JK BMS quite inaccurate at SoC (and also Current/Power ???) Measurement

[silverstone]

Yeah, it differs on each charge cycle. Depending on how long it's been since it was last balanced.
On a daily cycle, an hour is enough.
After a week, it usually takes 3 or 4 hours for mine.

Today I'm staying up late at night doing "small steps" of voltage change and slowly bringing the battery to those magic 56V without hitting the limit.

Currently at 55.4V or so. It starts to get faster and faster to do steps of 0.1V at least .

 

[740GLE]

Time is the solution, even a 2a active balancer will take time. I adjusted my “absorbtion” time for 4 hrs on my passive JBD and once “balance only while charging” was turned off that passive balancer was able to get the delta down to .002v at 3.45 per cell.

But also being said the best active balancer won’t fix a bad cell. It’s only a tool to polish our half decent grade b turds we buy, but it has its limitations.

 

[silverstone]

Time is the solution, even a 2a active balancer will take time. I adjusted my “absorbtion” time for 4 hrs on my passive JBD and once “balance only while charging” was turned off that passive balancer was able to get the delta down to .002v at 3.45 per cell.

But also being said the best active balancer won’t fix a bad cell. It’s only a tool to polish our half decent grade b turds we buy, but it has its limitations.

For sure. Right now I set my charger to 55.7V and it's (still) charging at like 2A. Still 123mV to balance (and 300mV of voltage to increase to get to the "magic" 3.5V/cell average).

But basically in the past I guess I kinda gave up and did the balance at 55.2V (then 55.0V, then 54.8V). Now I'm already doing a much better equalization than that.

So actually I was never doing a proper equalization for the past 4 months. Duh .

I wonder why some people had absorbtion set to 55.2V (and equalization disabled completely).

 

[Checkthisout]

I don't top balance manually before building my batteries. With active balancers it's not necessary.
And I charge to 3.5v per cell, daily.
It takes 3 or 4 charge cycles to get the cells balanced the first time, and then I'm done. (Well, not me. The balancer is done)
Quality cells make a big difference. But that probably doesn't matter either.
As long as you have active balancing. Top balancing before assembly is wasted time.
You can spend 3 days top balancing, before putting your battery to use. Or you can put your battery to use. And let the active balancer take care of it during the first 3 days of use.
Top balancing with passive balancers is beneficial. Because it's the only time your cells will be balanced.

To clarify, what I meant was you're more gentle on the cell to give the BMS time to balance the cell.

I also assume your BMS voltage cutoff is like maybe 3.65 while your charging voltage is set to 3.5.

This means the pack stays operational while giving some headroom for the new/unbalanced cell to balance.

 

[silverstone]

I'm kinda stuck at 55.7 VDC with like 4 cells at 3.580 VDC (OVP is set at 3.600 VDC). I am afraid this will take a while, especially considering cell #1 (minimum cell voltage) is like 3.432 VDC ...

 

[timselectric]

I also assume your BMS voltage cutoff is like maybe 3.65 while your charging voltage is set to 3.5.

3.6 and 3.5

This means the pack stays operational while giving some headroom for the new/unbalanced cell to balance.

On a new battery, it hits high voltage charging cutoff regularly.
Until the cells are balanced.
I don't baby my batteries at all. Bulk charge to 57v and float at 56v.
Balancing only above 3.45v per cell.

 

[timselectric]

I'm kinda stuck at 55.7 VDC with like 4 cells at 3.580 VDC (OVP is set at 3.600 VDC). I am afraid this will take a while, especially considering cell #1 (minimum cell voltage) is like 3.432 VDC ...

Give it time. The first time takes the longest.

 

[silverstone]

Give it time. The first time takes the longest.

I also think that the BMS causes a "Charge Pump" or "Capacitor Multiplier" effect.

When it's discharging the highest cell, it's lowering its voltage, thus causing some current to flow from the main supply, which kinda counteracts the charger.

I'm a bit surprised that the orange cell (lowest #4) charges before the lowest blue (lowest #1) though.

I would assume that if the balance leads were miswired would cause the delta voltage to INCREASE, not Decrease though, so I still think it's a "Charge Pump" kind of effect that happens during balance.

 

[timselectric]

View attachment 199707

I also think that the BMS causes a "Charge Pump" or "Capacitor Multiplier" effect.

When it's discharging the highest cell, it's lowering its voltage, thus causing some current to flow from the main supply, which kinda counteracts the charger.

I'm a bit surprised that the orange cell (lowest #4) charges before the lowest blue (lowest #1) though.

I would assume that if the balance leads were miswired would cause the delta voltage to INCREASE, not Decrease though, so I still think it's a "Charge Pump" kind of effect that happens during balance.

As long as the cells are in good condition, it will balance out.
The charger should maintain the overall pack voltage.
The active balancing will move small amounts of energy from the highest cell to the lowest cell.
And yes, each cell will bounce back a little after the balancer releases each time.
It just takes time. You just have to be patient.

 

[silverstone]

As long as the cells are in good condition, it will balance out.
The charger should maintain the overall pack voltage.
The active balancing will move small amounts of energy from the highest cell to the lowest cell.
And yes, each cell will bounce back a little after the balancer releases each time.
It just takes time. You just have to be patient.

I think the lowest cell went from 3.43 V (below the "steep" curve point of the voltage vs soc characteristic) to like 3.47 V (above the "steep" curve point of the voltage vs soc characteristic). And the delta_V went from 150+mV to like 105mV now. I also raised the pack voltage to 56.1 V now to have a bit more on the "well behaving" battery, since it was a bit low on the cells.

 

[timselectric]

I think the lowest cell went from 3.43 V (below the "steep" curve point of the voltage vs soc characteristic) to like 3.47 V (above the "steep" curve point of the voltage vs soc characteristic). And the delta_V went from 150+mV to like 105mV now. I also raised the pack voltage to 56.1 V now to have a bit more on the "well behaving" battery, since it was a bit low on the cells.

You're making progress.

 

[silverstone]

You're making progress.

I'm just a bit confused about the fact that, while there is definitively not sun and almost fog, the inverter seems to be getting all the power from the battery (and thus the charger and thus the grid). It's like it's ignoring completely the PV. Weird stuff on these Deye Inverters. Grid power is 1.6kW right now, should be much less given that a bit should come from the PV.

I think I also have "load first" set in the settings, but apparently if the battery voltage is ABOVE the Inverter Absorbtion / Equalization Voltage, then it just takes the energy from the battery .

A bit expensive equalization, but I guess it had to be done. I'll leave it full time tomorrow as well.

In the future maybe I just disconnect the breaker from the affected battery and come in through another connector for auxiliary power supply and charge from there. Then I can let it equalize for 3 days or so without losing on the energy production.

This just confirms me that I need that extra connector that I thought I was kicking out of the "low-cost" battery box v2 .

 

[silverstone]

72-74 mV left.

But I am worried about the last finding (Inverter ignoring PV when battery voltage is above Float/Absorbtion) ... That will have HUGE impacts on my control strategy .

 

[timselectric]

I can't help much with the inverter.
My Deye is still in its box.

 

[silverstone]

I can't help much with the inverter.
My Deye is still in its box.

Are you also going to install a Deye ? Were you in the US or EU, I cannot remember ? If US I assume you have this "split-load" type 120V/240V ?

53-55 mV left .

 

[silverstone]

@upnorthandpersonal : did you experience/notice a 0.7 V difference between the JK BMS v10 and JK BMS v11 ? I am guessing there is really some MOSFET body diode involved in the measurement circuit in one version of the JK BMS and not the other.

It's a bit of a coincidence at least that it's 0.7 V ... I originally thought it was offset or measurement tolerances. But what if the measurement circuit is different after all ?

 

[silverstone]

@timselectric 4mV delta left

3.54 V / cell on battery 1 and 3.58 V / cell on battery 2. Apparently some voltage measurement offset/method that is going on causing this issue as also the total pack voltage is different (by 0.7V ).




I think I will keep it like this for a few hours hoping that all the cells will settle.

 

[silverstone]

@timselectric But what do you set as "float" after you let your "equalization" take place ? 54.6 VDC (3.41 VDC/cell) ?

 

[upnorthandpersonal]

did you experience/notice a 0.7 V difference between the JK BMS v10 and JK BMS v11 ? I am guessing there is really some MOSFET body diode involved in the measurement circuit in one version of the JK BMS and not the other.

No, the MOSFET body diode should only there when either charge or discharge are disabled. Did you calibrate the voltage?

 

[silverstone]

No, the MOSFET body diode should only there when either charge or discharge are disabled. Did you calibrate the voltage?

Weird then that's its 0.7V. I don't believe in coincidences.

No I did NOT calibrate the voltage. Was I supposed to do that? Oups

 

[silverstone]

At least now the pack seems balanced though. Not sure how long if will last though.

 

[upnorthandpersonal]

No I did NOT calibrate the voltage. Was I supposed to do that?

In general, they're ok out of the box - but it's not a bad idea to calibrate them just in case. Use a decent multimeter, and without load. You can also calibrate the current (a 10A resistive load is what I tend to use), which should increase the accuracy of state of charge.

 

[timselectric]

Are you also going to install a Deye ?

Yes

Were you in the US or EU, I cannot remember?

US

If US I assume you have this "split-phase" type 120V/240V ?

Correct

 

[timselectric]

53-55 mV left .

Awesome
Getting there.

 

[timselectric]

@timselectric 4mV delta left

More awesome