BMS Balancing - Current Understanding

[n2mb_racing]

General question, what is a good level of balancing? < 10mV difference between cells? Asking for my eBike battery, which is currently 0.008V difference between the max and min cell groups. Obviously a different BMS, but similar cell chemistry (LG M50LT 2170 5ah cells).

 

[mr_hyde]

No need, the car balances at all SOC levels when it’s not charging.

So the bleed resistors are going off constantly whenever the car is not charging?

 

[aerodyne]

So the bleed resistors are going off constantly whenever the car is not charging?

Unlikely unless the BMS detects a problem.

Some have experienced a high vampire drain shortly before a BMS error or pack failure.

 

[mr_hyde]

Unlikely unless the BMS detects a problem.

Some have experienced a high vampire drain shortly before a BMS error or pack failure.

I agree but evrepair said the 'car balances at all SOC levels when not charging' which is not true in my understanding. I'm hoping he clarifies.

 

[aerodyne]

I agree but evrepair said the 'car balances at all SOC levels when not charging' which is not true in my understanding. I'm hoping he clarifies.

It certainly can be assumed it does not balance when sleeping, which in my case, is 99% of the time.

 

[jensk2]

No need, the car balances at all SOC levels when it’s not charging.

Not according to my 'tests' :-(


After my previous post I have charged 78% SoC to 93% SoC with 1 Phase x 5A x 230V = 1.150 kW - Charging Power Consumption of approx 0,180 kW.

This theoretically gives 10 hours of balancing each time. My Brick Voltage ImBalance improved as follows:

After driving down to approx 22% SoC, results are here:

So ImBalance at low SoC has been reduced by 6-7 mV. But Brick 12 (which has a high CAC) is still the 'Pull over safely' brick. I propose that the graph shows, that all Modules, but 2, 5, 6, 11 and 13 has had blead resistors on for some time.

Seing that it Recell is probably right, that it balances above 80% while charging, I did a 80% -87% Charge and Immediately left for my last long trip. If 2 x 13% ~ 2 x 10 hours improved 6 mV, then 1 x 7% should improve 6 mV 7% / 26 % or 1.6 mV, which should be 'just' measureable And I would avoid charging to above 90% WHICH I HATE!

Not so, here is the improvement which is nil:

 

[jensk2]

^^^This.

Even if there was a method to rebalance the pack strings better than the BMS normally would, what is the point? If there are weak strings they will become weak and out of balance again at some point.

If I wanted my pack to have greater life and higher capacity, I would charge in smaller increments when possible, at night when it is cooler, and store inside wherever possible.

This approach is highly doable and guaranteed to be better than not.

I know I posted a lot of graphs, but if you examine my CAC versus Cell Voltage double graph, you will see that Module 2 Bricks 7 and 12 are so low in Top Voltage (4,14 V) that they will terminate the drive first (Bottom most Voltage graph) EVEN though they both hav a Capacity (CAC) above average (Module 2 CAC is 223,Ah - 224 Ah!). That is not optimal.

I will not continue to Balance in vain, because it causes degradation, so I may stop now or at least as soon as the weakets bricks (which are 13, 14, 42 and 54) are bottom balanced.

 

[jensk2]

Not according to my 'tests' :-(



Not so, here is the improvement which is nil:
View attachment 1049160

Unfortunately I changed two things when I did my last 80% - 87% SlowFox Charge!
A: I only charged 7% (80% - 87%)
B: I drove of immediately after reaching 87%

My previous two 'balancing attempts' were:
A: I charged 13% (80% - 93%)
B: I turned on Camp mode and used that to drain a few hours off, before leaving

Some other interesting findings are:
1:
- When I used Camp Mode to reduce the stress at 93% SoC, my SMT Nominal Capacity fell from 65,2 kWh to 64,4 kWh after first test and fell further to 63.4 kWh after second test!!!
- When i drove immediately after reaching 87%, then during my drive SMT Nominal Capacity grew to a new record of 65,4 kWh
That suggest that some vampire Drain is not taken into the Total Capacity :-(

That would explain why my Nominal Capacity is 65,4 kWh despite my weakest Brick having 220,2 Ah, which should give something like 3,7 V x 222,2 Ah x 6 Bricks/Module * 14 Modules = 68,4 kWh (Q: Why 3,7 Volts and not 3,6 Volts, which is average of 3,0V and 4,2V? Because energy is less at the low voltage and the BMS confirms I have 49,5% energy left when voltage is 3,675 V and I have 50,4% Left when voltage is 3,714 V. So half the energy appears to be below 3,7 Volts

2:
- After my first Balancing, my Minimum CAC improved by 0,3 Ah and my Maximum CAC stayed unchanged (ImBalance fell from 4,3 Ah to 4,1 Ah, so some rounding errors in here)
This - to me - suggest that CAC is 'calculated' with the Voltage swing currently possible After Balancing my weakest Brick 54 can go slightly lower in minimum Voltage, because it is still not the 'terminating' brick (Which is still Brick #12) The high CAC Bricks have not been blead and are still the 'EndOfDrive' bricks and so have unchanged possible voltage swing and CAC.

Makes sense of course! Calculating that brick 7 WOULD have even more than the current 224,2 Ah (around 225,8 Ah) if it was charged to 4,2 V instead of the current 4,14V is true, but adds no value.

 

[n2mb_racing]

Where are you getting all this data? SMT?

 

[jensk2]

Where are you getting all this data? SMT?

Yes, - unfortunately I have to copy+paste from screen shots as I am on the Beta version of SMT in order to read out Min/Max CAC (Calculated Amperagehour Capacity) and the Beta version has 'lost' the existing ExportToCSV functionality. If I could get SMT to my old Windows Phone I could install the NON Beta version and export there.

 

[rowdy]

Some other interesting findings are:
1:
- When I used Camp Mode to reduce the stress at 93% SoC, my SMT Nominal Capacity fell from 65,2 kWh to 64,4 kWh after first test and fell further to 63.4 kWh after second test!!!
- When i drove immediately after reaching 87%, then during my drive SMT Nominal Capacity grew to a new record of 65,4 kWh

Have you done a 100% charge recently? I found my capacity used to jump around a lot (differing by a few kWh depending on SOC), but after a proper 100% charge and balance it barely moves now and is very stable regardless of SOC.

 

[jensk2]

Have you done a 100% charge recently? I found my capacity used to jump around a lot (differing by a few kWh depending on SOC), but after a proper 100% charge and balance it barely moves now and is very stable regardless of SOC.

I have charged once to 100%, then learned from FB/Tesla BMS_u018/BMS_u029/BMS_a066 Professionals, that balancing happens while charging in the 80% - 100% SoC window.

I have then twice charged to 93% with very low Amp setting to allow prolonged Bleading. It may or may not have helped, in reality I don't know whether it was the very first 100% charge or the next 93% Charge that improved my ImBalance with 6 mV. I do know that yet a 93% and next 87% charge did nothing :-(

The 2 x 93% Charging were very close in calendar, but car was properly discharged to low SoC in between. I am tempted at trying 100% before my next longer trip, which will be in late June

 

[mr_hyde]

I believe the balancing occurs higher than 80%, but I'm not sure. I'm still looking for a definitive breakdown of when it happens - what conditions trigger it.