DIY Electric Car Forums banner

1 - 14 of 14 Posts

·
Registered
Joined
·
4,326 Posts
Discussion Starter · #1 ·
closer and closer.... little tweaks with every charge, run a little, charge, measure,, tweak, and repeat....

I finally have all cells +/- .01v post charge, am I done balancing?!

My process is to allow a full charge, which ends at 139v, and then leave it plugged in and wait about 10 minutes for the pack voltage to settle to 135-136 volts where it seems to stabilize long enough for me to measure all cells without the pack voltage dropping much while I do it.

Then, I concentrated on the high cells, that were more than .01v above the theoretical average (135/38=3.55) and drain a little charge w/ resistors. I tried doing both the high cells and adding juice to the lower ones w/ power supply at the same time, but kept overshooting the next iteration... I would say that unless you have a WAY low cell to bring up, it is easier to take down the high ones. Or, at least finish with the high ones first, then catch the low ones next cycle.

so.... I now have whole pack showing between 3.53-3.55 when pack settles to v=134-135... Is this as close as I can reasonably expect, or should I stop whining and do a few more cycles to get them closer?
 

·
Registered
Joined
·
133 Posts
I'd say you are done. I've followed the same process as you have described and have most cells finishing near the terminal voltage of 3.65V, give or take 20mV.

For what it's worth, I bottom balanced the pack and used to see some cells get near 4V before the charger kicked off, but at less than an amp of charge current, I really couldn't see this as a potential problem.

However, I figured it would hurt to top balance them anyway.
 

·
Registered
Joined
·
4,326 Posts
Discussion Starter · #3 ·
...I've made a couple more iterations over the last couple days, and will take a couple rounds of measurements without any balancing just to make sure they are repeatable... then button up the tops for a month or so before re-measuring to see if there is any drift.
 

·
Registered
Joined
·
2,217 Posts
Just remember PTCrusin, since your cells are no longer balanced at the bottom you need to reliably limit discharge to prevent the smallest cells (the ones that used to go high on charge) from being reversed. That would destroy them.

I tend to favor top balancing because, if I understand the chemistry right, a lot of the capacity loss and internal resistance increase often attributed to aging is actually from time above around 3.7 volts. It seems counterproductive to subject the smallest cells to regular voltage excursions above the rest. The alternative is occasionally taking them a bit low. The catch is, it must stop at a bit low and never reverse the smallest cell (unless you get tired of it and want to match in a new cell.)
 

·
Registered
Joined
·
133 Posts
Just remember PTCrusin, since your cells are no longer balanced at the bottom you need to reliably limit discharge to prevent the smallest cells (the ones that used to go high on charge) from being reversed. That would destroy them.

I tend to favor top balancing because, if I understand the chemistry right, a lot of the capacity loss and internal resistance increase often attributed to aging is actually from time above around 3.7 volts. It seems counterproductive to subject the smallest cells to regular voltage excursions above the rest. The alternative is occasionally taking them a bit low. The catch is, it must stop at a bit low and never reverse the smallest cell (unless you get tired of it and want to match in a new cell.)

I count coulombs with an amp-hour meter and never discharge below 80% DOD. My commute usually results in 60 to 70 Ah used out of 100Ah capacity, so I'm not too worried about "finding the bottom".

I've read similar info concerning the loss of capacity if you charge well above the knee. I am considering adding two more cells to the 38 in the string. This would make the terminal voltage = 3.467V per cell instead of 3.65V in the CC/CV charge cycle. I am beginning to believe that this would still give me plenty of capacity, make the top balancing much easier, and reduce the capacity loss due to overcharging.

On a side note, we are converting one of the University's utility vehicles to electric. We are using an ac motor and controller and 16, 100Ah CALB cells. CALB says to charge to a max of 3.6V. I'm considering only charging them to 3.45v - 3.5V. Does this make sense?
 

·
Registered
Joined
·
4,326 Posts
Discussion Starter · #6 ·
I am concluding that initial balancing is best achieved well into the knee to narrow the actual state of charge to a narrow a band as possible, and then the operational charging can be set just 'a little' into the knee where the pack voltage just starts climbing enough for the charger to notice.

Sounds like the CALB curve is a little different than the TS, but in both is is amazing how fast that knee progresses once it starts climbing!
 

·
Registered
Joined
·
1,317 Posts
I've found with my headways that having all the cells under a light load (5A for a 30AH group) gave better results for measuring voltage. It got the surface charge off and let the cells settle into a real working voltage. Then measurements came within 0.003V for 3 of the 4 groups. Might be worth a try to turn your heater on or something then measure the voltage.
 

·
Registered
Joined
·
2,122 Posts
Even though resting voltage probably has some indication of SOC, I find it a very poor indication of balance across multiple cells.

Best indication of balance is the fact that cells reach the "knee" at the same time, either top or bottom, depending on which balance you are going for.

It doesn't have to be a perfect balance, don't kill yourself over it. As long as all cells reach the "knee" by the time charger shuts off, you are all set.

Definition of the "knee" area can be pretty wide from person to person and there is no point of arguing it. It also differs between cell brands. A "knee" is reached when rate of voltage rise (or fall) becomes very obvious, hard to miss. Define your "knee" range and try to get all cells within that range at the same time at the end of charge ( or discharge ).

For example, I use 3.6V-3.9V as my "knee" range for TS cells. As long as all cells are in this range at the end of charge, I consider this pack balanced.
 

·
Registered
Joined
·
2,318 Posts
I'm considering only charging them to 3.45v - 3.5V. Does this make sense?
Yes. I have 180Ah CALB cells. The knee during charging starts at around 3.45V or so. Around 1/5 Ah further charge will put them around 3.5V, and they climb in voltage faster after that.
 

·
Registered
Joined
·
2,318 Posts
Definition of the "knee" area can be pretty wide from person to person and there is no point of arguing it. It also differs between cell brands. A "knee" is reached when rate of voltage rise (or fall) becomes very obvious, hard to miss. Define your "knee" range and try to get all cells within that range at the same time at the end of charge ( or discharge ).
Here is an example of the "knee" for one of my 180Ah CALB cells (the red curve - goes with red axis numbers, green goes with green...). The voltage starts to take off a bit below 3.45V even though the charger is cutting back current at this point. That decreasing charge current is the reason the rate of voltage increase on the other two cells decreases a bit then starts to go up again as they get near the knee.
View attachment file3, cell log.pdf
 

·
Registered
Joined
·
4,326 Posts
Discussion Starter · #11 ·
I think we're all saying the same thing.... which is encouraging. ;)

I tried to define my balance procedure to bring the cells up to 'finish', and then allow them to settle just a little while still on the charger so that the pack voltage doesn't change WHILE I am measuring all the cells. I found that trying to catch the voltages right at the very top was a moving target since the pack seemed to settle from 139 to about 136 pretty fast and then hang and drop slowly to about 134 while left on the charger.

I am pretty confident that with cells +/- .01 at pack voltage of 135-136, then the furthest 'high' cell won't be more than 3.67 at most at end of charge when the 'average' is set to 3.65.

So, I'm gonna put the lids on and come back in a month.....
 

·
Registered
Joined
·
440 Posts
Here is an example of the "knee" for one of my 180Ah CALB cells (the red curve - goes with red axis numbers, green goes with green...). The voltage starts to take off a bit below 3.45V even though the charger is cutting back current at this point. That decreasing charge current is the reason the rate of voltage increase on the other two cells decreases a bit then starts to go up again as they get near the knee.
View attachment 9571
I am not sure I understand your chart. The red line (cell 1) shows discharged at about 3.28v (on left) and goes up to about 3.75v (using the y red axis on the right). On my 180 calbs I have resting voltage of 3.34v (about full) so that point on the chart is at about 1 hour 6 minutes and about 1/2 charge. I must be understanding this wrong.

Francis
 

·
Registered
Joined
·
2,318 Posts
I am not sure I understand your chart. The red line (cell 1) shows discharged at about 3.28v (on left) and goes up to about 3.75v (using the y red axis on the right). On my 180 calbs I have resting voltage of 3.34v (about full) so that point on the chart is at about 1 hour 6 minutes and about 1/2 charge. I must be understanding this wrong.
I'm not sure what you are asking spdas. Rest voltage on my cells after a full charge is about 3.38 to 3.39V. That is if I charge them to about 3.50V, with low current (< 2A) for the last half hour or so. Note that the voltages you see on the chart are while charging, probably at about 28A or so, so they are higher than rest voltage would be, and it is misleading to compare your rest voltage with voltage during charging.
 
1 - 14 of 14 Posts
Top