Victron charge controllers for lifepo4

[karrak]

So dude, in the end with my 40ah system made of GBS cells, I purposely derated by 30% (20% on the low side to stay out of the discharge knee, and 10% from the top for my relatively low daily cyclic 13.8v CV) to amount to basically a 28ah battery in cyclic conditions for best overall life. With the high up-front price of LFP in the first place, this is a bummer financially.

PNjunction, may I ask under what conditions, especially what charge current, and how you terminate the charge that you get an SOC of 90% when you use 13.8V as your charge voltage? Also could you define exactly what you mean by 100%SOC.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

---

[PNjunction]

This is such a GREAT question! Thank you for asking. It is a multi-part answer.


1) Since fully charging a LiFeP04 is not necessary to keep the cell healthy, as one would need with lead-acid, 100% SOC would be defined as any charge level that satisfies the user's capacity needs.

2) Manufacturers and users coming from lead-acid however define 100% SOC canonically as charging any cell (LiFeP04!) from between 3.45v to 3.6v and allowing it to stay there and absorb until charge current falls to .05C. To actually reach 3.45v, one needs about a minimum of 0.05C charge current, as even LFP has minimum efficiency limits. In this case, the 100% SOC usually equates to demonstrating the manufacturer's full rated capacity.

The key point is that it takes 0.05V over 3.4v, ie 3.45v charge voltage per cell (along with absorb time), to achieve this.

(For 12v / 4S LFP users, this would equate to charging at anywhere from 13.8 to 14.4v at a pack level (AND assuming all cells are relatively in balance of course!) and allowing absorb to fall to 0.05C. Naturally, it takes longer at the lower voltage)

Termination?

For many daily cyclic solar users with an adequately sized housebank they may NEVER reach 0.05C taper current anyway! If they do, the sun is no longer shining shortly thereafter.

If they are *not cyclic*, but irregular or standby use with daily exposure, then the termination solution is to charge at no more than 3.4v per cell (13.6v for you 12v/4S pack level users), thus never reaching a full state of charge and damaging the bank over time. Basically no termination either, as the long-term lithium plating voltages are never reached with this method.

** NOTE **
To ensure that ALL the cell material is evenly saturated upon receipt, taking the cells to 3.6v / 0.05C termination current at least once is recommended. This can be achieved by a "single cell" 3.7v charger (3.7v not a big deal for individual cell use) of at least 0.05C or larger current and applied to each cell individually. This is the safest method, and doesn't expose the cells to overcharge from questionable vampire boards, or hold them at full charge longer than necessary waiting for laggards to catch up. Do not buy trash or used cells in the first place too.

Other methods have been endlessly discussed elsewhere. I've used the single cell chargers, or my own programmable CC/CV bench supply for that duty. When using conservative methods that basically stay out of the knees, DERATE your stated cell capacity by 30%. Open wallet. Ask if you REALLY need LFP in the first place.

This also ensures that your cells start out with an equal SOC. That does not guarantee absolute capacity equality of each cell. But if you buy new from a reputable supplier, and not use TRASH / used/ abused/ cheapskate countereits, then your chances of being good enough for our relatively low-current housebank application will be satisfactory with simpler conservative methods such as these.

Another reason to run low charge voltages is that the extremes of cell balance issues won't be as extreme as if you always run to high voltages. A proper initial individual cell charge is good enough not to have to sweat out extreme voltage differences when running at conservative cv voltages later. I do a yearly PM of reapplying an individual cell charge before dropping back to my normal daily conservative levels. That may not work for some and a different method will be needed.

Of course, an LVD is needed. Cell or pack level - pick your poison. At a pack level, I've found that 12.8v is a nice way to stay out of the discharge knee without having to resort to cell-level lvd. Then again, I didn't buy used trash, nor run at EV like discharge currents where a cell-level lvd would be prudent. All discussed elsewhere.

This works for ME. It may not be sufficient for YOUR needs or application.

[Sunking]

Termination?

For many daily cyclic solar users with an adequately sized housebank they may NEVER reach 0.05C taper current anyway! If they do, the sun is no longer shining shortly thereafter.

.

No reason to ever Terminate the charge with a voltage equal to or less than 14.2 volts or greater than or equal to 13.6 volts as you have indicated PN. Karrak ****** thinks you must use BMS and Terminate when in fact none of that is needed.

With Solar last thing you want to do is terminate the charge. Only Karrak and Dax **** do that. At noon his battery is charged, terminates the charged and spends the rest of the rest of the day on battery instead of using Solar Power until the Sun Sets.

Karrak ***** you can Float Charge a LFP battery with absolutely no BMS. Just Float 13.6 to 14.2 volts. It is ***** to Terminate a charge on Solar or daily cycle. **** I have built 478 vol tLFP plants with Telco cell sites. They Float at 56 volts 24 x 7 x 365 and never discharged except in Emergencies when power goes out. You gotta lot to learn Karrak because right now you do not even know the basics or have any biz answering any questions. You are a pretender and *******.

[karrak]

2) Manufacturers and users coming from lead-acid however define 100% SOC canonically as charging any cell (LiFeP04!) from between 3.45v to 3.6v and allowing it to stay there and absorb until charge current falls to .05C. To actually reach 3.45v, one needs about a minimum of 0.05C charge current, as even LFP has minimum efficiency limits. In this case, the 100% SOC usually equates to demonstrating the manufacturer's full rated capacity.

The key point is that it takes 0.05V over 3.4v, ie 3.45v charge voltage per cell (along with absorb time), to achieve this.

With reference to the above how do you charge a 4S LFP battery to 90%SOC using a charge voltage of 13.8V (3.45V/cell) as you have stated in previous posts.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

[Sunking]

** NOTE **
To ensure that ALL the cell material is evenly saturated upon receipt, taking the cells to 3.6v / 0.05C termination current at least once is recommended. This can be achieved by a "single cell" 3.7v charger (3.7v not a big deal for individual cell use) of at least 0.05C or larger current and applied to each cell individually. This is the safest method, and doesn't expose the cells to overcharge from questionable vampire boards, or hold them at full charge longer than necessary waiting for laggards to catch up. Do not buy trash or used cells in the first place too.

PN I have no problem with that. However not practical for Joe Consumer. Easy as heck for you and I because we have the equipment to do that. As you know, you cannot use a BMS to balance a LFP Battery. Hear that Karrak, YOU CANNOT BULK or INITIAL BALANCE A BATTERY with a BMS. They are not made to do that. It could take weeks, months, or a year.A BMS can only make very small corrections. If you tried with a Vampire BMS you will harm the charged cells pumping current into fully charged cells for hours and hours waiting for the lower cells to catch up.

Nope there is only one-way to bulk Balance LFP Batteries. Put them all in parallel, and use a very high current charger set to 3.6, walk away until all current stops. Reassemble them in series and they are balanced pretty much for life at the top and you risk over discharge at the top. Easier way for Joe Consumer is to connect them all in parallel, put a load on them, and discharge to 2.5 volts. Reassemble in series and Float Charge to 13.6 to 14.2 volts. Work your way up from 13.6 to find the sweat spot.

Joe Consumer cannot Balance a Battery period without the right equipment and know how. Joe Consumer should therefore buy a Drop-In Replacement and any LFP can be charged exactly like Lead Acid at 14.2 volts. If you buy any Charge Controller made for LFP, they are set to 14.2 volts period. No reason to terminate. FWIW Drop in Replacements are factor balanced and discharged to 60% for shipping.

[Sunking]

I thought that was the point of the float setting in terms of lowering the voltage. My battery specs that it can be floated indefinitely at 13.6V I think. In terms of shutting it off, the BCU is there to prevent overcharge, or is this really only supposed to be there as a backup measure?

Back up only. Not needed if your Inverter has Low Voltage Disconnect.

If you are talking about Lead Acid Batteries and use a 3-Stage charger you would be correct about FLOAT. With Lead Acid you charge to 14.4 volts until current stops, then lower the voltage to 13.2 to 13.8 volts depending on battery. However you can also charge a Lead Acid Battery with just Float Voltage. Every professional system used as Emergency power operates that way. Float charging is the kindest gentlest method as it stay below gassing and corrosion voltages. The only down side and why Solar does not use 1-stage CC/CV aka Float Charging is it takes up to 24 hours. You do not have that luxury with Solar with a 4 Sun Hour day.

Here is a secret not many know about Lithium Battery. The faster you charge them, the slower they absorb, or put another way take longer to absorb. Take a fully discharged 100 AH cell, charge at 1c and it wil take you two hours to recharge to 3.6 volts @ .03 finnish current. At 1C you wil hit Absorb when the battery is roughly 60% SOC, or about 30 to 40 minutes into the charge cycle. From there another 1 to 1.5 hours to Absorb to 100%.

Charge at C/2 and in about 1.5 hours you hit 80% SOC when you reach Absorb. and another 45 minute to an hour to finish Absorb at 100%.

Now here is what you really want to know. If you use LFP, and size the bank correctly for 3 day reserve, your Charge Rate is going to be roughly C/4 to C/5. If you set the charger to 14.4 volts, your battery will hit Absorb phase when the battery SOC is roughly 90% SOC and you can then either Terminate the charge, or back the voltage off to 13.6 volts just like a Lead Acid Battery and use Solar power until sun sets. When Sun Sets you are at the perfect 90% SOC. Keep the SOC to 90% or less and you double the life of your LFP batteries. Any good Solar Controller can do this. It does not need to be made for LFP. All LFP chargers are set to 14.2 volts all day long.

Your BCU is to do two things. Turn the Charger Off when all 4 cells indicate they are 100% SOC a place you never want to go. On the flip side to disconnect your batteries if over discharged which is OK but redundant if your Inverter has a LVD.

What you have is a Vampire BMS the number one destroyer of Lithium batteries.

[PNjunction]

With reference to the above how do you charge a 4S LFP battery to 90%SOC using a charge voltage of 13.8V (3.45V/cell) as you have stated in previous posts.

Ah, you are quite right - in my case, the sun sets before I have completed an absorb down to 0.05C. But yes, the next day, it will achieve full saturation, so if I was non-cyclic, it would be best to drop that to 13.6v total.

Ideally, if I wanted to be a stickler about 90%, one would incorporate some form of coulomb-counting calibrated against the rated cell capacity and re-calibrate as needed. Blue Seas and others make them if the needs are that critical. In fact, that would be the only way to TRULY come close to any sort of SOC percentage accuracy with all the variables tossed in - CV voltage vs panel size vs battery capacity vs solar insolation vs absorb time etc etc.

The problem with all the variables is that it can lead to endless bench-racing / logic-trap specifications against one another. I see that has already started.

In the context of an experimental 60ah 4S battery, a bit of conservatism and leeway with exacting percentages can be made.

[PNjunction]

Now here is what you really want to know. If you use LFP, and size the bank correctly for 3 day reserve, your Charge Rate is going to be roughly C/4 to C/5. If you set the charger to 14.4 volts, your battery will hit Absorb phase when the battery SOC is roughly 90% SOC and you can then either Terminate the charge, or back the voltage off to 13.6 volts just like a Lead Acid Battery and use Solar power until sun sets. When Sun Sets you are at the perfect 90% SOC. Keep the SOC to 90% or less and you double the life of your LFP batteries. Any good Solar Controller can do this. It does not need to be made for LFP. All LFP chargers are set to 14.2 volts all day long.

I would just say for lurkers that the solar controller MUST have the ability to disable any temperature compensation (not to be confused with thermal controller protection) for LFP usage.

Topping off the variable list are the controllers which have fixed absorb timeouts, like only 2 hours when in fact you could do well with just waiting for the sun to set. (or tweak the voltage up on your "float" setting)

The variables are enough to make one tear their hair out - which is why I went so conservative with my setup.

[karrak]

Ah, you are quite right - in my case, the sun sets before I have completed an absorb down to 0.05C. But yes, the next day, it will achieve full saturation, so if I was non-cyclic, it would be best to drop that to 13.6v total.

So what charge regime would you recommend for those people that have an off-grid power system with an LFP battery who use 30-50% of their battery capacity on a daily basis?
BTW, charging a 4S LFP battery to 13.6V (3.4V/cell) with an end current of C/50 results is an SOC of over 98%.
Below is a graph of SOC versus time during the absorb phase if you are charging at 3.45V/cell(13.8V for 4S LFP). As usual reality doesn't match what Sunking is saying. As you can see at C/5 the absorb phase starts at an SOC of around 95%. The absorb phase starts at around 98.5% SOC at a charge rate of C/5 and a charge voltage of 3.6V/cell (14.4V for 4S LFP) not 90% as Sunking has stated. AbsorbTimes3.45V.jpg

Ideally, if I wanted to be a stickler about 90%, one would incorporate some form of coulomb-counting calibrated against the rated cell capacity and re-calibrate as needed. Blue Seas and others make them if the needs are that critical. In fact, that would be the only way to TRULY come close to any sort of SOC percentage accuracy with all the variables tossed in - CV voltage vs panel size vs battery capacity vs solar insolation vs absorb time etc etc.

I agree with you that this is the only way to accurately measure your SOC percentage. That is how I measure SOC. These gauges have to be re-calibrated by charging the battery to 100% at least once a week.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

[Sunking]

I would just say for lurkers that the solar controller MUST have the ability to disable any temperature compensation

True but the caveat is if you live in an area where Temp Comp for,Pb batteries is required, is not an area where Lithium batteries should be used. Good way to start a fire. .

[karrak]

No reason to ever Terminate the charge with a voltage equal to or less than 14.2 volts or greater than or equal to 13.6 volts as you have indicated PN. Karrak ****** thinks you must use BMS and Terminate when in fact none of that is needed.


With Solar last thing you want to do is terminate the charge. Only Karrak and Dax **** do that. At noon his battery is charged, terminates the charged and spends the rest of the rest of the day on battery instead of using Solar Power until the Sun Sets.

Not true, I do not terminate the charge. I set the CV/Bulk/Absorb charge voltage at 3.45V/cell and drop to a float voltage of 3.35V/cell when the charge current drops to C/50. This results in a battery around 99% full at the end of the day if there is enough sunshine. I would recommend the same to any one else who is charging at a rate less than C/5. If you can't set the end current but can set the time that your solar controller stays in the absorb phase, set the time to 1/2 an hour.

Karrak ***** you can Float Charge a LFP battery with absolutely no BMS. Just Float 13.6 to 14.2 volts. It is ***** to Terminate a charge on Solar or daily cycle. **** I have built 478 vol tLFP plants with Telco cell sites. They Float at 56 volts 24 x 7 x 365 and never discharged except in Emergencies when power goes out. You gotta lot to learn Karrak because right now you do not even know the basics or have any biz answering any questions. You are a pretender and *******.

If you float an 4S LFP battery at 14.2 volts (3.55V/cell) it will have an SOC of greater than 99.8%. The battery doesn't need to be very much out of balance before you get individual cells above 3.65V/cell which is the point where Lithium plating starts in earnest. This will make the imbalance worse until you could end up with a situation where most of the cells are at 3.35 and the fourth is at around 4.2 volts for a 4S battery or around 5 volts for a 8S (24v) battery!

Do your Telco installations use LFP batteries? If they are using lead acid battery it has little if any relevance to the discussion.Lead acid batteries have been around so long that all you have to do to design a Telco system is to consult the vast number of texts that have been written on the subject. As LFP batteries are so new especially in stationary applications one has to do the research and design oneself.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

[karrak]

Joe Consumer cannot Balance a Battery period without the right equipment and know how. Joe Consumer should therefore buy a Drop-In Replacement and any LFP can be charged exactly like Lead Acid at 14.2 volts. If you buy any Charge Controller made for LFP, they are set to 14.2 volts period. No reason to terminate. FWIW Drop in Replacements are factor balanced and discharged to 60% for shipping.

Unless the Drop-In Replacement battery has a well designed and built BMS built in the lifespan of the battery can be significantly reduced regardless how good the individual LiFePO4 battery cells within the battery are. For more information copy and paste this link into your browser.
ka7oei.blogspot.com/2013/05/lithium-iron-phosphate-lifepo4.html

Keeping 14.2V (3.55V/cell) on a 4S LFP battery for an extended period will reduce its lifespan, A123 and other LFP manufacturers recommend a float voltage of no greater than 3.45V/Cell.

Some Drop-In Replacement battery manufacturers like Bioennopower go to the trouble of disconnecting the battery from the charger after it is full to avoid the cells being subjected to high voltages for an extended period of time.

If you have enough electronic knowledge I would look at getting the individual cells and making up your own BMS or getting a complete unit with inbuilt BMS from a reputable supplier who is prepared to give a decent warranty with the battery. If you do get a complete unit, it is good to have external access to the individual cell voltages so you can check to see if the battery is keeping in balance.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

Admin note edited out link

[PNjunction]

So what charge regime would you recommend for those people that have an off-grid power system with an LFP battery who use 30-50% of their battery capacity on a daily basis?

The generic answer for me would be to just set their controller to 3.45v per cell and let it stay that way until the sun sets in a cyclic situation. For someone in Anchorage right now, looking at more than the autonomy they planned for and no genny, things might be different. I'd crank it up to 3.55v per cell in that case since solar insolation is limited, and so is the exposure to high(er) voltages.

BUT, this is for guys who don't think of LFP as "maintenance free", but will do some monitoring on a regular basis. Kind of like using a hydrometer with flooded, or at least doing sanity checks on agm's. LFP is no different. The average Joe who doesn't care might be better off with a black-box system, vampire boards and more, and rely solely on the warranty.

Below is a graph of SOC versus time during the absorb phase if you are charging at 3.45V/cell(13.8V for 4S LFP).

Your timing is good - I'm about to do the yearly test which I normally do in January.

I just finished charging my GBS 40ah 4S system to 13.8v, (an individual PM cell charge to all cells was done 6 months ago) and let the current drop to basically nothing. I'm now doing a discharge close to 0.1C at 4.825A load. How long will it last before it reaches 12.8v total? (my generic limit for no higher than 0.1C.) Let's find out - we'll see you guys in about 8 hours or more with the results...

[Sunking]

Unless the Drop-In Replacement battery has a well designed and built BMS built in the lifespan of the battery can be significantly reduced regardless how good the individual LiFePO4 battery cells within the battery are. Read the following for more information.
https://ka7oei.blogspot.com/2013/05/...e-lifepo4.html

Keeping 14.2V (3.55V/cell) on a 4S LFP battery for an extended period will reduce its lifespan, A123 and other LFP manufacturers recommend a float voltage of no greater than 3.45V/Cell.

Some Drop-In Replacement battery manufacturers like Bioennopower go to the trouble of disconnecting the battery from the charger after it is full to avoid the cells being subjected to high voltages for an extended period of time.

If you have enough electronic knowledge I would look at getting the individual cells and making up your own BMS or getting a complete unit with inbuilt BMS from a reputable supplier who is prepared to give a decent warranty with the battery. If you do get a complete unit, it is good to have external access to the individual cell voltages so you can check to see if the battery is keeping in balance.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

Karrak you are dumb as a Rock and have no biiz answering questions or giving advice. You do not know the difference between Solar and AC chargers. I AM TYPING SLOWLY SO YOU CAN UNDERSTAND Karrak. Every charge controller made for LFP is 14.2 volts from sunrise to sunset. All Automobiles use 14.2 volts because it is the universal standard for all 12 volt batteries regardless of the type. If you knew anything about solar you would know it cannot over charge a battery unless there is a serious malfuction or poorly designed by you.

You can Float any 4S LFP from 13.6 to 14.2 volts. Take your pick based on application and usage. No BMS required. I will say this, you and anyone who has no knowledge with LFP batteries should use a BMS. You are not smart enough to go without. The penalty you pay is wasting money on a BMS, shorten cycle life, and increase the likely hood of a destroyed battery. Good for BMS sales.

[PNjunction]

Well, here's the test data ... and sorry about the 12.8v lvd reference - that is my *at rest* reference, I use about 12.5v normally for lvd. In this case, I took it to 12.4 under load (about 3.1v per cell) I did not want to take it down any further.

DUT: 40Ah GBS 4S LFP
Age: approx 3 years old
Initial charge: 3.45v per cell, until absorb dropped to zero current (below my metering, although I bet there is still a trickle - will check later)
Target discharge: 3.1v per cell under load near 0.1C
Load: 4.825A as measured at the terminals by a Fluke 87v

Test terminated when pack voltage reachead 12.40v under load. Duration: 6 hours, 7 minutes.

If cell spec of 40A is still true, then this represents about 6.125 hours * 4.825A / 29.67aH available.

That is just shy of 75% total rated capacity when charged / discharged this way.

Cell data:
Under load just prior to termination: max delta .062v

3.100
3.109
3.124
3.062 < -- If this were an EV application, guess who the major problem is. Can you say bottom balance would be more appropriate here for that? Yes.

After 1 hour rest (should be more rest, but shows the major trend) max delta .028v

3.205
3.208
3.213
3.185

That was fun.