell will experience those LV and HV extremes
The goal afaic is no cells approach any such extremes.
> But is this meaningfully advantageous at the pack level? You're just hitting the weakest cell the hardest at every charge and
That inevitably happens no matter what, in the sense of "hardest" meaning compared to every other cell.
So, the healthiest strategy for every cell, is to make sure to coddle the weakest links; that ensures the rest are coddled too.
Just giving up (another) 10% capacity utilization can triple cycle lifespan or more.
> Cell capacities will also drift over time, necessitating balancing, though it may be less frequent.
Since the weakest cells are under greatest stress, they wear fastest, only QA anomalies would result in any "switching places."
> And I guess if you artificially constrain capacity, then capacity variance in absolute mAh will be reduced: say a 2s pack of 8Ah cells with 1% capacity variance have capacities of 8039 and 7973mAh, delta 66mAh. Charging to 4.1V (95%) gives 7637mAh and 7574mAh, delta 63mAh.
Not "artificial", capacity and SoC% are not objective attributes, tgey are **defined** arbitrarily by the test parameters selected.
A data sheet saying a cell has 20Ah is meaningless without the company's definition of Full and Empty
if those parameters are followed for every cycle in normal daily usage, the user will need to replace the pack 5x or 10x more frequently - that is not a coincidence!
By defining capacity more realistically, using **your** gentler / healthier definitions of Full and Empty to achieve decent longevity and thus value for money
you aren't really "sacrificing capacity", you are just acknowledging what the realistic capacity **actually is** , there is no reason to acknowledge the datasheet values as anything other than marketing-driven fantasies - those are what is "artificial"!!
john61ct wrote: ↑
Apr 03 2021 11:11pm
By pushing your charge voltage out past 4.05 - 4.10V and top balancing out there you are indeed **causing** the imbalances that now require frequent "fixing" and reducing longevity.
> I'm not sure about this. If you mean charging up to 4.2V reduces longevity first
, then I understand and agree, since that loss of capacity will not be uniform and the cells will be increasingly mismatched.
> I don't see how termination voltage alone (independent of capacity) can directly cause imbalance
Imbalances never revealed, in effect do not exist functionally.
john61ct wrote: ↑
Apr 03 2021 11:11pm
If you want to maximize range, you should bottom balance, but going down to a lower LVC drastically shortens lifespan, much more than charging too high
> I also don't follow this. If we agree that the high and low "shoulders" (the first and last 5-10% of capacity, respectively) are equally harmful to lifespan (and thus balance), then capacity is capacity and it shouldn't matter if you top-balance and use LVC on the weakest cell, or bottom-balance and use HVC on the weakest cell. The capacity would be equivalent.
No, the avoidance of the top shoulder is a much smaller range and much less impact tgan at the bottom.
Bottom balancing only is better for range when using **pack level** LVC, because you have greater peace of mind and don't need to be as conservative to prevent immediate gross damage.
Different from lifespan optimization, which requires a judgment call, giving up significant range, pack-level vs cell-level implementation is irrelevant.
> top-balance would give a higher average voltage and thus higher pack Wh than bottom-balancing, and so I would expect that top-balancing would actually maximize range, not bottom-balancing.
Wh are not nearly as accurate for battery capacity discussions, Ah match up with actual observed behaviour.
And "higher average" voltage is irrelevant, tge whole point is, voltage deltas represent much lower energy deltas once you go past 4.0V for EV chemistries, or 3.3V for LFP.
A very high % if the coulomb flow is not getting stored, just increasing useless even harmful chemical activity and producing waste heat.
That is why the standard for measuring energy deltas is not voltage nor charging Ah, but **discharged** Ah from an at-rest voltage down to an LVC
at a very slow current rate to give a bigger number for the nameplate and data sheet.
The SoC point at which you balance does not affect pack capacity at all, if HVC and LVC are measured at the per-cell level.