Struggling with basic LiFePO4 settings in Epever Tracer

[mikefitz]

Please look at my pictures and calm me down if everything is normal

The only normal one the when you were using lead acid.

From around the time 13.45 the system is unstable with intermittent current flowing into the battery. Either the battery BMS is switching off the charge path or the Tracer is having a 'senior moment'. This could be a thermal effect on a component or faulty software.
During the period from 13.45 the battery voltage voltage is falling below the rest voltage. this may be due to the battery load or the BMS switching off the charge path to the battery.
Things are starting to go wrong when the voltage and power reach a certain level at 13.45, late in the day when there is less power there is an improvement but still unusual performance.
Things to try,
Use a default lead acid charge profile in the Tracer.
Use a Tracer charge profile with lower boost and float voltages
Use an AC battery charger to charge the Belluna battery and monitor the charge process.
Examine the information provided via the Belluna Bluetooth.

Since the new addition to your setup is the battery then this is the most likely cause of the charging problem, perhaps due to a faulty BMS or a cell reaching over voltage and the BMS switching off the charge path.

Hope you are able to solve the problem without too much stress!

Mike

 

[TheeDude]

The only normal one the when you were using lead acid.

From around the time 13.45 the system is unstable with intermittent current flowing into the battery. Either the battery BMS is switching off the charge path or the Tracer is having a 'senior moment'. This could be a thermal effect on a component or faulty software.
During the period from 13.45 the battery voltage voltage is falling below the rest voltage. this may be due to the battery load or the BMS switching off the charge path to the battery.
Things are starting to go wrong when the voltage and power reach a certain level at 13.45, late in the day when there is less power there is an improvement but still unusual performance.
Things to try,
Use a default lead acid charge profile in the Tracer.
Use a Tracer charge profile with lower boost and float voltages
Use an AC battery charger to charge the Belluna battery and monitor the charge process.
Examine the information provided via the Belluna Bluetooth.

Since the new addition to your setup is the battery then this is the most likely cause of the charging problem, perhaps due to a faulty BMS or a cell reaching over voltage and the BMS switching off the charge path.

Hope you are able to solve the problem without too much stress!

Mike

Thanks for the answer, I knew something was wrong here!
The load shouldn´t be a problem, I was at work the whole day, and caravan is empty, just the raspberry pi and a router drawing power, approximately ~5W or so..

Is it okay to use a lead-acid charge profile to charge a Lifepo4 battery?

I thought I did lower the boost voltages compared to others out there, with the purpose of increasing battery life.
But maybe I will have to lower it even more.

I have asked Belluna seller before installation what settings to use, and he was like "just be below 14.4V and you will be fine, the BMS will cut off when it is fully charged".

If the BMS is cutting out on let´s say overvoltage, does it only cut off charging, and not discharging?
It is a JBD BMS with xiaoxiang application.

I´m gonna check what the BMS is saying right away when I get home from work, we are going out camping this weekend, maybe I will bring the laptop with me so I can adjust the voltages from there..

 

[mikefitz]

Is it okay to use a lead-acid charge profile to charge a Lifepo4 battery?

Sort of yes, its not ideal, just a testing suggestion to check out that the Tracer is operational, try the GEL setting.

If the BMS is cutting out on let´s say overvoltage, does it only cut off charging, and not discharging?

Yes, but you get a small volt drop through BMS, the battery will still power loads

It is a JBD BMS with xiaoxiang application.

The app should show if there has been a protection event. It would be useful to check the parameter settings in the BMS and confirm they they are usable values.

As a workaround for your trip you could try setting boost and float to a voltage below 14 volts ( the system seemed to behave OK below this), say 13.6 to 13.8 volts. This would still put charge into the battery and power loads during the day.

Have a good weekend.

Mike

 

[TheeDude]

Sort of yes, its not ideal, just a testing suggestion to check out that the Tracer is operational, try the GEL setting.

Yes, but you get a small volt drop through BMS, the battery will still power loads


The app should show if there has been a protection event. It would be useful to check the parameter settings in the BMS and confirm they they are usable values.

As a workaround for your trip you could try setting boost and float to a voltage below 14 volts ( the system seemed to behave OK below this), say 13.6 to 13.8 volts. This would still put charge into the battery and power loads during the day.

Have a good weekend.

Mike

Thanks!
I did have a good weekend, the battery didn´t die and I had power for the whole weekend.
And I did have some time to troubleshoot what is happening.

The problem is that the battery cells seems way unbalanced.
at least one cell, the battery is disconnecting "charge switch" according to the app, because cell #1 is way to high, and shows over 3.65V and the other cells are showing ~3.35V..


As you can see on theese pictures, "cell overvoltage times" are increasing.


Here is a picture with it shwing that charge is off, because cell protection is on.




And here is pictures with all cell voltages, where you can see that cell #1 is way out of balance if compared to #2-4.

I would think that there is something wrong with either battery cell #1, or the BMS/wiring inside.
I did e-mail the seller/manufacturer the pictures in my first post, and this was his answer.

Hello

1.The voltage of the lithium battery will be close to 14.6v when it is just fully charged. When it is still for a period of time, the battery voltage will fall back13.2 ~ 13. 4V. This is the difference between lithium batteries and lead-acid batteries.
2.Lifepo4 battery charger is recommended.

After that, I did e-mail my pictures in this post, with the unbalanced cells, and haven´t got an answer in a week.
I have also reminded with another e-mail, but no answer on that also.

Can we agree that this is a battery/cell problem?
I have also asked Epever support, and they told me to lower boost duration time to 10 min, instead of 120 min, haven´t done that yet.
But I will do that, and if it doesn´t help, maybe I will make a PayPal claim to get a refund or something..
Something else to try?

What do you think?

 

[jnrhome]

did you fully balance the cells before putting them into production?

 

[MisterSandals]

The weird chit always happens at cell 1 for some unknown reason. If you have not top balanced, you should do that. Label your cells A, B, C, D.
Rearrange them after top balance as B, C, A, D just for a change.

 

[TheeDude]

did you fully balance the cells before putting them into production?

I though that the manufacturer would do that, so no I have only connected it to my Epever solar charge controller.

Cell voltage when new was
1 3.320V 2. 3.318V 3. 3.320V 4. 3.322
Total voltage 12,68V according to Bluetooth application.

The weird chit always happens at cell 1 for some unknown reason. If you have not top balanced, you should do that. Label your cells A, B, C, D.
Rearrange them after top balance as B, C, A, D just for a change.

Yes, the manufacturer told me that, because the #1 cell is the first one or something like that.
The battery is sealed, and I cannot rearrange them by moving cables or so.

So, how do i Top balance?
I have a 0-30V power supply, and also a charger that is in my caravan (that is not connected) that outputs 14.4V continuously.

I guess top balance happens when the voltage is high enough?
The seller told me to only discharge for a moment, and not charge, to make it use more cycles, and after that it will balance better, I don´t know?

 

[jnrhome]

That's almost certainly the problem. Disconnect the cells completely, then reconnect the cells in parallel. Add the power supply to the paralleled group of cells and allow the cells to charge up to 3.6v or so. If your power supply can do constant voltage, constant current, let them saturate until the charge current is less than .5A or so. Then I would separate the cells and let them rest overnight. In the morning, measure each cell. They should be within 5mV or so of each other. If not, hook them back up in parallel and do it all over. Once they measure within a few mV of each other after an overnight rest, connect them back up in series and put back into production. The cell variance should be gone and stay gone. If it comes back, the odd cell may have a problem. If you can measure the internal resistance of each cell, they should also be very close, within a couple milliohms if they were matched cells originally.

 

[TheeDude]

That's almost certainly the problem. Disconnect the cells completely, then reconnect the cells in parallel. Add the power supply to the paralleled group of cells and allow the cells to charge up to 3.6v or so. If your power supply can do constant voltage, constant current, let them saturate until the charge current is less than .5A or so. Then I would separate the cells and let them rest overnight. In the morning, measure each cell. They should be within 5mV or so of each other. If not, hook them back up in parallel and do it all over. Once they measure within a few mV of each other after an overnight rest, connect them back up in series and put back into production. The cell variance should be gone and stay gone. If it comes back, the odd cell may have a problem. If you can measure the internal resistance of each cell, they should also be very close, within a couple milliohms if they were matched cells originally.

Thanks for your answer!
But as I wrote in my last post, the battery is sealed in a plastic moulded box that isn´t supposed to be opened, and therefore I can´t disconnect the cells.
And this was something that the manufacturer should have done, and maybe have done, but there is something wrong with one cell?
Or is there something else that I could do besides that?
It should be 5 years warranty on this battery, it might seem that I have to convince them if there is something wrong with it..



Here is the last mail from the manufacturer/technical support, blue text is my questions to him, it is not the whole question, so it may look a bit out of context.
Can it be normal, or is he just trying to convince me it is normal and just "leave it like it is"?

 

[jnrhome]

hmm...
That's a toughie. How are you getting individual cell voltages if the unit is sealed?

 

[jnrhome]

Guess I should have looked at everything you posted before responding. What does your Bluetooth app look like? Does it have a balance feature and what are the parameters?

 

[rickst29]

Thanks for your answer!
But as I wrote in my last post, the battery is sealed in a plastic moulded box that isn´t supposed to be opened, and therefore I can´t disconnect the cells.
And this was something that the manufacturer should have done, and maybe have done, but there is something wrong with one cell?
Or is there something else that I could do besides that?
It should be 5 years warranty on this battery, it might seem that I have to convince them if there is something wrong with it..

View attachment 50712

Here is the last mail from the manufacturer/technical support, blue text is my questions to him, it is not the whole question, so it may look a bit out of context.
Can it be normal, or is he just trying to convince me it is normal and just "leave it like it is"?

View attachment 50713

Hmmm. I'll SWAG that they used insufficient bus bars, if charging of the first cell is significantly "favored" over charging of "later" cells. The charge current on the pack passes through ALL of the bus bars, and Voltage drop with each cell (and willingness to be charged) should be roughly the same. If, OTOH, the bus bars have significant resistance, then each additional bus bar drops Voltage by a bit, and the "later" cells will see less charging Voltage (the cells themselves being roughly equal in internal resistance, which should have been the case in building the batteyr pack.

 

[mikefitz]

lower boost duration time to 10 min, instead of 120 min,

Have you done this yet , you dont need any boost time for a small hobby system.
You could have a usable system by reducing boost voltage to 13.9 or 13.8. Try this, it will still charge the battery with a slightly longer charge time. However with the high imbalances of cell one it wont solve the issue if you charge to higher voltage. Charging at the lower voltage will, over time, allow the bms to bring the cells more in balance.
The battery is faulty in some way. If this is the company in Germany who sold you the battery, request a replacement,

Lithium-Batterien

Note this is a very small organization that specialize in converting household fridges to 12/24v operation by replacing the compressor. Importing batteries from China is a recent addition to their product range, I doubt they have any in depth technical experience with batteries.

The Epever works as follows:
When it wakes up it charges at the maximum power the panel can give until the battery volts equal Boost Volts.
It then holds the battery at a constant voltage, the boost volts, for the Boost duration.
At the end of Boost duration, the voltage drops to Float Voltage.
If the battery volts drop below Boost reconnect volts, it starts the charge cycle again.

With small charging currents the battery will be full by the time the battery volts equals the boost volts, so no or a short boost duration is needed, as advised by Tracer.

If the problem cannot be resolved by a replacement, and charging at a lower boost voltage does not balance the cell, you can cut the top of the battery and fix the issue.You should not have to do this , approach the seller for a replacment.

Mike

 

[rickst29]

I think that the battery voltages varies/fluctuates pretty much, I can see it charges up to 13.8 and sometimes up to 14.5V and it stops charging, it goes down to 12.6-12.8V, which is lover than the resting voltage "at night" it seems to go down to 13.3-13.4V at night when no charging is on.

The battery is new, and installed 3 days ago, I have a couple of logs from last year when I had a lead-acid battery and the voltages there seems to be more stable..

Please look at my pictures and calm me down if everything is normal, this is my first lifepo4 battery.
Theese logs are from a day with strong sun nearly all day.

....

And some bonus pictures, some installation pictures in my caravan..

View attachment 48823

THIS battery, in this picture, appears to specify a maximum charging Voltage of only 14.4 volts. You are likely 'tripping' the BMS if you use Battleborn values (or my own values, with "Bulk Charging" at 14.5V). Please cut back to 14.4 maximum charging voltage and then advise us if it behaves better. Thanks.

 

[mikefitz]

If you look at the plot below you can see the steeply rising voltage as the battery approaches full charge. Once above 14 volts, a high volts on cell 1 shuts of charging. the BMS recovers but each time the volts is greater than 14v it shuts down again ( the sampling missed the over 14 volts ).
Reduce the charge volts to 13.8 ( boost volts) and charge the battery, with for now, a long boost duration. This may help the balance and you can see via the app how things are going.
Try increasing the boost volts in 0.1 steps for subsequent charge cycles.
Note that any voltage above 13.6 v will charge to full.

Just cutting back to 14.4v wont help at this stage, lower to a value where the BMS is not tripped and hope the cell balance will improve.


Mike

 

[rickst29]

If you look at the plot below you can see the steeply rising voltage as the battery approaches full charge. Once above 14 volts, a high volts on cell 1 shuts of charging. the BMS recovers but each time the volts is greater than 14v it shuts down again ( the sampling missed the over 14 volts ).
Reduce the charge volts to 13.8 ( boost volts) and charge the battery, with for now, a long boost duration. This may help the balance and you can see via the app how things are going.
Try increasing the boost volts in 0.1 steps for subsequent charge cycles.
Note that any voltage above 13.6 v will charge to full.

Just cutting back to 14.4v wont help at this stage, lower to a value where the BMS is not tripped and hope the cell balance will improve.

View attachment 50855
Mike

I agree with this procedure. GREAT advice from @mikefitz

 

[RCinFLA]

Settings on the Epever are confusing and some of the values relate to the charge control and some to the load outputs, the manual is not very clear on this.)


Over volts disconnect : if the battery volts exceed this , the load outputs disconnect from the load from the battery.

Charging limit voltage: if the battery volts exceed this, charging the battery from solar is stopped.

Over Voltage reconnect: if the load outputs have been disconnected due to the battery exceeding over voltage a reconnect will occur at this value.

Equalize charge voltage: this term is used by Epever to indicate the voltage level that will be applied for a period (default 2 hours) once the boost target voltage has been reached. Most everybody else use the term absorption for this stage.

Boost charge voltage: under the boost mode, (referred to as 'bulk' by almost every body else) the controller will charge the battery at maximum power from the solar panels until this value is reached. At all times before this 'target' voltage is reached the maximum power control process will will try to 'pull' maximum power from the panels. Under poor solar conditions with a discharged battery it should remain in this mode for some time, Epever limit this mode to 2 hours as a default.

Float charge voltage: once the boost stage has been completed the controller will modify the maximum power search and load the panels to produce a constant float volts at the battery.

Boost reconnect voltage: once the unit is in float mode the voltage may vary due to solar conditions and any load on the battery, if the battery voltag falls to this value the controller re enters the Boost stage.

Low voltage reconnect: if the load outputs have been disconnected due to a low battery, this voltage is the turn on value.

Under voltage warning re connect: warning turned off at this voltage.

Under voltage warning: warning set at this voltage.

Low voltage disconnect: load outputs are disconnected from the battery at this voltage.

Discharging limit voltage: other than issuing a warning at the set voltage the stand alone unit cannot do anything about this.

Equalize Duration: the time duration where the voltage is held constant once the boost voltage has been reached, (Absorption period).

Boost Duration: the time period that power will be applied to the battery until the boost target voltage is reached, ( Bulk period).

.
Battleborn suggest the following settings,

Over Voltage Disconnect 14.7 V ( 3.675v/cell, with some cell imbalance, likely have a BMS cell overvoltage disconnect before this happens)
Charging Limit Voltage 14.6 V (3.65v / cell)
Over Voltage Reconnect 14.6 V (3.65v / cell)
Equalize Charging Voltage shut off or 14.4 V (don't use equalize)
Boost Charging Voltage 14.6 V ( 3.65v / cell)
Float Charging Voltage 13.6 V ( 3.40v / cell )
Boost Reconnect 13.3 V ( 3.325v / cell, too high but depends on use case and how much you accept stressing cells)
Low Voltage Reconnect 10 ( 2,5v / cell, I would raise to 2.8-3.0v /cell unless you will be running high % C discharge current from battery)
Under Voltage Warning Reconnect Voltage 11.5 ( 2.875v / cell, set at least 0.1v higher then LV disconnect, greater if running high load current)
Under Voltage Warning 11.5
Low Voltage Disconnect 11.0
Discharging Limit Voltage 10.5
Equalize Duration 0 or set as low as possible
Boost Duration 180 minutes (this is time in boost mode, unit does not have current taper level termination of boost)

Ideally for user programming you will need the MT50 or interface cables for Android or PC.

The Boost at 14.6 and float at 13.6 may be considered by some as too high and may wish to adopt more conservative values.

Mike

For off grid PV, the Boost reconnect is most important to stress on battery. If you boost to 3.65v per cell it is stressful but gives maximum cell capacity for given time in boost mode. This is usually desired for off-grid PV where you want to store as much energy possible when sun is out.

Any unit that only does timer period for boost mode termination and does not have boost charge current taper-off termination for boost mode will cook battery at boost voltage for boost timer period specified regardless of actual state of charge of battery. This is not good when using LiFePO4 battery.

The main issue is a boost cycle can be re-triggered everytime the inverter draws high current, momentarily dropping battery voltage below boost retrigger voltage. This can result in battery spending a high percentage of time at 3.65v which is very stressful on cells. Hopefully the charge controller requires the boost retrigger voltage be below set level for some minimal time period so it does not initiate a boost cycle for battery voltage slump on a short inverter surge load current everytime a water pump or other high starting current device turns on. I did not see this time specified in the manual.

PV charge controller should be wired directly to battery terminals to avoid high load inverter cables voltage drop.

There is no 'standard' settings. The controller needs to be tuned based on your PV array size, battery AH size relative to inverter power and what your peak and average inverter loads are. Also, how hard do you want to be on your batteries.

 

[TheeDude]

Thanks for all your answers!
They are all helping me very much!

Have you done this yet , you dont need any boost time for a small hobby system.
You could have a usable system by reducing boost voltage to 13.9 or 13.8. Try this, it will still charge the battery with a slightly longer charge time. However with the high imbalances of cell one it wont solve the issue if you charge to higher voltage. Charging at the lower voltage will, over time, allow the bms to bring the cells more in balance.
The battery is faulty in some way. If this is the company in Germany who sold you the battery, request a replacement,

Lithium-Batterien

Note this is a very small organization that specialize in converting household fridges to 12/24v operation by replacing the compressor. Importing batteries from China is a recent addition to their product range, I doubt they have any in depth technical experience with batteries.

The Epever works as follows:
When it wakes up it charges at the maximum power the panel can give until the battery volts equal Boost Volts.
It then holds the battery at a constant voltage, the boost volts, for the Boost duration.
At the end of Boost duration, the voltage drops to Float Voltage.
If the battery volts drop below Boost reconnect volts, it starts the charge cycle again.

With small charging currents the battery will be full by the time the battery volts equals the boost volts, so no or a short boost duration is needed, as advised by Tracer.

If the problem cannot be resolved by a replacement, and charging at a lower boost voltage does not balance the cell, you can cut the top of the battery and fix the issue.You should not have to do this , approach the seller for a replacment.

Mike

Yes, I have lowered boost time to 10 min, no improvement yet, but I have for some days ago disconnected solar panel, because the manufacturer told me so, after that, no answer from him, again, he is really bad at answering e-mails...

You are really sure that the battery is faulty even with the other comments below yours?
Just want to know if I am ready to request a replacement.
I can of course try to reduce boost voltage, just to try if it balances better.

I think that the battery voltages varies/fluctuates pretty much, I can see it charges up to 13.8 and sometimes up to 14.5V and it stops charging, it goes down to 12.6-12.8V, which is lover than the resting voltage "at night" it seems to go down to 13.3-13.4V at night when no charging is on.

THIS battery, in this picture, appears to specify a maximum charging Voltage of only 14.4 volts. You are likely 'tripping' the BMS if you use Battleborn values (or my own values, with "Bulk Charging" at 14.5V). Please cut back to 14.4 maximum charging voltage and then advise us if it behaves better. Thanks.

Actually, the BMS is set to "Pack overvoltage" to 14.6V so it should accept 14.4V charging but I have never used Battleborn values at 14.5V
I have in an earlier post showed my charging parameters, but I can add it again.
My boost voltage is set to 14.2V and Charging limit to 14.3V.
But I can lower it more just to try if it is behaving better..

If you look at the plot below you can see the steeply rising voltage as the battery approaches full charge. Once above 14 volts, a high volts on cell 1 shuts of charging. the BMS recovers but each time the volts is greater than 14v it shuts down again ( the sampling missed the over 14 volts ).
Reduce the charge volts to 13.8 ( boost volts) and charge the battery, with for now, a long boost duration. This may help the balance and you can see via the app how things are going.
Try increasing the boost volts in 0.1 steps for subsequent charge cycles.
Note that any voltage above 13.6 v will charge to full.

Just cutting back to 14.4v wont help at this stage, lower to a value where the BMS is not tripped and hope the cell balance will improve.

View attachment 50855
Mike

As I said to rickst29 above here, I haven´t had the boost voltage set to 14.4, so the epever shouldn´t be up at 14.4V, it is set to 14.2V.
But of course, I will try your advice pointed out here.
I will go out there right away and change theese settings!
Thanks again!

Guess I should have looked at everything you posted before responding. What does your Bluetooth app look like? Does it have a balance feature and what are the parameters?

I will check this!

Edit: Solar panel has been unplugged for a couple of days, and the cells are having an equal voltage at 3.26V, and I have lowered boost voltage to 13.8V, so hopefully I get sun here tomorrow to try it out.

Here comes balance parameters from the BMS bluetooth app.
What I can understand with that is that it will balance or try to balance when the voltage is above 3.4V.
Maybe it will get there with a little slower charge..

 

[TheeDude]

For off grid PV, the Boost reconnect is most important to stress on battery. If you boost to 3.65v per cell it is stressful but gives maximum cell capacity for given time in boost mode. This is usually desired for off-grid PV where you want to store as much energy possible when sun is out.

Any unit that only does timer period for boost mode termination and does not have boost charge current taper-off termination for boost mode will cook battery at boost voltage for boost timer period specified regardless of actual state of charge of battery. This is not good when using LiFePO4 battery.

The main issue is a boost cycle can be re-triggered everytime the inverter draws high current, momentarily dropping battery voltage below boost retrigger voltage. This can result in battery spending a high percentage of time at 3.65v which is very stressful on cells. Hopefully the charge controller requires the boost retrigger voltage be below set level for some minimal time period so it does not initiate a boost cycle for battery voltage slump on a short inverter surge load current everytime a water pump or other high starting current device turns on. I did not see this time specified in the manual.

PV charge controller should be wired directly to battery terminals to avoid high load inverter cables voltage drop.

There is no 'standard' settings. The controller needs to be tuned based on your PV array size, battery AH size relative to inverter power and what your peak and average inverter loads are. Also, how hard do you want to be on your batteries.

Thanks!
I understand what you are meaning about that..

Voltage can drop pretty much when high load occurs, I do not have pretty high loads though, it should be the waterpump which draws about 4A or something like that.
That shouldn´t drop the voltage that much, not with a Lifepo4 anyway..
My Boost reconnect is set to 13.2V, but maybe I should go lower?
On a more cloudy day, the voltage is kind of steady at 13.2V, and it behaves pretty normal..
This screenshot is from a day with clouds all day, and I watched TV a couple of hours, working pretty good.

 

[TheeDude]

If you look at the plot below you can see the steeply rising voltage as the battery approaches full charge. Once above 14 volts, a high volts on cell 1 shuts of charging. the BMS recovers but each time the volts is greater than 14v it shuts down again ( the sampling missed the over 14 volts ).
Reduce the charge volts to 13.8 ( boost volts) and charge the battery, with for now, a long boost duration. This may help the balance and you can see via the app how things are going.
Try increasing the boost volts in 0.1 steps for subsequent charge cycles.
Note that any voltage above 13.6 v will charge to full.

Just cutting back to 14.4v wont help at this stage, lower to a value where the BMS is not tripped and hope the cell balance will improve.

View attachment 50855
Mike

Hello again!

I have tried out your advice now.
It seemed to be working just fine for the first day, until the evening, when the battery got nearly full, then it started to do the same thing again.
And the day after it did the same thing over and over again.

So, I guess the battery is faulty, or should I go even lower? 13.6V?
Oh wait, I forgot long boost duration, it was set to 10 minutes when I did this..
What do you think, does it matter? should I try to increase it to 120 min again and try?

.