Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Acceptable cell imbalance range?

This site may earn commission on affiliate links.
I had a '15 85D with similar cell mismatch for years.

It had very high range until last year when it suddenly died, replaced under warranty.

The cells are actually strings of cells in parallel, in the case of my 85 pack, 72 cells in a string..one dead cell in said string would reduce the voltage about the amount you are seeing as string mismatch.
 
Hi.
First of all thanks for your quick answer.
My Tesla is 2015 model s 70d 86k miles.
Based on Scan My Tesla app it until last month it had all cells perfectly balanced in 4-6mv range but now 20-30mv.
If your pack 60 mv imbalance is 1 dead cell, than I could have no dead but weak cell?!
I will monitor the pack in next weeks (months) to see, if it will stay in this range.

Thanks one more time
My 2015 70D is also showing about 20-25 mv balance, from an original of about 4 mv. My rated range is now down to about 195 miles, with about 162k miles on the car.
 
My 2015 70D is also showing about 20-25 mv balance, from an original of about 4 mv. My rated range is now down to about 195 miles, with about 162k miles on the car.
My rated range was 219 miles at 100% SOC but now it is down to about 215 in one month.
maybe it is also because of cold whether we have no here: day tems max 8 Degree Celsius and nights min minus 2 Degree Celsius.
any way as I understand, if my current imbalance will stay in this range, than everything will be ok.
 
I had a '15 85D with similar cell mismatch for years.

It had very high range until last year when it suddenly died, replaced under warranty.

The cells are actually strings of cells in parallel, in the case of my 85 pack, 72 cells in a string..one dead cell in said string would reduce the voltage about the amount you are seeing as string mismatch.
As I found out the battery pack module has 74p6s 444 cells (Tesla 5.3 kWh Module (85 kWh Pack) | Batteries | Stealth EV). 85d has 16 modules in all and my 70d has 14 modules.
I my problematic module all 5 brick voltages are 3.924-3.925v and only one has 3.905v. in this case the module has 20mv imbalance.
If I calculate its percentage diff between 3.925v(good brick) and 3.905v(problematic brick) it would be 100-(3.905*100/3.925)=100 - 99.49044586=0.50955414%
after that, if I take the whole brick cells (in my case 74 cells) and calculate from this we will get 74*0.50955414/100=0.377070064 cell
So, from this calculation I'm getting 0.38 dead cell. Because of this, I think, that the cell is weak or hes weak contact.
Based on this I can translate it in this: about 53mv imbalance will give us a picture of one dead cell.

BUT.... I'm not Electrician, so I'm not sure if this calculations are correct for this situation. I'm not sure if dead cell(s) can be determined with such calculations. It' could be that in case of 53mv imbalance 3 cells are weak, but not dead.
Please, correct me if this all is wrong, what I wrote here!
Thanks and Happy New Year
 
As I found out the battery pack module has 74p6s 444 cells (Tesla 5.3 kWh Module (85 kWh Pack) | Batteries | Stealth EV). 85d has 16 modules in all and my 70d has 14 modules.
I my problematic module all 5 brick voltages are 3.924-3.925v and only one has 3.905v. in this case the module has 20mv imbalance.
If I calculate its percentage diff between 3.925v(good brick) and 3.905v(problematic brick) it would be 100-(3.905*100/3.925)=100 - 99.49044586=0.50955414%
after that, if I take the whole brick cells (in my case 74 cells) and calculate from this we will get 74*0.50955414/100=0.377070064 cell
So, from this calculation I'm getting 0.38 dead cell. Because of this, I think, that the cell is weak or hes weak contact.
Based on this I can translate it in this: about 53mv imbalance will give us a picture of one dead cell.

BUT.... I'm not Electrician, so I'm not sure if this calculations are correct for this situation. I'm not sure if dead cell(s) can be determined with such calculations. It' could be that in case of 53mv imbalance 3 cells are weak, but not dead.
Please, correct me if this all is wrong, what I wrote here!
Thanks and Happy New Year
Have you tried charging to 100% yet? While the BMS now days always works on balancing out the pack, going to 100% forces a more "Thorough" job of it. It can take a long time depending on how bad. However, it does kind of "force" the BMS to bring all bricks to the same or as close of voltage as it can. It may not happen all in one charge though, it may take 3-5 100% charges in a row. I do find it takes longer and does a better job doing so on a supercharger, and cuts off charging & balancing much sooner on AC charging.
After doing it 3-5 times in a row, letting it go all the way until "Charge Complete", then drive it until it's lower SOC (10-20%) and let the car sit for 15-30 minutes after you are done driving. After it sits, connect TM-SPY or SMT again, and check voltages. If you see bricks that are lower than the rest, those either have a dead cell/blown cell level fuse, or are becoming weaker/less capacity.
If at 100%, and left for a couple hours after charging, those weak looking bricks are still holding voltage, you are fine, if after charging to 100% and checking after a couple hours, those "weak" looking bricks are starting to drop in voltage, then you have a parasitic cell, that is self discharging and slowly bringing the voltage of that brick down. Either way, not much you can do about it other than keep the battery in a charged condition and don't store it unplugged (such as leaving it at the airport) for any period of time. Doing so will eventually trigger the BMS and an error limiting max charge, basically forcing either a battery pack replacement or third party repair. Even with a parasitic cell, you can keep the pack in operation quite a long time as long as it doesn't get over-discharged.
One not best option with a parasitic cell is to "snip" the cell level fuse. This still costs money, time, or both depending on if you do it or have someone else do it. It's not a walk in the park for most people. I personally do not believe you have a parasitic cell though based on your numbers, just throwing out more information.

In all honesty, your pack is older and used. It's not just mileage but also age with lithium. You just happen to have a brick or two that are starting to show it's age. It really only takes one weak cell to start bringing down a brick. Heck, one dead cell in one single brick can reduce the entire pack capacity by 1-2%. Though additional single cell failures in other bricks will have little to no effect. Pack is only as strong as it's weakest brick.

In the end, keep an eye on it, not much you can do right now, drive it and enjoy. Don't stress over it. If/when it's ready to fail, it will fail. Not much you can do to change that. Just keep the pack charged. Charge to 100% on a regular basis, at least once/month, to help keep everything in balance and happy, and just drive on.
 
  • Like
Reactions: nicog
In the end, keep an eye on it, not much you can do right now, drive it and enjoy. Don't stress over it. If/when it's ready to fail, it will fail. Not much you can do to change that. Just keep the pack charged. Charge to 100% on a regular basis, at least once/month, to help keep everything in balance and happy, and just drive on.

Agree except:

There is no need to ever go above 93%. on a similar vintage pack, I saw no improvement in capacity or imbalance going above 90%.
You loose all Regen at 93/95%, and with some packs, the pumps run at 100% over 77%.

I would add to keep it plugged in and store at 50%-60% or what you need in case of an emergency trip, and then charging additional just before a drive. There is no doubt on just about all battery chemistries that smaller charge/discharge cycles result in a non linear benefit in reduced degradation.

This is supported by numerous studies and Elon's tweet back in 2018.
 
Last edited:
  • Like
Reactions: nicog
Agree except:

There is no need to ever go above 93%.
Well, I disagree with that, and I back my opinion with 11 years and 450,000+ miles of tesla ownership, testing, BMS access, many friends whom are currently or were Tesla Engineers & service personal & Tesla Battery refurbishers that I question regularly. Despite the updates to the BMS over the years, where as in the "old days" the BMS wouldn't do much balancing at all below 90%, so you HAD to basically do a 100% charge to even things out, again, it does balance at nearly all SOC's now days, there is still more of a limit, that it just won't balance out until it tries cramming in a full, 100% charge.
I've sat, and watched my MX take HOURS to even out the pack at 100% (At a supercharger none the less!) and it even take over 6 hours at a supercharger while balancing one back in May.
Not trying to turn this into a pissing match. I'm just trying to make it clear I back everything I do & say with real world experience.
 
Last edited:
  • Like
Reactions: nicog
Ah, edited and expounded on while I was writing my response.

Yes, some cars do run the pumps to bring it back down. Shady of Tesla to be doing that. For storage, yes, 50-60%. still, will do a better job balancing out at 100%. Obviously, do not store it at 100%. The 100% would then be used to drive the next morning. Amperage adjustment is available, so can be timed to complete closer to when it will be used.

Lose regen, big deal. Depends on location, but here in Wisconsin, I won't see very much regen until the weather warms in April, and even then, no regen after being parked overnight. And charging to 100% once/month and not having regen for a short time after is no big deal. Still helps keep things balanced out much better.
On the 2012 I sold to a friend, that is stored in my driveway 10 months out of the year, is kept at 50%. Even though it kicks in and charges every coups days to keep it at 50% SOC, I still been checking the pack with SMT & TM-SPY, and cells still are going out of balance. So every month, a 100% charge bring everything from about 30mV out of wack back to ~6-10mV.
 
  • Like
Reactions: nicog
I can only offer my own observations regarding cell balancing.

About 35,000 miles ago, my 90% charge limit held firmly at 294-295 RM. It was very solid for years. I rarely, if ever, charged and held it any higher than that. If I was going on a road trip, I would schedule it to finish at 100%, but it never stayed there long.

On this particular road trip, I left at 100% charge and before my first stop with about 11RM indicated, I got a "pull over immediately" warning with red steering wheel and that terrible alert sound blaring continuously. I was able to make it less than 1 mile to the SC exit--blew through two stop signs--and coasted in perpendicular between SC stalls. As soon as the car stopped moving, the e-brakes autolocked and car powered down. I WAS able to jump out, open the charge port and plug in and start a charge, but nothing happened for at least 10 minutes. Zero charging KW. Green flashing charge port was a welcome sign, though. Then nirvana as it ramped slooooowly up to 1kw, 2kw...5kw and so on.

The rest of the 4000 mile CC roadtrip was uneventful from a charging standpoint.

After I got home (10 days later), I charged to 100% and held it for 24 hours. BMS seemed to reset itself, and over the course of a few weeks my 90% charge dropped to 284 and holding. It took several charging cycles before it dropped to 284. Some may say that I damaged my battery, but I do not think so. I think the BMS was out of tune and needed to see bookends in order to recalibrate.


90% is still at 284 RM crossing 98,000 miles and 6+ years
 
I can only offer my own observations regarding cell balancing.

About 35,000 miles ago, my 90% charge limit held firmly at 294-295 RM. It was very solid for years. I rarely, if ever, charged and held it any higher than that. If I was going on a road trip, I would schedule it to finish at 100%, but it never stayed there long.

On this particular road trip, I left at 100% charge and before my first stop with about 11RM indicated, I got a "pull over immediately" warning with red steering wheel and that terrible alert sound blaring continuously. I was able to make it less than 1 mile to the SC exit--blew through two stop signs--and coasted in perpendicular between SC stalls. As soon as the car stopped moving, the e-brakes autolocked and car powered down. I WAS able to jump out, open the charge port and plug in and start a charge, but nothing happened for at least 10 minutes. Zero charging KW. Green flashing charge port was a welcome sign, though. Then nirvana as it ramped slooooowly up to 1kw, 2kw...5kw and so on.

The rest of the 4000 mile CC roadtrip was uneventful from a charging standpoint.

After I got home (10 days later), I charged to 100% and held it for 24 hours. BMS seemed to reset itself, and over the course of a few weeks my 90% charge dropped to 284 and holding. It took several charging cycles before it dropped to 284. Some may say that I damaged my battery, but I do not think so. I think the BMS was out of tune and needed to see bookends in order to recalibrate.


90% is still at 284 RM crossing 98,000 miles and 6+ years
You did not damage your battery. You are correct with your assumption. Very Low SOC to Very High SOC allowed the BMS to recalibrate and get a more accurate measurement of what the actual pack capacity is. When it gets so low that it shuts down, when it starts charging again, it goes through a series of checks before charging starts. Then it will of course, ramp up extremely slowly. I've seen cars hold 1kW with no range gain for 15-20 minutes. It will especially do this if "range" was used below shown 0. This is also part of the calibration. If for some reason you find yourself with it this absolute low of a charge, if you have the time to do so or are in a position to let it do it, such as just arriving home, it is a good opportunity to let it go to 100% uninterrupted to let it measure out the pack capacity.
I would not recommend doing this on purpose, sub-10% is usually good enough starting point, but it's an opportunity none the less.
 
  • Informative
Reactions: aerodyne
@islandbayy
thank you for your detailed explanation. I already charged it one time to 100% until charge was completed. The second time I charged it to 100% but had to go, so did not waited until "charge complete" indication.
Any way I'll do as you wrote and will do 100% charge until "charge complete" 3-5 times.
But I'd like to specify one thing: Have I to charge it to 100% and discharge to 10-20% this 3-5 times or can I charge it to 100% and than use for example 30% and on 70% and charge it again to 100% an so on 5 times. And only after that charge it to 100% and discharge to 10-20% and let it sit for 30 mins and get voltage readings with SMT?

@aerodyne
Thank you too for the information. The pump kicks at 81% SOC and work some time. The regen is disabled as you wrote, but now dais, in cold weather, its disabled daily, so I can live with this :)

@MN-MS100D
and than you too for sharing your experience. I think you are right - the BMS must have information from Low and High SOCs to calculate real capacity of the pack and not to under or overestimate it.

Have a nice day all of you and thanks one more time for shearing such a really good infos! :)
 
@islandbayy
thank you for your detailed explanation. I already charged it one time to 100% until charge was completed. The second time I charged it to 100% but had to go, so did not waited until "charge complete" indication.
Any way I'll do as you wrote and will do 100% charge until "charge complete" 3-5 times.
But I'd like to specify one thing: Have I to charge it to 100% and discharge to 10-20% this 3-5 times or can I charge it to 100% and than use for example 30% and on 70% and charge it again to 100% an so on 5 times. And only after that charge it to 100% and discharge to 10-20% and let it sit for 30 mins and get voltage readings with SMT?

@aerodyne
Thank you too for the information. The pump kicks at 81% SOC and work some time. The regen is disabled as you wrote, but now dais, in cold weather, its disabled daily, so I can live with this :)

@MN-MS100D
and than you too for sharing your experience. I think you are right - the BMS must have information from Low and High SOCs to calculate real capacity of the pack and not to under or overestimate it.

Have a nice day all of you and thanks one more time for shearing such a really good infos! :)
Does not need to go down to 10-20%. One time every once in a while is nice to help the BMS calculate total pack capacity more accurately, however, for these 3-5 100% charges, no need. Just needs that 100% to let the BMS do it's balancing act. The first 1-2 times it will do the majority of the balancing, but will go for only so long before it will stop charging. doing it 3-5 times in a row will let it run that top end balance more so.
 
  • Like
Reactions: nicog
Does not need to go down to 10-20%. One time every once in a while is nice to help the BMS calculate total pack capacity more accurately, however, for these 3-5 100% charges, no need. Just needs that 100% to let the BMS do it's balancing act. The first 1-2 times it will do the majority of the balancing, but will go for only so long before it will stop charging. doing it 3-5 times in a row will let it run that top end balance more so.
thank you so much. I will try it and write here what I will get after this charging process!
Happy New Year!
 
  • Love
Reactions: islandbayy
Does not need to go down to 10-20%. One time every once in a while is nice to help the BMS calculate total pack capacity more accurately, however, for these 3-5 100% charges, no need. Just needs that 100% to let the BMS do it's balancing act. The first 1-2 times it will do the majority of the balancing, but will go for only so long before it will stop charging. doing it 3-5 times in a row will let it run that top end balance more so.
charged already 2 times to 100%. until now no luck to get better imbalance.
Yesterday it was trying to balance half an hour at 100%. But today it completed charge at 99%.
Today I began to write the data and here they are:

DateTimeVoltageImbalanceSOC%Remark
30.12.2023​
10:59​
4.132​
0.043​
98.9​
15 mins after charge was complete
30.12.2023​
11:18​
4.084​
0.025​
95.6​
after driving
30.12.2023​
11:44​
4.094​
0.024​
95.4​
after sitting
30.12.2023​
12:01​
4.028​
0.031​
89.3​
after driving
30.12.2023​
13:27​
4.030​
0.025​
87.3​
after driving
30.12.2023​
21:23​
4.015​
0.028​
85.1​
after sitting in the garage

I will continue to monitor this process, but maybe @islandbayy you can already see something interesting in this data.

Happy New Year TMC :)
 
  • Like
Reactions: islandbayy
charged already 2 times to 100%. until now no luck to get better imbalance.
Yesterday it was trying to balance half an hour at 100%. But today it completed charge at 99%.
Today I began to write the data and here they are:

DateTimeVoltageImbalanceSOC%Remark
30.12.2023​
10:59​
4.132​
0.043​
98.9​
15 mins after charge was complete
30.12.2023​
11:18​
4.084​
0.025​
95.6​
after driving
30.12.2023​
11:44​
4.094​
0.024​
95.4​
after sitting
30.12.2023​
12:01​
4.028​
0.031​
89.3​
after driving
30.12.2023​
13:27​
4.030​
0.025​
87.3​
after driving
30.12.2023​
21:23​
4.015​
0.028​
85.1​
after sitting in the garage

I will continue to monitor this process, but maybe @islandbayy you can already see something interesting in this data.

Happy New Year TMC :)
Yes! Voltage is quite low after charge completion. Usually, 100% charge will have cell voltage around 4.175-ish to 4.185-ish. Keep with it. The fact that it's not "going" to 100% shows it's still trying to balance out the pack. Given the lower "Charge Complete" voltages, that would be where I believe the 1-2% loss is at. I'd keep with it. If you have the Luxury to, and get free supercharging, try letting it go to charge complete at a supercharger. I think I mentioned, it will keep balancing in my experience, for a longer period of time, at a supercharger. Another thing I forgot to mention, try adjusting the charge amperage at home, as low as you can, so it will still charge to full, but still will be done charging by the time you need the vehicle.
 
Agree the voltage is low, but I expect the pumps were running and the car was awake 15 min after charge complete.
Yes, there was probably a little load.

Also, old cells that have degraded will have a lower resting voltage than new ones.

Also @nicog did’nt specify the voltage, if it was max, min or avg. Having an imbalance of 43mV when the charging just finished means that the average voltage would be somewhere around 21-22mV below the maximim cell voltage.
 
Here are new data with average min - avg - max voltages plus today morning's valtages:


DateTimeBrick VoltageMin. VAvg. VMax. VImbalanceSOC%Remark
30.12.2023​
10:59​
4.132​
4.132​
4.168​
4.175​
0.043​
98.9​
15 mins after charge was complete
30.12.2023​
11:18​
4.084​
4.084​
4.103​
4.110​
0.025​
95.6​
after driving
30.12.2023​
11:44​
4.094​
4.094​
4.115​
4.118​
0.024​
95.4​
after sitting
30.12.2023​
12:01​
4.028​
4.028​
4.045​
4.059​
0.031​
89.3​
after driving
30.12.2023​
13:27​
4.030​
4.030​
4.053​
4.056​
0.025​
87.3​
after driving
30.12.2023​
21:23​
4.015​
4.015​
4.041​
4.043​
0.028​
85.1​
after sitting in the garage
31.12.2023​
08:28​
4.006​
4.006​
4.033​
4.035​
0.029​
84.4​
after sitting in the garage whole night
31.12.2023​
08:47​
3.978​
9.978​
4.001​
4.008​
0.030​
83.0​
after short driving

normally the pack has good imbalance values: max 8mv. but only the brick 6 in the module 4 has this problem.
I already tried to charge it with 5 amp current. will try it today too.
The supercharger is located far from me, so I will try to charge it on SP at 02.01.2023 when I will make my trip to other city. I will try to charge it full to 100%.
the voltages after charging was complete, was never higher as now. So I think this is because of 2019 update as @aerodyne mentioned.
And one thing I monitored after charge: the car is not going to sleep very long time. After I charged yesterday, it was online until today's morning and went offline only for 6 Hours today morning.
 

Attachments

  • Screenshot 2023-12-31 100851.png
    Screenshot 2023-12-31 100851.png
    157 KB · Views: 34