GUIDE to properly Top-Balance and Charge a LFP Battery: Part 1

upnorthandpersonal said:

To implement tail current, you need some kind of communication with your charge controllers to stop putting in power, or a BMS that can limit current.

Click to expand...

Yep, But even in the absence of comms, we can try to closely approximate/approach this with the charging parameters we are provided in our chargers.
We can both agree that it is never wise for BMS to do anything regularly regarding charging. It should be the duty of charger.

Even when it comes to communication, it doesn't have to be full fledged communication.
I think it should be very easy to signal just charge request to controller using single conductor pulled HIGH/LOW. Or two differential pairs for long er distances. I'm bad with comms and signalling but I'm sure creative geniuses could easily figure this out.

upnorthandpersonal said:

or tell me why I should right away involve my battery to power the load when solar (and the charge controller) can continue to power my load, but are told by the BMS to stop.

Click to expand...

You're substituting my understanding with your own. It doesn't have to be like that. The charge current going into the battery can stop and let it switch to float.

upnorthandpersonal said:

When in practice, I have a fluctuating source of power from which I want to capture the most amount as possible, and just stopping at 3.5V (or 3.55V) charging voltage will do just that without any balancing issues (again, five years in with my packs if you want real world experience). Also, again, imagine you're at 3.5V (due to a very quick burst of incoming solar) and you have a cloud come over which makes your current drop. How do you distinguish that from meeting the tail current?

Click to expand...

That is the question I wanted you to ask all this time! There is an extremely elegant approach to this whole "charging at very low C rates" problem.

Case A: When instantaneous charging current is more than tail current:​

Let's assume my equipment can muster 0.25 C charge current.
If I'm aiming for 0.05 C cut off @ 3.65V,
I charge until battery voltage rises to 3.65 V/Cell and hold it there until charging current drops to 0.05 C.

You personally want to target 'safer' voltage like 3.5 V/Cell. The corresponding tail current from the table is ~0.02 C. No need to be absurdly precise.
I charge until battery voltage rises to 3.5 V/Cell and hold it there until charging current drops to ~0.02 C.

Case B: When instantaneous charging current dips to less than tail current:​

Suppose, our equipment can only muster a C-rate lower than 0.05 C. This is specially applicable for DC benchtop power supply charging multiple cells in parallel with smaller current compared to its capacity.

Or that while charging using Case A above, one of your concerns came up.

(due to a very quick burst of incoming solar) and you have a cloud come over which makes your current drop.

Click to expand...

Then the calculation is inverted. Now, we look up from the table/calculate the voltage according to our instantaneous charging current and terminate charging when voltage is reached.
So, if the charging current is 0.02 C, charging is considered complete when voltage rises/reaches 3.5 V/Cell.
If the charging current is 0.05 C, charging is considered complete when voltage reaches 3.65V/Cell and so on.

OR If you don't want voltages beyond 3.5V/Cell, once 3.5 V is reached with 0.05 C, the situation flips to Case A. 3.5 V/Cell is then held until current drops to 0.02 C.

For all C-rates below 0.05 C, there exists a voltage below 3.65 V beyond which charging is considered complete.


Just for self-consistency purposes, think about what will happen if charging current is 0.06 C.
Naturally, Case A is followed.
The battery will be charged with 0.06 C until battery voltage rises to 3.65 V/Cell and held until charging current drops to 0.05 C.

At all time, the charging current can be either more or less than what a user-specifies for charging parameter. In either situations, the charging behaviour is determined. And because, I've described all possible situations that can arise, There is no question of ambiguity or things to guess anywhere.

This according to me should be the hallmark of any information complete, self consistent and more even more importantly physically correct implementation. Implementing this elegant logic for a charger is something I guess is less than few pages of code.z
If You've understood this far, then we can proceed further.

shvm said:

There is an extremely elegant approach

Click to expand...

But how is this in any way easier to understand/implement, especially for newcomers (with the equipment readily available today, and not just Victron and Batrium), compared to just saying: stop charging at 3.5V. Or more specifically (for a 48V system): 'put the charge controller to 56V bulk, no absorb time, and float at 3.37V, and set the balancer start voltage in the BMS to 3.45V per cell'? This works perfectly fine (from experience of a multitude of people here on the forum and elsewhere), is easy to set up even with the cheapest of equipment, and doesn't stress the batteries in any way.

upnorthandpersonal said:

But how is this in any way easier to understand/implement, especially for newcomers (with the equipment readily available today, and not just Victron and Batrium), compared to just saying: stop charging at 3.5V.

Click to expand...

It might not make sense there but it matters for guidelines on top-balancing, maintaining balancing and cycle life in general.

upnorthandpersonal said:

'put the charge controller to 56V bulk, no absorb time, and float at 3.37V, and set the balancer start voltage in the BMS to 3.45V per cell'?

Click to expand...

My Luxpower SNA 5k unit has no option to set absorb time for Lead Acid Charging.

T1=10*T0, minimum 10min, maximum 8hour

What seems to work fine for other users and hardware is not even an option for us. This is a common problem of inverters relying too much on BMS comms.

If I set 3.5 V going by the consensus, I will overcharge my battery. If I set a lower value like 3.45, I am wasting solar and my battery will never get charged.

This works perfectly fine (from experience of a multitude of people here on the forum and elsewhere), is easy to set up even with the cheapest of equipment, and doesn't stress the batteries in any way.

Click to expand...

Saying it doesn't stress the batteries in any way requires data on cycle life not getting affected.

shvm said:

Saying it doesn't stress the batteries in any way requires data on cycle life not getting affected.

Click to expand...

If datasheet tests go from 2.5V to 3.65V with whatever tail current to determine cycle life, and I stay from 3.0V to 3.5V with no tail current nor absorption (i.e., stop charging at 3.5V period), there is less stress on my battery than the default test pattern.

shvm said:

What seems to work fine for other users and hardware is not even an option for us.

Click to expand...

It sounds like you’re onto something but are talking to the wrong people who don’t own the same equipment as you.

shvm said:

My Luxpower SNA 5k unit has no option to set absorb time for Lead Acid Charging.

Click to expand...

Maybe you might consider changing the title of your post to:

GUIDE to properly Top-Balance and Charge a LFP Battery using a Luxpower SNA 5k: Part 1​

shvm said:

My Luxpower SNA 5k unit has no option to set absorb time for Lead Acid Charging.

Click to expand...

I have the new EG4 6000XP and am FORCED to use lead acid settings without an option for absorb time. It IS a problem. I am all ears.
Thank you for your efforts here.

Tomthumb62 said:

It sounds like you’re onto something but are talking to the wrong people who don’t own the same equipment as you.




Maybe you might consider changing the title of your post to:

GUIDE to properly Top-Balance and Charge a LFP Battery using a Luxpower SNA 5k: Part 1​

Click to expand...

I don't know how you think but any recommendation in general always start from top, (generalized) to bottom (specialized).

Anyone can derive the specialized from the generalized but the same is not possible in the opposite direction.
What I guess happened over years is that the specialized model approximating LFP charging using Lead Acid Charger has caught on.

That's exactly what I followed for the first month before stumbling upon and learning about the correct charge model. I expect that to be told first and foremost before any mention of specialized advice.

Even if lead acid charging approximation appears to works somehow, other advice is still mostly wrong because the very foundation is wrong.
Like, floating LFP which can never be more than 3.37V , or top-balancing using DC power supply which should never be set to 3.65V with current dropping to zero.

upnorthandpersonal said:

But how is this in any way easier to understand/implement, especially for newcomers (with the equipment readily available today, and not just Victron and Batrium), compared to just saying: stop charging at 3.5V. Or more specifically (for a 48V system): 'put the charge controller to 56V bulk, no absorb time, and float at 3.37V, and set the balancer start voltage in the BMS to 3.45V per cell'? This works perfectly fine (from experience of a multitude of people here on the forum and elsewhere), is easy to set up even with the cheapest of equipment, and doesn't stress the batteries in any way.

Click to expand...

This 100%

shvm said:

Do you know that LFPs need to be balanced every few months and not at every charge cycle?

Click to expand...

Fun theory. Sure, you could do more balancing every couple months. Or you could do the industry standard and leave the balancer to do it's thing anytime needed, in tiny increments.

shvm said:

1. Correct/Standard charge model for a LFP Cell
pay careful attention to the "0.05C cut off" part.

Click to expand...

Is this 0.05c your entire point? Or 3.65 volts?

The company I work for has worked directly with EVE to establish our charge profile. EVE provided us an extended warranty if we terminated charge at 3.55 volts/cell and 0.05c
There is almost no capacity lost by only charging to 3.55 volts when compared to 3.65 volts.

shvm said:

2. Initial Top-Balancing of a LFP battery​

Combining multiple Cells in series is required to achieve voltages higher than 3.2V.

Click to expand...

False, based on your info above. You are saying 3.65 volts above.
I understand that you mean voltage levels like 12 or 48 volts, but you're the one coming here an insisting you're are the smartest person in the room and that everyone else is an idiot. At least he clear about your facts.

shvm said:

3. Modified/improved charging model for a LFP Cell/Battery​

In section 2 where we were top-balancing 4 EVE 230 Ah Cells, it is obvious that pushing 46 Amps @ 3.65 V according to standard charge model is beyond spec of commonly available benchtop DC power supplies. Most of them are rated for 5/10 Amps max.

What this means is we need to improvise on the standard charging model a little. And this is where things get 'somewhat' complicated and the main source of confusion lies.

And the insight into that comes from determining the fully charged resting voltage for a LFP Cell.
Let's say you take a brand new LFP Cell and charge it exactly like the Standard Charge Model says above.
Can you guess how its Voltage will change with time at rest with no current going in or out?
Naturally it can be observed that it will settle down to a voltage lower than 3.65 V. If you've got a spare Cell and lab equipment lying around, this value is easy to determine.

Click to expand...

Have you tested a cell? Do you know what this newly invented "fully charged resting voltage" might be?

shvm said:

This 'full charge voltage' or FCV which will be obviously <3.65 is the resting voltage of a fully charged LFP Cell. Basically it represents the charging limit beyond which reversibility of LFP chemistry starts diminishing.

Click to expand...

Underline is my added emphasis. That's not the resting voltage.

shvm said:

To keep a cell topped up, standard charging model can be re-initiated once the cell has discharged somewhat and its voltage drops below FCV.

Click to expand...

Voltage isn't a good indicator that the cell is charged and unless you've done lots of research, on your cell voltage curve, it would be much easier to use coulomb counting/SOC to know when to restart a charge cycle.

shvm said:

So, if you don't want to (or are not able to) cut off current as mentioned in the standard charging model in section 1, you can also (in theory) fully charge any LFP Cell by maintaining FCV across its terminals and allowing current to go zero. It will take basically forever but it should also fully charge a LFP Cell.
This is basically the same as saying cut off current is 0 C for charging at FCV for a LFP Cell.

Then it is a simple matter of applying linear regression to find the desired (voltage, current) pair connecting these two extremes. (FCV, 0 C) and (3.65 V, 0.05 C).

For example, Let's tackle the example of top-balancing 4 EVE 230 Ah Cells in parallel using a 10 Amps power supply.
The equivalent cell capacity is 230 × 4 = 920 Ah of which 10 Amps charging current constitutes ~0.011 C rate.

The charging voltage for this cut off current will obviously lie above FCV @ 0 C and at the same time below 3.65 V @ 0.05 C.

This is the proper way to determine what charging voltage to target in your DC power supply for top-balancing.

Click to expand...

Honestly, I feel like you left out part of the math here. I expected a formula to say to charge at 0.011c to x.xx volts.

shvm said:

Anyone saying top-balancing of cells in parallel is done by setting DC power supply to 3.65 V beforehand and letting current drop to 0 is unaware of complete charging model for LFP.

Click to expand...

Or they understand that a single, less than idea charge cycle isn't going to cause catastrophic cycle loss.

shvm said:

The second part of this guide will cover maintaining top-balance in the context of BMS settings, good LFP charging parameters and how to closely emulate it using Lead Acid charging parameters.​

Click to expand...

I can't wait to hear your wisdom, master ?

You demonstrate a well educated student's ability to digest a spec sheet and treat it as the gold standard (that's good) but you seem to be lacking any real life experience, testing, exposure to how LifePO4 cells actually work. If your entire point here is to warn people about the tail current when charging, good job. That's marginally important.

400bird said:

Fun theory. Sure, you could do more balancing every couple months. Or you could do the industry standard and leave the balancer to do it's thing anytime needed, in tiny increments.


Is this 0.05c your entire point? Or 3.65 volts?

The company I work for has worked directly with EVE to establish our charge profile. EVE provided us an extended warranty if we terminated charge at 3.55 volts/cell and 0.05c
There is almost no capacity lost by only charging to 3.55 volts when compared to 3.65 volts.



False, based on your info above. You are saying 3.65 volts above.
I understand that you mean voltage levels like 12 or 48 volts, but you're the one coming here an insisting you're are the smartest person in the room and that everyone else is an idiot. At least he clear about your facts.

Have you tested a cell? Do you know what this newly invented "fully charged resting voltage" might be?

Underline is my added emphasis. That's not the resting voltage.

Voltage isn't a good indicator that the cell is charged and unless you've done lots of research, on your cell voltage curve, it would be much easier to use coulomb counting/SOC to know when to restart a charge cycle.


Honestly, I feel like you left out part of the math here. I expected a formula to say to charge at 0.011c to x.xx volts.


Or they understand that a single, less than idea charge cycle isn't going to cause catastrophic cycle loss.

I can't wait to hear your wisdom, master ?

You demonstrate a well educated student's ability to digest a spec sheet and treat it as the gold standard (that's good) but you seem to be lacking any real life experience, testing, exposure to how LifePO4 cells actually work. If your entire point here is to warn people about the tail current when charging, good job. That's marginally important.

Click to expand...

Yeah, but shvm is an engineer with a degree!??

Partimewages said:

Yeah, but shvm is an engineer with a degree!??

Click to expand...

It's a perfect example of education/experience/competence in one area making one believe they have inherent competence in other areas.

I too have an engineering degree and 30 years of engineering experience in a technical field, and aside from learning V = I * R and P = I * V, it didn't grant me ANY competence with batteries.

However, a few thousand hours of working with literally 10s of thousands of battery modules of multiple chemistries over the last eight year has bestowed something above a "layman's" level of expertise. I still have a lot to learn. Just not from @shvm until he/she/they/it demonstrates some level of competence.

Oh, and @Partimewages, you should be more selective on who you follow over at Community...

This thread is starting to remind me of a story I once heard on the Art Bell show (a late night radio show on paranormal topics), called "Mel's Hole". It was about a hole in the ground of his family property in Washington state that was mysterious to them. Anything they dropped into the hole, a dead cow, a dead car, garbage, etc, would never make a single sound. No crash or bang. They stopped trying to measure the depth at about 80,000 feet (24000m)?! They lowered a bucket of ice cubes deep into the hole, left it there for a few hours, then pulled it up. The ice cubes were still frozen solid but now were hot to the touch. Rumors it was a hole built by aliens they used to go into inner earth and all that It got weirder, too, especially when he did Part 2 a couple years later. Really weird.

Finally, after what must've been over 4 hours of this story on the radio, the guy, Mel, finally let's out that the reason he had to move from Washington state, was the feds were after him. "Oh so why where they after you?" Art Bell asked. He said they wanted what he was growing in his greenhouse and put in him their special ops program but he didn't want to work for them (or that's my memory of it anyways). Long story short, it took still a bit longer for Art Bell to finally suss it out of Mel...Mel was growing psilocybin mushrooms in his greenhouse near the hole!

"AHhhhhhhhhhhhhh......." said Art. "You've got to be kidding me! You've been growing mushrooms???"

@sunshine_eggo tagged since you seem to like weird shit, lol.


If you want to listen to the original episode (there are two more even weirder), here's the link:
https://www.coasttocoastam.com/videos/art-bell-vault/guest-only-mel-of-mel-s-hole/ (you need to pay to listen, $7/mo)

Despite the fact that "Mel" never existed in the area he claims to have lived and he was likely in la-la land of shrooms, it was hands down one of the single most entertaining stories I have ever heard in my life!

I have no degree of any sort. Ate slept and drank electronics for my last two years of vocational school. The last 40 years have been spent in the electrical/DDC/HVAC industry. Throughout that time I've seen a lot.
I can usually spot BS from a mile away including those that have everything figured out.
I'm still however a student!

Tomthumb62 said:

@sunshine_eggo tagged since you seem to like weird shit, lol.

Click to expand...

Hmmm... both offended and feel like you get me...

sunshine_eggo said:

Hmmm... both offended and feel like you get me...

Click to expand...

Well I like weird shit, too, lol! Check my comment again, I updated and posted the audio link if you dare listen to that crazy story. It's worth it!

Late night with Art Bell!

Partimewages said:

Late night with Art Bell!

Click to expand...

I used to LOVE that show, especially since I was a night owl in that time of life. So many crazy stories I got to hear. Hearing of Dr Greer and watching his The Disclosure Project was eye-opening...lots of ex high-level military (or a good hoax) who shared their knowledge of aliens and UFO's was the pinnacle of it (along with Mel's Hole).

Eventually I grew tired of the occult and paranormal stuff, it gave me an ongoing sense of paranoia. Over things of which I have no control, if true. I've found other interests now, like LFP and camping for weeks off-grid in our trailer!

Heh! Mel's Hole was one of Art's greatest hits. I'm sure it was well within the 98% of the show's content that was complete BS, but it was great fun. Art knew how to put on a show, that's for sure.

400bird said:

EVE provided us an extended warranty if we terminated charge at 3.55 volts/cell and 0.05c

Click to expand...

Now THAT says something.

Tomthumb62 said:

Well I like weird shit, too, lol! Check my comment again, I updated and posted the audio link if you dare listen to that crazy story. It's worth it!

Click to expand...

Think I found it without a paywall, listening to part 1 now.

42OhmsPA said:

Think I found it without a paywall, listening to part 1 now.

Click to expand...

Yes that’s it! Good find, I could never find it. Let us know when you come up for air ?

ghostwriter66 said:

ME = "Hey whatever most of us are doing seems to work - there's no fire -- good enough for me"

Self-appointed "LFP Expert" - OMG your using a 2/0 instead of 3/0, your torqueing your batteries at 8 n/m not 8.5, your charging them to 3.65 instead of 3.6255, you did not compress your batteries - you did not grind the terminal base down to a shining marble finish before connecting busbar, YOU realize that you have taken at least 2 days off the lifespan of a 15 year battery!!!"

Click to expand...

Haha, yeah totally, I just try to follow all these guidelines (except for springs, just tighten bolts) and crossing my fingers, hoping I can add a couple extra bonus days to my battery lifespan (before all the new battery tech hits the market and obsoletes LFP). I hope I didn't lose too many cycles for not using springs and just tightening my cell clamps with my trusted arm-wrench and no torque wrench..

Hey, this is a good place for a stickup (might as well add it to the mix):

SeaGal said:

View attachment 180837

yea.. right... nuff said... I'm joining @42OhmsPA with his ? again...

Almost everyone is charging their LFP wrong!

Some background: I built my own LFP battery consisting of 16 CALB L160F100 with the JK-BD6A24S10P paired to a LuxPowerTek SNA 5k unit. It is 50 days old showing 21 cycle counts on the BMS as of 26 Oct 2023. It took the battery (51.2V, 100 Ah, 0.6 A active balance current) a full month to...

diysolarforum.com

Click to expand...

Seagal,
Shvm's background got clipped:
Masters in Industrial & Manufacturing Engineering

Thank you for putting efforts to read what I wrote.

400bird said:

I understand that you mean voltage levels like 12 or 48 volts,

Click to expand...

You and I both know that nominal voltage of LFP is 3.2 or 3.22 voltage. You understand that I'm talking about 12/24/48 V systems still you chose to nitpick.

but you're the one coming here an insisting you're are the smartest person in the room and that everyone else is an idiot.

Click to expand...

You're cooking things up at this point trying to be demeaning.

Have you tested a cell? Do you know what this newly invented "fully charged resting voltage" might be?

Click to expand...

My emphasis is that anyone can determine this voltage by experimenting if they don't trust me. This is well known to be ~3.37 V for LFP.

400bird said:

Underline is my added emphasis. That's not the resting voltage.

Click to expand...

Nice try but you missed the 'less than' '<' mark to the left of 3.65V. Resting voltage is lesser than 3.65 V.
3.37 V < 3.65 V.

400bird said:

Voltage isn't a good indicator that the cell is charged and unless you've done lots of research, on your cell voltage curve, it would be much easier to use coulomb counting/SOC to know when to restart a charge cycle.

Click to expand...

You are ignoring the what I term continuity aspect of Solar applications.
If a cell is fully charged at say, afternoon, and there is excess power available, the cell can't accept more charge.
Now suppose, if demand comes and the excess power can't meet the load, then power must be drawn from the cell.

Now if that demand is completed (no need to discharge from cell) and excess energy is there, then the OCV of the Cell has to be below FCV.
Voltage in the absence of current is a good indicator of when to initiate top-up charging cycle again.

400bird said:

Honestly, I feel like you left out part of the math here. I expected a formula to say to charge at 0.011c to x.xx volts.

Click to expand...

I knew this would come up.

The last edit to that post was in November. Your concerns are addressed.

I don't feel the need to bog down everyone with math where it is unnecessary.

sunshine_eggo said:

a few thousand hours of working with literally 10s of thousands of battery modules of multiple chemistries over the last eight year has bestowed something above a "layman's" level of expertise. I still have a lot to learn.

Click to expand...

If the underlying foundation of your very understanding is at fault, there is only so much difference those 'thousands of hours' and experience can make.

I can smell from miles away who is confrontational possessing a good understanding and who isn't.

This you??

Balancing above 3.40 is productive. Cells in near perfect balance at 3.45V nearly always extrapolate to near perfect balance at 3.65V.

Click to expand...

From how I understand things, (Top) - balancing basically implies bringing cells to 100% SOC. It's not some beauty pageant of aligning voltages.
The resting OCV of a LFP Cell at 100% SOC is 3.37 V
I can't understand nor do I want to know whatever you are doing above 3.37V regarding balancing.

Want more?

Fast charging to full: 3.55-3.65V/cell - 30 minute absorption time.
Slow charging/longer cycle life (98%+ charge): 3.45V/cell - 4+ hour absorption time.
Float (95%+): 3.4V/cell

Click to expand...

This is you talking of floating LFP above 3.4/Cell when 3.37 Volts is the voltage corresponding to 100% SOC.

'absorption time', 'slow charging', 'longer cycle life' - hot damn !!

It is you who needs to get down your high horse.
The reason you were ignored because your criticism is hollow. Engaging with you will derail any chance of meaningful discussion happening.