diy solar

diy solar

Float LiFePO4 or not??

rhino

Solar Wizard
Joined
Jun 6, 2020
Messages
2,554
Location
Minnesota
Can someone give a definitive answer if you should turn off float stage in charge controller for LiFePO4 cells? I see conflicting information in messages here. Most of the information I see says Float should be disabled completely. cc: @Steve_S
 
Can someone give a definitive answer if you should turn off float stage in charge controller for LiFePO4 cells?
With most things in life it depends. In other words the assumptions on use will determine whether a "float" charge will hurt. For example if your inverter is on and powering loads it cannot hurt a Lithium battery to have some of that draw offset by a low Amperage constant voltage. On the otherhand, unattended Lithium batteries with no load can be damaged long term by continuosly applying even a small current while they are fully charged.

The term float or trickle charge is a process designed to offset the typical self discharge of a Pb (Lead Acid) battery. Since there is very little self discharge of Lithium cells, the definitive answer is that no trickle or float charge is necessary.
 
What @Ampster said is correct. The float stage of charging is beneficial to a lead acid battery, but it is not beneficial to a LiFePO4 battery.

However, most people use the float setting on their charge controller to allow the solar to power loads while there is sun available. If you disable float, the SCC will not push any current (or allow any to be pulled from the SCC) until the battery voltage falls to the recharge voltage. That wastes lots of available solar power.

A float voltage of 3.4V per cell or thereabouts is perfectly fine. It doesn't harm your LiFePO4 cells, and lets your SCC power all your loads during daylight hours.
 
Please refer to this post on the next page for the latest & correct posting.
 
Last edited:
The Rabbit hole... shoot the rabbit & eat it.

In the context of Lithium Float is Constant Voltage Variable Current. I USE FLOAT - It Works and I'll explain with a pinch of Curmudgeon.
Bulk Charge or CC get's you to 95% of your TARGET Voltage, in my case 27.0V (3.375Vpc), the ABSORB takes it to 28.2 (3.525Vpc) for a MAX of 1 hour or until EndAmps (Tail-Current) of 14A is reached.

Then float kicks in at 27.9 and starts to decrease Amps as the packs take it, while this is occurring the cells are a bit Deviated after the 28.2V so the balancing is underway as well as the cells themselves settling post charge*, while Floating the deviations decrease as all cells level up as the lower charge cells are taking more than the higher ones within the pack, This IS observable.
* Post-charge LFP will always settle if not provided with power, depending on "saturation" (depth of charge at voltage) that settling can be significant.
-- Typically by 12-13:00 hours I am in FLOAT, Amps decrease from 14A as the packs top off & balance out, By 14-15:00 batteries are in "storage Mode" (BMS function) and float is servicing whatever demands my Inverter has, even if I start the Coffee Maker that pulls 50A, if the panels can provide it, the batts won't be touched.
-- IF the batts are pulled from, the SCC flips back to Bulk and tops off again and returns to Float when done, this actually is fairly quick.

NOTE: I use Chargery BMS, with Passive Balancing (only ON during Charge) above 3.30V and allowing for 30mv differential. All my Production Packs also have a QNBBM-8S Active Balancer.

A LAST POINT because it is a PEEVE (being extremely polite OK)
A FOOL charges to 3.65Vpc every day ! An IDIOT keeps charging above Working range {3.50} (THAT is where you are aging / hurting cells) and that means TOPPING OFF to <2A Taken ... ie Charging and Floating to 3.500Vpc 28.0 & <2A.

But BattleBorn Says ! Those are Grade A+++ Matched & Balanced Premium Cells which they test the crap out of. Even they say you can Goto 29.2 (3.650Vpc) but they recommend 28.4 (3.550Vpc) And forget what Chinese Battery Packers say, they want to sell more, so burn em up and buy more... "Consume & Spend".

Hope that helps, Good Luck.
PS: I came from Lead World, am familiar with LEADisms.. Humbug ! Not entirely applicable in context.
Battleborn cell packs are 30 cylindrical cells per pack. 4 packs per 12v battery. Top balancing of 12v battery with the BB BMS is passive per 'cell pack' balancing, not per cell in each pack balancing. Top balancing starts at 14v + for a 12v BB battery.
I have looked at Lithium-ion Cylindrical LFP 3.2/3.2 26650 LFP data sheet (CEGASA). Discharge cuttoff voltage is 2.5v, charge voltage is 3.65v (per cell or cell pack). I did not see a SOC usage recommendation on that data sheet. BB states use it's 26650 cells 100% to 0% SOC.
I also looked at Prysmatic cell data sheets. In particular a CALAB 100 AH cell. Discharging cutoff voltage is 2.5v. Charging upper limit voltage is 3.65v. The data sheet includes Recommended SOC Usage Windon SOC 10%-90%.
I think to be correct when comparing charging parameters, consideration should be given to whether the battery cells are cylindrical or prismatic. Including battery float voltage spec/usage consideration.
 
Last edited:
I think to be correct when comparing charging parameters, consideration should be given to whether the battery calls are cylindrical or prismatic. Including battery float voltage spec/usage. consideration.
I was not aware of any significant difference in charging parameters within the same chemistry. Your examples support that they all use the same voltage for discharge cuttoff and maximum charge voltage. As far as Float is concerned it is my opinion that it is more about the use case and not the physical form factor of the cells. For example, for many years I used Headway LFP cylindrical cells then evolved to Winston and Thundersky LFP primatic cells for an EV and a stationary pack.
 
I have my SCC set to charge to 13.90 and hold for 10 minutes. Then drops to float at 13.00 volts.
Generally the battery will be charged once a day and then off. Have to be extraordinary circumstances to reach 13.00 volts.
I like the idea of cycling the battery daily.
 
I was not aware of any significant difference in charging parameters within the same chemistry. Your examples support that they all use the same voltage for discharge cuttoff and maximum charge voltage. As far as Float is concerned it is my opinion that it is more about the use case and not the physical form factor of the cells. For example, for many years I used Headway LFP cylindrical cells then evolved to Winston and Thundersky LFP primatic cells for an EV and a stationary pac
I wonder why the CALAB 100ah prysmatic cell data sheet states: Recommended SOC usage Windon - SOC 10%-90% ?
 
Any time you take LiFePO4 into either voltage knee you reduce its lifespan.

If you set “float” on a charge controller to 3.4V, your charger will be at zero amps without taking your cell into its high voltage knee.
 
I was not aware of any significant difference in charging parameters within the same chemistry. Your examples support that they all use the same voltage for discharge cuttoff and maximum charge voltage. As far as Float is concerned it is my opinion that it is more about the use case and not the physical form factor of the cells. For example, for many years I used Headway LFP cylindrical cells then evolved to Winston and Thundersky LFP primatic cells for an EV and a stationary pack.
Why do some prismatic cells require compression by the manufacturer ? Cylindrical cells do not require compression, they are each in a steel case that will not expand. What causes the prismatic cells to expand. What does that expansion do to the structure of the cell elements ?
 
The Rabbit hole... shoot the rabbit & eat it.

In the context of Lithium Float is Constant Voltage Variable Current. I USE FLOAT - It Works and I'll explain with a pinch of Curmudgeon.
Bulk Charge or CC get's you to 95% of your TARGET Voltage, in my case 27.0V (3.375Vpc), the ABSORB takes it to 28.2 (3.525Vpc) for a MAX of 1 hour or until EndAmps (Tail-Current) of 14A is reached.

Then float kicks in at 27.9 and starts to decrease Amps as the packs take it, while this is occurring the cells are a bit Deviated after the 28.2V so the balancing is underway as well as the cells themselves settling post charge*, while Floating the deviations decrease as all cells level up as the lower charge cells are taking more than the higher ones within the pack, This IS observable.
Strictly speaking, I don’t believe this is correct. In a series string, all cells ‘take’ exactly the same charge current during float (and at any phase of the charging process).

What happens once cell charge has reached the level to activate the BMS balancing function is that cells that are higher that the lowest cell will have a modest amount of balance current ‘bled off.’

If you want to take Icharge - Ibalance as the net charge into a cell, fine, the higher-voltage cells are ‘taking’ less charge than the lower-voltage cells.

But in general, all cells in a series string recieve or ‘take’ the same charge current and it is only the BMS balance currents that vary between series cells (and only once charge voltages are high enough to activate them…).

* Post-charge LFP will always settle if not provided with power, depending on "saturation" (depth of charge at voltage) that settling can be significant.
-- Typically by 12-13:00 hours I am in FLOAT, Amps decrease from 14A as the packs top off & balance out, By 14-15:00 batteries are in "storage Mode" (BMS function) and float is servicing whatever demands my Inverter has, even if I start the Coffee Maker that pulls 50A, if the panels can provide it, the batts won't be touched.
-- IF the batts are pulled from, the SCC flips back to Bulk and tops off again and returns to Float when done, this actually is fairly quick.

NOTE: I use Chargery BMS, with Passive Balancing (only ON during Charge) above 3.30V and allowing for 30mv differential. All my Production Packs also have a QNBBM-8S Active Balancer.

A LAST POINT because it is a PEEVE (being extremely polite OK)
A FOOL charges to 3.65Vpc every day ! An IDIOT keeps charging above Working range {3.50} (THAT is where you are aging / hurting cells) and that means TOPPING OFF to <2A Taken ... ie Charging and Floating to 3.500Vpc 28.0 & <2A.

But BattleBorn Says ! Those are Grade A+++ Matched & Balanced Premium Cells which they test the crap out of. Even they say you can Goto 29.2 (3.650Vpc) but they recommend 28.4 (3.550Vpc) And forget what Chinese Battery Packers say, they want to sell more, so burn em up and buy more... "Consume & Spend".
 
Any time you take LiFePO4 into either voltage knee you reduce its lifespan.

If you set “float” on a charge controller to 3.4V, your charger will be at zero amps without taking your cell into its high voltage knee.
When I set battery float at 13.60v (cell float 3.40v), after a Bulk and absorb (absorb time 10 minutes) of 14.1 v, my cylindrical cell 12v battery settles at 13.60 v (the float setting). The SCC settles at 13.60 v, 0.60 a. The battery settles at 13.60v 0.00a. The SCC suppliess the 0.60 amp parasitic load to my system. If I turn the inverter on, with no ac load, the dc load on the system is 3.60 a at 13.60 v. Thats 0.60 a parasitic and 3.00 a to inverter idle at no ac power out. The battery sits at 3.60 v, 0.00 a. If the SCC powered float voltage is not able to keep up with the system load (clouds, shade, more load current), the voltage will drop 0.10 v for more than one minute and the SCC will go into Bulk charge mode and repeat the solar charge cycle (less absorb duration). 13.60 v at 0.25C load is 100% SOC, at zero current load is 99% SOC. Perfect for maintaing max solar energy in BB battery on sunny days. The battery sits at nearly 100% SOC ready for the night or a few cloudy days in the future.
I just noticed that 13.60v (3.40v per cell pack) is up pretty high on the charge/discharge knee. 13.40v (3.350v per cell pack) (99% SOC and 90% SOC) is down a little lower on the knee and I am sure it would serve the same purpose as the float setting I am trying out now).
 
Last edited:
Because the cases are not able withstand the normal swelling of ingredients like cylindrical cells can.
It is my understanding that during charging to 100% SOC of a prysmatic battery the litthium Ions attach to the Cathode. The ions are rather large and over time can cause cracking in the cathode if cell expansion is allower. You do not have this phenomina with cylindrical cells. The swelling contributes to less battery longevity in prysmatic cells.
I am not sure I understand this correctly, can anyone set me straight. Reference to a white paper would be appreciated.
 
It is my understanding that during charging to 100% SOC of a prysmatic battery the litthium Ions attach to the Cathode.
The Lithium ions also attach to the Cathode in a cylindrical cell. The only difference between a prismatic and cylindrical cell is the form factor not the chemistry. Some prismatics use a jelly roll configuration shaped into more or less a rectangle and others use a series of individual anode and cathode sheets. A cylindrical cell is a circular jelly roll. I don't have a link to a white paper but a Google search can probably get you links to various types of cell construction.
 
The Lithium ions also attach to the Cathode in a cylindrical cell. The only difference between a prismatic and cylindrical cell is the form factor not the chemistry. Some prismatics use a jelly roll configuration shaped into more or less a rectangle and others use a series of individual anode and cathode sheets. A cylindrical cell is a circular jelly roll. I don't have a link to a white paper but a Google search can probably get you links to various types of cell construction.
Got it. I read somewhere that in Prismatic cells the cathodes expand as large lithium ions attach to them. Thus the swelling of prysmatic cells. But no swelling can occur in cylindrical cells. This causes cracking of the cathode material over time in prismatic cells. Thus a need to compress prysmatic cells. This process reduces the life of prysmatic cells. I think thats why the CALB data sheet mentions using the battery cell in the 90%-10% SOC range (to extend their life). I do not see this on a cylindrical cell data sheet I viewed. If I'm wrong here please straighten me out. If you have a link to a white paper that discusses this phenomina, please post a link.
Thanks.
 
Denis talks about lifepo4 batttery float with Will:
 
Last edited:
As for differences in data sheets and state of charge suggestions, the soc usage determines whether or not it'll hit rated cycle counts. If they rate it for 0 to 100% then it'll do rated cycles for that range, or should.

Plenty of manufacturers rate them at different capacities and the only reason is to increase cycle counts. Has little to do with varying chemistry so long as youre comparing like chemistry in the first place.


I just noticed that 13.60v (3.40v per cell pack) is up pretty high on the charge/discharge knee.
*checks my own discharge curve*

Yep.

At any rate my 1.5 cents here is that I use a crappy wfco charger which absorbs at 13.6 for a day then floats at 13.2v, and I see a slight discharge of about half an amp when it's on float.

This suggests to me that 13.2 will never be a problem because the resting voltage of the battery after 13.6v is higher than the float anyways.

This is also due to all the various parasitic draws in my camper. I can reduce this a fair bit by shutting everything off of course but it's not in "storage" as yet.


Would be nice if I could cut that 13.6v time down to just a few hours but I'm not going to stress about 3.4v/cell too much.
 
I read somewhere that in Prismatic cells the cathodes expand as large lithium ions attach to them.
Do you have a link? I do not doubt that it said that. However physics says the same Lithium attachment has to take place in any similar Lithium chemistry. By its shape, a cylindrical cell is in compression automatically and prevents the impact of that expansion just like compressing prismatics does.
 
Do you have a link? I do not doubt that it said that. However physics says the same Lithium attachment has to take place in any similar Lithium chemistry. By its shape, a cylindrical cell is in compression automatically and prevents the impact of that expansion just like compressing prismatics does.
I'm sorry, I do not have the link. Was in hope that someone elese hear understood the same thing and might have some link to more info.

Some compress the prysmatic cells, many do not.
 
Back
Top