The first part of this guide can be read here.
What is the most appropriate way to top-balance this battery?
The solution is what can be termed as balancing after 'Follow Charge'.
Follow Charge is basically applying voltage (charging a battery) such that the highest voltage Cell gets charged according to the charge model for LFP outlined in Part 1.
The voltage across the battery is adjusted in such a manner that it follows the cell with highest voltage.
It is obvious that any FOLLOW CHARGED battery will have at least one fully charged cell.
And this fully charged Cell has to have the highest voltage compared to others (serving as reference).
This is the intuition behind 'threshold voltage deviation' balancing parameter that distinguishes lower voltage Cells.
Those who are reading carefully will easily realize an important implication:
Full Bulking/charging voltage should not be immediately used for an initially unbalanced pack.
For example, if you're building a 16S battery and targeting a charging/bulking voltage of 3.5 V/Cell.
Setting 16 × 3.5 = 56 V as your bulking/charging voltage for an unbalanced battery will result in some of your Cells 'peaking' earlier than rest of the pack and triggering BMS Over-Voltage Protection. Start with a lower voltage and increase gradually as balancing progresses.
Once, the battery has been fully Follow Charged, top-balancing can then commence bringing the 'laggers' to charge.
Whether it is achieved using active or passive/bleed balancing is a matter of time and efficiency.
There are two extremely important implications that I'd like to point out.
1. Balancing should take place at the END of a full Follow Charge cycle once the cells are resting and there is no current going in/out of the battery.
One can also reason this another way. Every Cell has some internal resistance and in the presence of current going in/out, voltage across its terminals will deviate from the intrinsic cell voltage. This important implication is extended below.
2. The first implication basically means that 'threshold voltage deviation' balancing parameter increases proportionally to Current (but only to a certain limit).
As current going out/ into battery increases from 0, this 'threshold voltage deviation' also increases proportionally. Even then, once current is no longer 'low enough', balancing shouldn't take place at all.
An example for this would be,
If a battery is balanced to within 0.01 V (say) at rest with no current going in/out.
Then balancing should also not kick in, if cell voltage deviation stays below 0.04 V (say) at some low current charge/draw.
This low current threshold must be strictly determined and balancing shouldn't happen above it regardless of deviation.
This means balance trigger also depends on current besides well known Cell voltage and voltage deviation.
This according to me is the expected behaviour from any BMS geared towards LFP capable of negotiating charging voltages.
The voltage across the battery follow its highest voltage cell.
Individual LFP Cells are charged. Batteries are to be Follow Charged.
By definition, all Cells of a top-balanced battery will achieve full charge at the same time.
And as one may have quickly realized, balancing requirements are the exact same:
Whether your battery is newly built from factory new cells and you're top balancing for the first time.
Or, maintaining top balance on a heavily cycled battery in years of service.
Because LFP has a long cycle life, measures used to prolong lifespan of denser Li chemistries used in mobile applications, such as charging slowly once above 80-85% are unnecessary.
Solar applications with instantaneously varying charging currents that can go lower than cut-off are covered by the modified/improved charging model described in Section 3. After which the cells are made to float at a voltage lower than or equal to resting voltage of a fully charged LFP cell (~3.37 V / Cell).
It is elaborated in more details in this post.
Any older charge controller made for Lead Acid will almost always lack the ability to monitor current and thus doesn't have the capability to implement cut-off.
It then becomes a matter of implementing or rather compromising on LFP charging using charging parameters provided by such a hardware.
The goal always remains battery health by emulating proper LFP charging model as closely as possible.
The motive of this two part guide is to enable readers to realize the physical reasoning behind commonly 'thrown around' numbers & terms for LFP and putting them in a proper context ranging all the way from procuring cells, assembling a battery and top-balancing it, to commissioning it in field regarding solar applications.
4. Top-balancing and maintaining Balance with help of BMS
Suppose you got a battery with N factory-new cells connected in series.What is the most appropriate way to top-balance this battery?
The solution is what can be termed as balancing after 'Follow Charge'.
Follow Charge is basically applying voltage (charging a battery) such that the highest voltage Cell gets charged according to the charge model for LFP outlined in Part 1.
The voltage across the battery is adjusted in such a manner that it follows the cell with highest voltage.
It is obvious that any FOLLOW CHARGED battery will have at least one fully charged cell.
And this fully charged Cell has to have the highest voltage compared to others (serving as reference).
This is the intuition behind 'threshold voltage deviation' balancing parameter that distinguishes lower voltage Cells.
Those who are reading carefully will easily realize an important implication:
Full Bulking/charging voltage should not be immediately used for an initially unbalanced pack.
For example, if you're building a 16S battery and targeting a charging/bulking voltage of 3.5 V/Cell.
Setting 16 × 3.5 = 56 V as your bulking/charging voltage for an unbalanced battery will result in some of your Cells 'peaking' earlier than rest of the pack and triggering BMS Over-Voltage Protection. Start with a lower voltage and increase gradually as balancing progresses.
Once, the battery has been fully Follow Charged, top-balancing can then commence bringing the 'laggers' to charge.
Whether it is achieved using active or passive/bleed balancing is a matter of time and efficiency.
There are two extremely important implications that I'd like to point out.
1. Balancing should take place at the END of a full Follow Charge cycle once the cells are resting and there is no current going in/out of the battery.
One can also reason this another way. Every Cell has some internal resistance and in the presence of current going in/out, voltage across its terminals will deviate from the intrinsic cell voltage. This important implication is extended below.
2. The first implication basically means that 'threshold voltage deviation' balancing parameter increases proportionally to Current (but only to a certain limit).
As current going out/ into battery increases from 0, this 'threshold voltage deviation' also increases proportionally. Even then, once current is no longer 'low enough', balancing shouldn't take place at all.
An example for this would be,
If a battery is balanced to within 0.01 V (say) at rest with no current going in/out.
Then balancing should also not kick in, if cell voltage deviation stays below 0.04 V (say) at some low current charge/draw.
This low current threshold must be strictly determined and balancing shouldn't happen above it regardless of deviation.
This means balance trigger also depends on current besides well known Cell voltage and voltage deviation.
This according to me is the expected behaviour from any BMS geared towards LFP capable of negotiating charging voltages.
The voltage across the battery follow its highest voltage cell.
Individual LFP Cells are charged. Batteries are to be Follow Charged.
By definition, all Cells of a top-balanced battery will achieve full charge at the same time.
And as one may have quickly realized, balancing requirements are the exact same:
Whether your battery is newly built from factory new cells and you're top balancing for the first time.
Or, maintaining top balance on a heavily cycled battery in years of service.
5. Approximating LFP charging on chargers meant for Lead Acid
The standard charging model for LFP Cells is simple: CC followed by a CV phase with a well determined cut-off. This is mostly used for EV.Because LFP has a long cycle life, measures used to prolong lifespan of denser Li chemistries used in mobile applications, such as charging slowly once above 80-85% are unnecessary.
Solar applications with instantaneously varying charging currents that can go lower than cut-off are covered by the modified/improved charging model described in Section 3. After which the cells are made to float at a voltage lower than or equal to resting voltage of a fully charged LFP cell (~3.37 V / Cell).
It is elaborated in more details in this post.
Any older charge controller made for Lead Acid will almost always lack the ability to monitor current and thus doesn't have the capability to implement cut-off.
It then becomes a matter of implementing or rather compromising on LFP charging using charging parameters provided by such a hardware.
The goal always remains battery health by emulating proper LFP charging model as closely as possible.
The motive of this two part guide is to enable readers to realize the physical reasoning behind commonly 'thrown around' numbers & terms for LFP and putting them in a proper context ranging all the way from procuring cells, assembling a battery and top-balancing it, to commissioning it in field regarding solar applications.
