Chevy Bolt EV Forum banner

81 - 100 of 108 Posts

·
Registered
Joined
·
239 Posts
The way I understand it, the weak cell's voltage changes faster during charging and discharging. It goes up faster during charging, so it becomes the highest cell. The strong cell's voltage goes up slower and becomes the lowest cell. During discharging, the weak cell becomes the lowest cell, and strong cell becomes the highest. Somewhere in-between the voltage spread narrows and then widens.

It is the weakest cell that we need to watch out. It determines when the charging and discharging stop, and hence the usable capacity of the battery pack.

-TL

Sent from my SM-N960U using Tapatalk
 

·
Registered
Joined
·
1,368 Posts
The way I understand it, the weak cell's voltage changes faster during charging and discharging. It goes up faster during charging, so it becomes the highest cell. The strong cell's voltage goes up slower and becomes the lowest cell. During discharging, the weak cell becomes the lowest cell, and strong cell becomes the highest. Somewhere in-between the voltage spread narrows and then widens.

It is the weakest cell that we need to watch out. It determines when the charging and discharging stop, and hence the usable capacity of the battery pack.

-TL

Sent from my SM-N960U using Tapatalk
If that is the case, since my data was from a DCFC charging session, the cell I have labeled "high" should be labeled "weak" and the one I have labeled "low" should be labeled "strong". I would have to do some data logging on a loooong drive to get data under discharge load to see if the "high" cell is actually lower voltage than the "low" cell under load.

Later,

Keith
 

·
Registered
Joined
·
6,256 Posts
The way I understand it, the weak cell's voltage changes faster during charging and discharging. It goes up faster during charging, so it becomes the highest cell. The strong cell's voltage goes up slower and becomes the lowest cell. During discharging, the weak cell becomes the lowest cell, and strong cell becomes the highest. Somewhere in-between the voltage spread narrows and then widens.

Not sure what you think defines a cell as "weak" or "strong." I have been monitoring our Bolt's cells for over two years. The highest, and lowest voltage cells have always remained the same two, whether at 100% SOC or empty. The lowest voltage cell is of lower capacity, and falls over the voltage curve knee first at the bottom of charge.

12-29-18-1.jpg 12-29-18-2.jpg 12-29-18-3.jpg
 

·
Registered
Joined
·
240 Posts
At the end of each charge cycle (hill top reserve, or full charge) the BMS conditions the battery temperature. If it is cold, it heats it until it reaches 60F... not sure what temperature it cools it down to in hot weather charging.
Interesting - that explains the tail at the end of a full charge [full being whatever you have set it to]. Of course now I have to verify that by looking at the battery V when it starts to taper and when it is done. Or just wait until Boston gets over 60F, and the tail should go away...
 

·
Registered
Joined
·
239 Posts
Not sure what you think defines a cell as "weak" or "strong." I have been monitoring our Bolt's cells for over two years. The highest, and lowest voltage cells have always remained the same two, whether at 100% SOC or empty. The lowest voltage cell is of lower capacity, and falls over the voltage curve knee first at the bottom of charge.

View attachment 28802 View attachment 28803 View attachment 28804
Weak cell is the one with slightly less capacity. For the same amount of charge, its terminal voltage is higher than the others.

-TL

Sent from my SM-N960U using Tapatalk
 

·
Registered
Joined
·
1,368 Posts
Interesting - that explains the tail at the end of a full charge [full being whatever you have set it to]. Of course now I have to verify that by looking at the battery V when it starts to taper and when it is done. Or just wait until Boston gets over 60F, and the tail should go away...
When it gets warm, the AC system will still be running at the end of a charge to cool the battery :) The way to check now, if you have torque pro is to just look at the battery heater KW reading... and in warm weather look for the AC system RPM.

Keith
 

·
Registered
Joined
·
6,256 Posts
Weak cell is the one with slightly less capacity. For the same amount of charge, its terminal voltage is higher than the others.

-TL

Sent from my SM-N960U using Tapatalk
Please explain why any cell would have a lower final voltage on a top balanced pack. The whole point of top balancing is to bleed off excess charge from each, and all cells at the end of charge, so that they all end up at the exact same voltage. The extent to which this is accomplished is a reflection of the level of tolerance allowed in the electronic BMS components. It suggests nothing about the capacity or impedence of the individual cells.
 

·
Registered
Joined
·
66 Posts
Balancing will technically allow all cells to have the same state of charge (theoretically 100%). That just happens to be about the same voltage for every cell since all cells in the pack are the same chemistry.

When balancing the current is bled off from the weaker cells, which have already reached 100% capacity, to allow the stronger cells to charge up to 100%. The pack is always as strong as the weakest cell (or group of parallel cells, in many cases).
 

·
Registered
Joined
·
6,256 Posts
Balancing will technically allow all cells to have the same state of charge (theoretically 100%). That just happens to be about the same voltage for every cell since all cells in the pack are the same chemistry.

When balancing the current is bled off from the weaker cells, which have already reached 100% capacity, to allow the stronger cells to charge up to 100%. The pack is always as strong as the weakest cell (or group of parallel cells, in many cases).
Weak and strong implies some of these cells are inferior in some way. They are not. They vary in capacity. Do you guys make fun of short people? :)

However, you are correct that the lowest capacity cell, of all 288 cells in the pack, determines the capacity for the entire pack
 

·
Registered
Joined
·
66 Posts
Unlike people, no two cells are created equal, although many are very close. Fortunately they don't have political affiliations or constitutional rights ?
 

·
Registered
Joined
·
1,368 Posts
Unlike people, no two cells are created equal, although many are very close. Fortunately they don't have political affiliations or constitutional rights ?
Why do you say fortunately? We need to organize a coalition of privileged suburban youth to fight for the rights of underprivileged battery cells! How else can we make ourselves feel special?

Keith
 

·
Registered
Joined
·
239 Posts
Please explain why any cell would have a lower final voltage on a top balanced pack. The whole point of top balancing is to bleed off excess charge from each, and all cells at the end of charge, so that they all end up at the exact same voltage. The extent to which this is accomplished is a reflection of the level of tolerance allowed in the electronic BMS components. It suggests nothing about the capacity or impedence of the individual cells.
The "capacity-challenged" cell hits the max voltage of 4.167V while the "normal" cell is still below that voltage. A resistor is then switched to be in parallel with the weak (clumsy to be politically correct) cell. Not to bleed off its charge but to make its charging current to zero as it is already fully charged. The strong cell is still being charged. This goes on untill all cell voltages are within tolerance, 20mV in my case. So the strong cell ends up with lower voltage. I believe the charging system also counts columbs(Ah). It may stop charging when it reaches certain limit.

I have been collecting data on torque pro. The weak cell's voltage indeed goes down faster than the strong cell. At about soc of 30%, the voltage difference is zero. That's based on no-load condition (car stationary). When the car is moving (discharge and regen) the voltage difference actually varies quite a bit more in either directions, as much as 70mV. But very quickly it gets "regulated" back to 20mV ish. That makes me believe load balancing is in action.

Will put up the plots later.

-TL


Sent from my SM-N960U using Tapatalk
 

·
Registered
Joined
·
6,256 Posts
Political correctness has nothing to do with it. It is about precision of language. "Weak" is not an engineering term. It isn't meaningful or helpful. Low capacity, high internal impedance, shorted...these are descriptions of an actual condition.

" A resistor is then switched to be in parallel with the weak (clumsy to be politically correct) cell. Not to bleed off its charge but to make its charging current to zero as it is already fully charged."

OK. It is diverting the incoming current. Sorry for using imprecise language. We all get lazy.

I would be amazed if they were balancing cells in the middle of the charge. What you are seeing is the cells voltage rebounding at rest. Sorry.I don't actually understand the chemistry that goes on inside a cell, but battery engineers say this happens.

My experience does not support your theory. My highest reading cell, which according to you should be my lowest capacity cell, is never the first to have a large voltage drop at the bottom of charge, but precisely the opposite.

12-29-18-1.jpg 12-29-18-2.jpg 12-29-18-3.jpg
 

·
Registered
Joined
·
239 Posts
It is not only a theory, but also data. A few more days I will have it ready.

-TL

Sent from my SM-N960U using Tapatalk
 

·
Registered
Joined
·
6,256 Posts
It is not only a theory, but also data. A few more days I will have it ready.

-TL

Sent from my SM-N960U using Tapatalk
Your theory may exactly fit your data from your battery, but it does not fit data from my battery. I wish there were some actual battery chemists/engineers on this forum.
 

·
Registered
Joined
·
6,256 Posts
Since we see completely opposite results, I would suggest that my theory, that full charged, balanced voltage is the result of random variation in balancing components, and that actual capacity difference is observable only at the bottom of charge is a better fit for the data.
 

·
Registered
Joined
·
239 Posts
Since we see completely opposite results, I would suggest that my theory, that full charged, balanced voltage is the result of random variation in balancing components, and that actual capacity difference is observable only at the bottom of charge is a better fit for the data.
Not quite completely different. Here is what I see from your numbers.

Soc_100%, #16_4.180V, #70_4.155V, delta_25mV

Soc_50%, #16_3.662V, #70_3.650V, delta_12mV

Soc_2%, #16_3.351V, #70_3.241V, delta_10mV

The voltage difference is shrinking as the battery discharges. It just doesn't go down to zero or cross into negative. But the trend is similar.

The other observation is that at 100%, #16 has 4.180V. It is a bit too high, is it not?

-TL

Sent from my SM-N960U using Tapatalk
 

·
Registered
Joined
·
239 Posts
These guys are pretty much experts on the subject and use the terms weak and strong to describe cells - Active Battery Cell Balancing | Analog Devices

This also describes the cell imbalance over charge/discharge without balancing.
Thanks for the link. I like, and have used, analog devices' power management products. They make good stuff. The article matches what I think, so one more data point.

-TL

Sent from my SM-N960U using Tapatalk
 

·
Registered
Joined
·
6,256 Posts
Not quite completely different. Here is what I see from your numbers.

Soc_100%, #16_4.180V, #70_4.155V, delta_25mV

Soc_50%, #16_3.662V, #70_3.650V, delta_12mV

Soc_2%, #16_3.351V, #70_3.241V, delta_10mV

The voltage difference is shrinking as the battery discharges. It just doesn't go down to zero or cross into negative. But the trend is similar.

The other observation is that at 100%, #16 has 4.180V. It is a bit too high, is it not?

-TL

Sent from my SM-N960U using Tapatalk

Sorry. Look again. The 1.6% SoC delta is 0.110 mV. As for 4.180V being too high. It is the result of a cheap BMS, which apparently GM engineers figure is good enough to keep them within budget, and the 60% degradation warranty.
 
81 - 100 of 108 Posts
Top