JK BMS SOC - what happened?

[LucianM]

Hi all,

I'm running a 24V 280Ah LiFePO4 battery using a 200A JK BMS + 2.4 KW Easun inverter.
Recently I ran into a strange issue: I used a lawn mower machine rated at 1.1 KW for 5 min, all worked well, but then I saw the battery SOC dropped from ~80% to 3% (JK BMS app).
No error logs on the BMS or inverter.
I then charged the battery again, at 23% the cells reached 3.45V .
I have reset the cell ever voltage protection parameter and then the SOC was recalculated at 95%, which should be about right...
I do not understand what happened. Which events can cause SOC recalculation? And how did it get to this value? Do I need to check anything?

 

[DIYrich]

Do you fully charge on a regular basis? The BMS looses calibration over time and needs the full charge to reset.

 

[LucianM]

Do you fully charge on a regular basis? The BMS looses calibration over time and needs the full charge to reset.

Yes it mostly stays between 60-98%.
But it was showing 80% before, it only charged about 20% until full so the 80% was kind of correct...
How did it jump from 80% do 3% in 5 min?

 

[TomC4306]

When you change settings, recalibration of SOC is required on many BMS's. Recalibration requires a full charge, with absorption. So, voltage held, declining tail current.
Days of patience are your friend. You will not get satisfaction immediately after changing settings.
And 98% is not 100%.

 

[LucianM]

When you change settings, recalibration of SOC is required on many BMS's. Recalibration requires a full charge, with absorption. So, voltage held, declining tail current.
Days of patience are your friend. You will not get satisfaction immediately after changing settings.
And 98% is not 100%.

It seems I did not explain the situation properly.
It was calibrated, 80% was correct, proof is that it only accepted 20% more until full.
The qorstion is why it trigered recalibration during the lawn mower usage and why it only estimated 3% when in fact it was 80% full.

 

[MisterSandals]

proof is that it only accepted 20% more until full.

Just out of curiosity how did you add this 20% and how did you measure this amount?

 

[LucianM]

Just out of curiosity how did you add this 20% and how did you measure this amount?

Charging started at 3% and cells reached 3.45 V at 23%.
3.45 V means the cells are almost full as far as I know, yet the BMS showed 23%

 

[TomC4306]

Charging started from 3% and cells reached 3.45 V at 23%.
3.45 V means the cells are almost full as far as I know, yet the BMS showed 23%

To 3.45 volts and then stop immediately or did you hold 3.45 volts until current declined to near zero?
You can get a full charge at 3.45 but you must absorb there for quite some time.

 

[TomC4306]

Try charging to 3.5.
You are paying too much attention to percentage first then voltage. Pay attention to voltage first then percentage will follow.

 

[LucianM]

I don't understand why everybody is focused on the full charge discussion, that was not my question. Let's just assume the the starting 80% SOC was correct.
The question is what event could cause re-evaluating the SOC from 80% to 3%, because I suspect it's a malfunction and I would like to fix it.
All I can think of is a big voltage drop that lasted long enough for the BMS to think that this is the real battery voltage and set the SOC accordingly.

 

[JJJJ]

Sometimes a huge draw has the potential to make the batteries look lower than they are. If this is the case then the state of charge will bounce closer to the original level.

If this is the case then it might be worth considering an additional battery in parallel to share the load.

 

[MisterSandals]

The question is what event could cause re-evaluating the SOC from 80% to 3%,

Ok, lets look a this then.

I used a lawn mower machine rated at 1.1 KW for 5 min,

Do you have a link for this machine? Or the specs?

I saw the battery SOC dropped from ~80% to 3% (JK BMS app).
No error logs on the BMS or inverter.

Can you post a pic of your BMS settings, particularly the Soc related ones?

And just for good measure, a pic of your battery and inverter wiring.

 

[LucianM]

Ok, lets look a this then.


Do you have a link for this machine? Or the specs?

Can you post a pic of your BMS settings, particularly the Soc related ones?

And just for good measure, a pic of your battery and inverter wiring.

1. Makita ELM3310 series, attached specs screenshot.
2. Unfortunately it's at anoff-grid cabin and won't get there until weekend. Which settings are related to SOC? I only changed capacity - 280ah, cell type - LiFePO4, under voltage protection - 2.9V, over voltage protection 3.5V as far as I remember.
3. Attached pic.

 

[MisterSandals]

under voltage protection - 2.9V, over voltage protection 3.5V

These are unnecessarily tight for a BMS. As aconservative LiFePO4 user, i use 2.7Vpc and 3.62Vpc. The idea of a BMS is to protect the cells from going outside safe voltages (2.5V - 3.65V). Using the BMS to control your charging and day to day use patterns is just wrong.

It would be interesting to see the cell voltages (as displayed in BMS, screenshots please!) before using mower and then 5 minutes in. This would likely reveal what is going on.

Regarding your Soc settings, they could be set initially to misleading Soc estimates. Thats a great place to look. These match Soc percentages to voltage which is used until several charge/discharge cycles.

Looks like a nice and tidy build. You need a T class fuse coming off the positive on your battery.

I am not familiar with that AIO, be sure to check the battery type and any settings related to Soc. Also charge settings since it was very throttled with 3.5V ovp.

I dunno. This is how i'd proceed tracking this down.

 

[LucianM]

These are unnecessarily tight for a BMS. As aconservative LiFePO4 user, i use 2.7Vpc and 3.62Vpc. The idea of a BMS is to protect the cells from going outside safe voltages (2.5V - 3.65V). Using the BMS to control your charging and day to day use patterns is just wrong.

It would be interesting to see the cell voltages (as displayed in BMS, screenshots please!) before using mower and then 5 minutes in. This would likely reveal what is going on.

Regarding your Soc settings, they could be set initially to misleading Soc estimates. Thats a great place to look. These match Soc percentages to voltage which is used until several charge/discharge cycles.

Looks like a nice and tidy build. You need a T class fuse coming off the positive on your battery.

I am not familiar with that AIO, be sure to check the battery type and any settings related to Soc. Also charge settings since it was very throttled with 3.5V ovp.

I dunno. This is how i'd proceed tracking this down.

I only have a screenshot a few hours before the incident, attached. The battery has a few cycles and I really think it was displaying SOC correctly.
The voltages after the incident were similar, above 3.3V .
There is a fuse behind the switch, this one https://www.amazon.de/dp/B017XDIX7S/ref=pe_27091401_487027711_TE_SCE_dp_1

 

[MisterSandals]

The voltages after the incident were similar, above 3.3V .

Then i would disregard the 3% reading, it seems bogus if your cells were all 3.3V. I would guess you have an odd Soc value to voltage setting (for display only and before enough cycles for calculated Soc).

There is a fuse behind the switch

The fuse should be the closest thing to the battery to protect the wire and switch. I an not familiar with that fuse be it needs to be rated to blow before the wire and switch.

 

[LucianM]

Then i would disregard the 3% reading, it seems bogus if your cells were all 3.3V. I would guess you have an odd Soc value to voltage setting (for display only and before enough cycles for calculated Soc).


The fuse should be the closest thing to the battery to protect the wire and switch. I an not familiar with that fuse be it needs to be rated to blow before the wire and switch.

It is bogus, but I am wondering if it's just a lack of calibration or could indicate some other malfunction.

 

[TomC4306]

It is bogus, but I am wondering if it's just a lack of calibration

Hmm.
Sounds familiar.
Get a FULL hard charge on the battery FIRST, then try to replicate the "fault".

 

[cityel.uk]

I get a similar issue. Mine is in a small electric car. It goes into UVP randomly, then after UVPR it is at 0% despite the voltage indicating a charge of >80%.

 

[TomC4306]

Charge to full, tippy top.
It's not full when the meter says 100% full. It's full when the battery will take no more amps. Then the meter should reset to 100% and then you can determine where you're at with your battery. Those state of charge indicators require regular recalibration. Recalibration is accomplished by FULLY charging the battery.