Seplos 200 16s queries

[Logic28]

A few days ago I started my newly built 304A Battery managed by the Seplos 200 16s BMS and was trying to obtain a full charge and full discharge so that the BMS could asses the full capacity correctly and stopped the process when it read 10% but not knowing if that was the real value.
1) Do I have to set all parameters such to reduce the cells voltage down to 2.5V in order to read the full capacity just as I've done for each cell before assembly?
2) Today during normal operation, charging/discharging, the battery suddenly dropped from 95% to 12.5% interrupting the inverter and the charger but still showing a Pack V of 53.39 and all cells were balanced to 1/2 mv deviation, all this with no reasons or warning what so ever, so I quickly disconnected it and carried on with my spare 5.2K battery.
Surely that reading of 12.5% doesn't match the voltage status.

I could not have drained it as the PV had produced today 11.9kWh and at the time only 6.1kWh had been drawn for usage and all the rest kept the battery charged up to 94% until the sudden drop.
Is this a BMS mistake perhaps?
3) if so how can I reset the BMS to a vergine status so to start from scratch by fully charging it and fully discharge it or will it fix itself after another full charge?
I really appreciate your inputs as I could not find any clue on the manual or in related threads so far.
Many thanks

 

[a619]

I'm no expert, so take this with a grain of salt. Below is my understanding:

1. No. Factory defaults should be sufficient.
2. What's the warning code? This sounds like a cell low voltage warning, but you already mentioned that all the cells are well balanced. Another possibility is that there was a large load applied and BMS could not supply sufficient current. This would be possibly caused by a bad/ loose busbar connection on one of your cells/ BMS board.
3. Not sure if there's a way to 'factory reset' BMS. Doing a full charge/ discharge cycle at something like >0.3C should be sufficient to calibrate the BMS.

 

[Logic28]

I'm no expert, so take this with a grain of salt. Below is my understanding:

1. No. Factory defaults should be sufficient.
2. What's the warning code? This sounds like a cell low voltage warning, but you already mentioned that all the cells are well balanced. Another possibility is that there was a large load applied and BMS could not supply sufficient current. This would be possibly caused by a bad/ loose busbar connection on one of your cells/ BMS board.
3. Not sure if there's a way to 'factory reset' BMS. Doing a full charge/ discharge cycle at something like >0.3C should be sufficient to calibrate the BMS.

No warnings other than Victron shutting of the inverter when it receive the "erroneous" battery SOC. All battery cells torques checked and ok, bit more difficult to check the BMS but it was built only a week ago and it was checked twice then.
I had a look at yesterday's graphs and they show that at the time of the event the battery was only discharging 33.6A at 52.29V.


I'm beginning to understand what might have caused the issue but need to recharge first and see if it goes back to normal.

 

[houseofancients]

no need to change parameters, there is very little power left at the bottom, so use the default values to "train" your bms.

check your connections, there might be a sensor wire not torqued to 4-5 Nm, the alternative is a shot cell

resetting to virgin status isn't required, the bms is trained every time you do a full charge till cut off and discharge till cut off...

not sure which version of the BMS you are using, if v2 , then i would suggest lowering the charge current from 57.2 to 56.8 so the passive balancer has the time to do it's work.
also if using something like a neey active balancer, i certainly hope that you didn't splice the sensing wires, as the bms actively adjust itself based on what it senses

 

[a619]

No warnings other than Victron shutting of the inverter when it receive the "erroneous" battery SOC.

I asked because the picture of screen you shared suggest that it had issued at least 1 warning. Weird if you don’t see that on BMS Studios or Seplos app

 

[Tulex]

no need to change parameters, there is very little power left at the bottom, so use the default values to "train" your bms.

check your connections, there might be a sensor wire not torqued to 4-5 Nm, the alternative is a shot cell

resetting to virgin status isn't required, the bms is trained every time you do a full charge till cut off and discharge till cut off...

not sure which version of the BMS you are using, if v2 , then i would suggest lowering the charge current from 57.2 to 56.8 so the passive balancer has the time to do it's work.
also if using something like a neey active balancer, i certainly hope that you didn't splice the sensing wires, as the bms actively adjust itself based on what it senses

According to the Seplos tech I dealt with, it only retrains if within 30% of current settings.

 

[a619]

According to the Seplos tech I dealt with, it only retrains if within 30% of current settings.

Can you clarify this comment? Does it mean the BMS will only train again if you charge/ discharge at > 0.3C or > 0.7C, where C being 200A here

 

[Tulex]

His wording was-

Updates are allowed only if the deviation between the new learning capacity and original battery capacity does not exceed 30%.

So I would say if someone trained the battery without fully charging it, they would get a low capacity. The only way to train it properly would be to do it in steps, charging a bit higher but not more than 30% higher.

 

[houseofancients]

His wording was-

Updates are allowed only if the deviation between the new learning capacity and original battery capacity does not exceed 30%.

are you sure he wasn't talking about the v3 version ?
haven't seen this behavior in mine ( 10E)

 

[Tulex]

are you sure he wasn't talking about the v3 version ?
haven't seen this behavior in mine ( 10E)

We were definitely working with my version, which is 10E.

 

[houseofancients]

We were definitely working with my version, which is 10E.

so same HW version....
maybe firmware difference....
mine are actually calibrated every time i do a full charge till cut of and full discharge untill cut off

 

[Logic28]

no need to change parameters, there is very little power left at the bottom, so use the default values to "train" your bms.

check your connections, there might be a sensor wire not torqued to 4-5 Nm, the alternative is a shot cell

resetting to virgin status isn't required, the bms is trained every time you do a full charge till cut off and discharge till cut off...

not sure which version of the BMS you are using, if v2 , then i would suggest lowering the charge current from 57.2 to 56.8 so the passive balancer has the time to do it's work.
also if using something like a neey active balancer, i certainly hope that you didn't splice the sensing wires, as the bms actively adjust itself based on what it senses

It's a 204-10E, thanks for your answer and, just as I assumed and in fact after a couple of hours on a slow 5A charger up to 58.4v it restored itself to 100% proving that it was indeed left at over 90% when it happened.
I haven't had a chance to change the parameters in order to make it set to 100% at other max voltages also because I really would like to drain the battery down to its min at least once in order for it to learn its own capacity.
This still does not explain why it happened so suddenly with very little discharge at the time (only 37A), it has seen double that (60+A) only one hour prior to this.
I checked all cells torque and they are correct, a bit more difficult reaching the BMS at this stage but they were torqued only a week ago and with conductive paste after thorough polishing of all surfaces.
All bus bars purchased are Flexi copper other than the one I fabricated for Battery 8 & 9 carefully trimming the size so to have the exact same resistance of the rest of them.

The passive balancer seems to work fine an I do not have more than 1/2 mv deviation however I have bought the Nee but I'm waiting to develop the two circuit boards which will replace the standard loom (harness) before I can install it.
This is how it's going to be mounted with one connector to the BMS, one to the batteries and the middle one for the Balancer if and when needed.


Thank s again for your comment, nice to belong to such an helpful fraternity.

 

[houseofancients]

It's a 204-10E, thanks for your answer and, just as I assumed and in fact after a couple of hours on a slow 5A charger up to 58.4v it restored itself to 100% proving that it was indeed left at over 90% when it happened.
I haven't had a chance to change the parameters in order to make it set to 100% at other max voltages also because I really would like to drain the battery down to its min at least once in order for it to learn its own capacity.
This still does not explain why it happened so suddenly with very little discharge at the time (only 37A), it has seen double that (60+A) only one hour prior to this.
I checked all cells torque and they are correct, a bit more difficult reaching the BMS at this stage but they were torqued only a week ago and with conductive paste after thorough polishing of all surfaces.
All bus bars purchased are Flexi copper other than the one I fabricated for Battery 8 & 9 carefully trimming the size so to have the exact same resistance of the rest of them.
View attachment 198828View attachment 198819View attachment 198829View attachment 198838
The passive balancer seems to work fine an I do not have more than 1/2 mv deviation however I have bought the Nee but I'm waiting to develop the two circuit boards which will replace the standard loom (harness) before I can install it.
This is how it's going to be mounted with one connector to the BMS, one to the batteries and the middle one for the Balancer if and when needed.
View attachment 198824

Thank s again for your comment, nice to belong to such an helpful fraternity.

To add, my max charge is 56.8 , which allows the balancer to keep up, after a while will reset SOC to 100% , and will keep my cells clear from 3.65 to have some wiggle room

 

[Logic28]

so same HW version....
maybe firmware difference....
mine are actually calibrated every time i do a full charge till cut of and full discharge untill cut off

What firmware version are you running please? I'm trying to familiarise with this subject and avoid falling in upgrades which are worst than the previous one.

 

[houseofancients]

What firmware version are you running please? I'm trying to familiarise with this subject and avoid falling in upgrades which are worst than the previous one.

Probably the oldest out there