DIY Electric Car Forums banner

Bad cells or manual balancing needed?

5K views 19 replies 7 participants last post by  GerhardRP 
#1 · (Edited)
I have 45, 180AH CALB batteries that I have been using for 2.5 years. Each winter the resistances seem to get higher and the voltages lower. I have an Orion BMS which top balances. The BMS gives me the capability to download detailed data on cell resistances and voltages. In reviewing the data, I see that the majority of the cells will drop to 2.8-3.0 volts under load, but there are 3 cells that will range from 2.5-2.6 volts. This is with a fully charged pack. The BMS will cutoff discharge at 2.5 volts so I am starting to hit that quite often under normal acceleration. It is in the 20's and 30's where I live. I do not have a pack heater. Those same three cells will range in resistance from 2-3.5 mOhm while the rest of the pack is 1-1.8 mOhm.

Heaters might help but I don't think that is the only problem. Two of the weak batteries are on the outside of the pack but one is in the dead center.

I have never done a manual balance. Would that help? If so, what is the easiest method to do that? Should I charge those three cells on their own or do I have to disconnect every single cell and put them in parallel for a time? (seems like a lot of work)

When people decide they have a bad cell that needs to be replaced, what criteria do you use? Am I at the point that I should consider getting those cells replaced?

I have sent similar notes to the Orion BMS developer and Keegan at CALB. I'll let you know what their perspectives are.

Cell voltages including kickout


Cell voltages during 2 periods of light acceleration


Cell resistance during 1 min drive
 
See less See more
3
#3 · (Edited)
I updated the graphs in my first post to include current and provide more data on the timeline

The first graph is 9 seconds. The current ramps from 100 amps to 340 amps when it kicks out on low voltage.
The second graph is 13 seconds The current ranged from 150 to 250 amps.
The last graph (resistance) is 2 minutes, 20 seconds

You can download the Excel spreadsheet with raw data here: https://app.box.com/s/ysv8uubs6w8xv67c4po2
(This link will work for 30 days)
 
#4 ·
A quick back of the envelope calculation indicates that your cells vary between about 1.4 milliohms and 2.4 milliohms. I don't like to see a fill 0.001 ohm difference between cells, but I don't have cell level data to compare to my pack.

My pack's summer internal resistance is about 1.4 milliohms each for my 60 amp hour cells (pack average) and in the winter it jumps to around 2.4 milliohms (pack average) if I insist on driving in near freezing weather.
 
#5 ·
That is really interesting data, thanks for sharing.

In the 9 second test and the resistance chart it looks like there is a broad range of cell voltages under load. Was the pack fully charged for the 9 sec test? The cells seem a little low at ~ 3.275 with a 10 mV spread. The cells diverge under load to about 300 mV, then recover with about 50 mV spread. This seems like an excessive range under load that could be due to either damaged cells or cells not equally charged (balanced capacity).

The resistance chart shows the same thing with a more dramatic spread, but the values are very low and cluster in the same range magnitude. If a cell were damaged then i think the resistance would be out of range by an order of magnitude (factor of 10 to 100 times greater). The broad spread makes me think more unbalanced than damaged.

How do you charge--what is your charging procedure/voltage/current, shunting to top balance, etc?
 
#6 ·
Thanks for your perspective. The pack was fully charged when the data above was taken. The only abnormality is that I hadn't used the car for a week. It is parked in my garage so it is not exposed to outside temperatures around freezing but it was cold.

I have an Orion BMS. I chose it primarily because its resistance to noise as I had problems with other monitors. It has a lot of cool features such as the ability to log detailed data above. Regarding balancing, this is what I pulled from its website:

The Orion BMS uses passive balancing to remove charge from the most charged cells in order to maintain the balance of the pack. The passive shunt resistors dissipate up to 200mA per cell. While that amount may seem small, that current is more than sufficient for maintaining balance in very large battery packs. Difference in cell self-discharge rates are often measured in the tens to hundreds of uA (with a uA being 1/1000 of a mA.) Even with a very high difference in self-discharge rate of 1mA, the 200mA balancing current is still 200 times that of the discharge rate.
I don't have a single cell charger but I do have a power supply I could use to charge some of the really bad cells to see how that would impact them.
 
#7 ·
I don't see the data needed to determine if you need to manually balance. For that you need to log a charge cycle and see how much they spread at the end of charge. If they hang tight until they go over 3.45 volts each then the pack is quite well (top) balanced. Most EVs with a BMS are top balanced.

If you are bottom balanced you would need to do a full discharge to determine pack state of balance. So long as all the cells spring back up over 3.1 volts when the load is removed it isn't really possible to determine a bottom balance condition.
 
#8 ·
What are your voltage and current limits--What charger are you using?

Usually the battery manufacturer has a charging procedure to reach a full charge, such as constant current at a rate of up to C/3 until the voltage hits 3.5, then hold constant voltage at 3.5 until the current drops off to C/20. In your case that would be CC charging at 60 amps with CV cutoff at 9 amps. Then the voltage would settle to about 3.33 the next day after charging and be considered fully charged. Continuing to charge at lower than 9 amps would be overcharging by this procedure, not recommended.

If you are using a different procedure then your cells may not be fully charged (as seen by the 3.27 starting point) and they may not be equally charged or balanced (as seen by the broad range under load).

You could certainly do a manual charge of individual cells using a hefty lab power supply--i would recommend to find CALB's charging procedure. No reason to dismantle the pack if you can access all the terminals. It may be a little tedious but can be done. If the lab supply can be set for CC and CV modes it will make life easier.
 
#9 ·
I am balancing the three batteries with my 3 amp lab supply set to 3.6V. I'll take another data set after that is complete. I am not too hopeful though. Simply being out of balance would not account for a 0.2V difference at full charge.

This what CALB says regarding charging:
-Individual cell charging constant voltage: 3.6V (CCCV charging mode, charging voltage for the period which from CC mode to CV mode)
-Charging voltage for N cells connected in series:nx3.6V (CCCV charging mode, charger shall automatically convert to CCCV when individual cells voltages raise up to 3.6V)
-Cut-off charging voltage for individual cell: 3.9V (charging maximum charging voltage for individual cell charging current shall be cut-off immediately when individual cell voltage rasie up to 3.9V).
Discharging:
-Individual cell discharging alert voltage (dynamic value @0.3C)(When OCV of individual cell falls down to 3.1V, the depth of discharge has been over 85%. WHen the OCV of the individual cell falls down to 3.0V, the depth of discharge has been over 90%. Strongly recommend shallow charge and discharge. If the BMS finds individual cell voltages below 3.0V, the cell shall be charged.
-Individual cell discharging cut-off voltage: 2.5 V (dynamic @0.3C)(Minimum discharge voltage is 2.0V@20C. WHen individual cell voltage falls down to 2.0V, output current shall be cut-off immediately. If cell is over discharged seriously, the damage can be unrecoverable).

The Orion BMS controls charging. It is setup to provide full current capability of the charger (CC) until cells near 3.6V. At that point, it calculates a charging current to make sure that not cell goes over 3.6V. When the first cell reaches 3.5V, it starts balancing and it continues balancing until all cells are within 0.02V.

I have not logged data for a charge cycle since the original installation so I'll do that after I get done balancing the three and post results here.
 
#10 ·
Make sure your battery connections are clean and tight. A little bit of corrosion might cause some resistance.

Take the lowest cell (maybe the yellow one) and clean it up. If that makes a big change in its location on the graph, you may need to do that to the rest of the ones that go to low voltage under load.
 
#11 ·
I have 45, 180AH CALB batteries that I have been using for 2.5 years. Each winter the resistances seem to get higher and the voltages lower. I have an Orion BMS which top balances. The BMS gives me the capability to download detailed data on cell resistances and voltages. In reviewing the data, I see that the majority of the cells will drop to 2.8-3.0 volts under load, but there are 3 cells that will range from 2.5-2.6 volts. This is with a fully charged pack. The BMS will cutoff discharge at 2.5 volts so I am starting to hit that quite often under normal acceleration. It is in the 20's and 30's where I live. I do not have a pack heater. Those same three cells will range in resistance from 2-3.5 mOhm while the rest of the pack is 1-1.8 mOhm.

Heaters might help but I don't think that is the only problem. Two of the weak batteries are on the outside of the pack but one is in the dead center.

I have never done a manual balance. Would that help? If so, what is the easiest method to do that? Should I charge those three cells on their own or do I have to disconnect every single cell and put them in parallel for a time? (seems like a lot of work)

When people decide they have a bad cell that needs to be replaced, what criteria do you use? Am I at the point that I should consider getting those cells replaced?
Hi, We have not used CALB (only Thundersky/Wilson). I consider myself not a battery-specialist but from experience I would replace these those three cells now, in my opinion they are damaged. The specs of the other cells look normal.
 
#14 · (Edited)
I manually charged the 3 lowest cells. After charging and resting for a few days, I took it for a drive and did another recording. I found that two of the three problem cells are still the lowest of the pack under load despite being manually charged to the top of the pack. Net, I learned that balancing was not my problem.

A couple people requested voltage profile at the top of charge. I got that but it is messed up because I now I have 3 high cells (the ones I manually charged). At the moment the charger shut off, the difference between the highest battery and lowest battery was 0.07V (excluding three that were manually driven higher). The BMS balances until all of the cells are within 0.02V. When I checked in the morning, the entire pack was within 0.0195V. Excluding the 3 high batteries, the entire battery pack was within 0.009 volts. Again, I am completely convinced that balancing is not my issue.

My next step per the recommendation twright and Ewert Electric is to clean up the terminations and re-tighten. I'll post back after that is complete.
 
#15 ·
I use an Orion and it top balances the pack upon every charge (if needed). From my understanding there is no option to "manually balance" since the Orion would immediately top balance the same way it always does and nullify whatever work the "manual balance" does.

Would it be possible to move one of the temperature sensors to be directly above or beside the suspect cells so you can compare contrast that data?

josh
 
#16 ·
I cleaned and tightened the connections on the lowest battery. It seems to have helped a little bit. Instead of getting 2.5V when other batteries are above 2.8V, I am now seeing 2.6V. This will give me a few more amps before hitting 2.5V kickout. I am also asking CALB what they recommend for cuttoff voltages. Their manual says: "Individual cell discharging cut-off voltage: 2.5V(dynamic at 0.3C)(Minimum discharge voltage is 2.0@20C when the individual cell voltage falls below 2.0V, output current shall be cut-off immediately)"
 
#18 ·
Hi Jasper,
3 days of silence??
I hope you have not taken too many corrective measures yet, because I think you are in a position to shed some light on BMS usefulness. I think your batteries are out of balance, partly because you are charging in the flat region of the Voltage vs. SOC portion of the curves for these batteries. But instead of theory and typing, I wonder if you could do a series of measurements for us.

1 Fully charge your system as usual.
2. Drive a circuit which consumes about 10% of your range, recording as you did already. At the end of the circuit, rest the car for at least 15 minutes, recording. [You could just use the car normally but not recharge it immediately.]
3. Repeat step two until some battery reached less than 3 volts resting.
4.Share and we will analyse.

I bet a small beer that the high resistance cells will show the smallest capacity, not because of actual capacity, but instead undercharge.

Gerhard
 
#19 ·
Sorry for the delay in posting.

I will do the test you suggest if that will help determine the usefulness of my BMS. I assume I don't really need to do it in 10% increments right? I can just drive until I get to the point that the lowest battery has open circuit voltage under 3V, right? I have to drive 60-80 miles in the winter to get to that point with 45, 180AH cells. My normal workday doesn't usually take me to that point but I'll see what I can do. For sure, the next time I get to the end of the lowest batteries, I'll let them rest and then collect some data before charging. I am certain that the three lowest batteries will be the first to hit 3.0V and confirm that they have the lowest capacity. I am not sure how you can conclude that is due to balance or simply a weak cell when you are at the bottom of the pack.

I'll do another data record on the charge profile as well. What I seem to remember is that the charging current does drop near the top of charge but doesn't hold the top battery at 3.6V for very long before it stops charging. It continues to balance after the charge is complete. All of the cells are within 0.02V when the balancing stops. I am certain of that. I can set that tighter if necessary.

Most importantly, I sent the data to Calb batteries. They agreed to replace three cells at their cost. Impressive support considering I bought them 2 1/2 years ago.
 
#20 ·
Sorry for the delay in posting.

I will do the test you suggest if that will help determine the usefulness of my BMS. I assume I don't really need to do it in 10% increments right? I can just drive until I get to the point that the lowest battery has open circuit voltage under 3V, right?
Hi, glad you will do a test. It is true that the 10 percent intervals are not essential, but I would like to see at least three driving intervals [with a rest period between each] recorded so we can see the OCV vs. charge delivered.
Gerhard
EDIT: Also, if you could, drive a short distance with the pack fully charged and then rest, so we can see the OCV minus the surface charge at that point.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top