Acceptable cell imbalance range?

[rowdy]

Had a similar issue. Imbalance rose quite suddenly to 26-28mV. Advice from Tesla nowadays is to charge to 100% and let it sit for at least 6 hours to balance.

I was watching the levels via SMT after charging to 100%. Balancing was slow, with imbalance reducing around 1mV per hour. It took around 24 hours for imbalance to return to 4-6mV.

Coolant pumps were running the whole time. In the 24 hours the car lost 8%.

All has been good since with imbalance typically 6-8mV. Going forward I think I'll do a 100% charge every couple of months and let it sit at that level overnight.

 

[khorton]

Had a similar issue. Imbalance rose quite suddenly to 26-28mV. Advice from Tesla nowadays is to charge to 100% and let it sit for at least 6 hours to balance.

I was watching the levels via SMT after charging to 100%. Balancing was slow, with imbalance reducing around 1mV per hour. It took around 24 hours for imbalance to return to 4-6mV.

Coolant pumps were running the whole time. In the 24 hours the car lost 8%.

All has been good since with imbalance typically 6-8mV. Going forward I think I'll do a 100% charge every couple of months and let it sit at that level overnight.

Let it sit at 100% plugged in to the "charger" or unplugged?

How has the displayed miles at 100% changed since before you started this series of charges?

 

[rowdy]

Let it sit at 100% plugged in to the "charger" or unplugged?

I unplugged once it finished charging at 100%. There was a constant drain of around 250W so I assume the car would've started charging again after some hours if plugged in.

How has the displayed miles at 100% changed since before you started this series of charges?

There was an increase in capacity of around 2% (nominal capacity increased from 72kWh to 73.6kWh). Also a slight drop in CAC imbalance from 3.20aH to 3.10aH

 

[nicog]

Had a similar issue. Imbalance rose quite suddenly to 26-28mV. Advice from Tesla nowadays is to charge to 100% and let it sit for at least 6 hours to balance.

I was watching the levels via SMT after charging to 100%. Balancing was slow, with imbalance reducing around 1mV per hour. It took around 24 hours for imbalance to return to 4-6mV.

Coolant pumps were running the whole time. In the 24 hours the car lost 8%.

All has been good since with imbalance typically 6-8mV. Going forward I think I'll do a 100% charge every couple of months and let it sit at that level overnight.

Well, you mean to charge the car to 100% and let it sit 100% charged and not discharge down to 90%? will it not hurt the HV Pack?
I already charged it to 100% 5 times in a row as @islandbayy advised, but I noticed no improvement at all. Yesterday I charged it at DC Fast charger station watching levels via SMT. As I found out, from 93% to 100% charge level the problematic brick's voltage has stopped charging at 4.144v and begun to fluctuating between 4.144 - 4.146v. between this charge levels the voltage reading mostly was 4.144v and only some times it went up to 4.145v - 4.146v.

While this charging - discharging process I found out that the imbalance is going higher at the high SOCs and lower at the low SOCs.
For example the best imbalance was at SOC 39% - 19mv and the worst at 99% - 43mv. And while charging I got the max imbalance of 54mv at 100% SOC.

this is the chart of soc - imbalance correlation:

Could anyone tell me please, what it does mean?

I will try your (@rowdy) method as well to charge it to 100% and let it sit over 24 hours.

 

[David99]

As I found out, from 93% to 100% charge level the problematic brick's voltage has stopped charging at 4.144v and begun to fluctuating between 4.144 - 4.146v. between this charge levels the voltage reading mostly was 4.144v and only some times it went up to 4.145v - 4.146v.

That is physically not possible. All bricks are in series and there is no wiring that would allow some bricks or modules to be charged but not others. All the BMS can do is "bleed off' some power from individual bricks. But the amount it can do is rather small.

I think one fundamental misconception is that balancing can even out weaker and stronger cells (bricks). That's impossible. As the pack ages, some cells/bricks degrade a bit more than others. That means the cells/bricks that have degraded a bit more can hold less total energy. At this point all the BMS can do is make sure the charge of the weaker bricks overlap 100% with the stronger ones.

Ideally all bricks reach 100% state of charge at the same time and when you drive, the BMS uses the weakest brick as the determining factor for the total pack capacity. Once the weakest brick is at it's lowest level, the battery has to shut down. It is irrelevant that other bricks still have more charge left. In a chain (bricks are in series) the weakest cell determines the cut off point and thus the capacity.

When some bricks have degraded more than others, all that balancing can do is make sure all other bricks still have charge left when the weakest one is at it's lowest allowable point. Or at least have equal charge as the weakest. Also the weakest brick gets charged to true 100% because it is the limiting factor for the entire pack. If it wasn't charged to 100%, it would further limit the entire pack. So with an older pack you will see the voltages drift apart. No amount of balancing will bring back degraded capacity. You can't 'balance' more capacity into them. Seeing bricks at different voltages doesn't mean the battery is not well balanced. It might just be due to the fact that some bricks have aged more than other and as a result will drift apart in voltage both during charging and during discharging.

 

[islandbayy]

Well, you mean to charge the car to 100% and let it sit 100% charged and not discharge down to 90%? will it not hurt the HV Pack?
I already charged it to 100% 5 times in a row as @islandbayy advised, but I noticed no improvement at all. Yesterday I charged it at DC Fast charger station watching levels via SMT. As I found out, from 93% to 100% charge level the problematic brick's voltage has stopped charging at 4.144v and begun to fluctuating between 4.144 - 4.146v. between this charge levels the voltage reading mostly was 4.144v and only some times it went up to 4.145v - 4.146v.

While this charging - discharging process I found out that the imbalance is going higher at the high SOCs and lower at the low SOCs.
For example the best imbalance was at SOC 39% - 19mv and the worst at 99% - 43mv. And while charging I got the max imbalance of 54mv at 100% SOC.

this is the chart of soc - imbalance correlation:
View attachment 1005701

Could anyone tell me please, what it does mean?

I will try your (@rowdy) method as well to charge it to 100% and let it sit over 24 hours.

That is not quite making sense. As David said, no way for the car to stop charging specific bricks or modules. What is the kW charge rate when the problem brick's "stop" charging while the rest continue?

 

[islandbayy]

That is physically not possible. All bricks are in series and there is no wiring that would allow some bricks or modules to be charged but not others. All the BMS can do is "bleed off' some power from individual bricks. But the amount it can do is rather small.

I think one fundamental misconception is that balancing can even out weaker and stronger cells (bricks). That's impossible. As the pack ages, some cells/bricks degrade a bit more than others. That means the cells/bricks that have degraded a bit more can hold less total energy. At this point all the BMS can do is make sure the charge of the weaker bricks overlap 100% with the stronger ones.

Ideally all bricks reach 100% state of charge at the same time and when you drive, the BMS uses the weakest brick as the determining factor for the total pack capacity. Once the weakest brick is at it's lowest level, the battery has to shut down. It is irrelevant that other bricks still have more charge left. In a chain (bricks are in series) the weakest cell determines the cut off point and thus the capacity.

When some bricks have degraded more than others, all that balancing can do is make sure all other bricks still have charge left when the weakest one is at it's lowest allowable point. Or at least have equal charge as the weakest. Also the weakest brick gets charged to true 100% because it is the limiting factor for the entire pack. If it wasn't charged to 100%, it would further limit the entire pack. So with an older pack you will see the voltages drift apart. No amount of balancing will bring back degraded capacity. You can't 'balance' more capacity into them. Seeing bricks at different voltages doesn't mean the battery is not well balanced. It might just be due to the fact that some bricks have aged more than other and as a result will drift apart in voltage both during charging and during discharging.

The amount the BMS bleeds off is small, yes, hence the need for repeated 100% charges. "Chips" away at the imbalance a little at a time. Would be nice if Tesla gave us a service menu feature along the lines of, "Force Full Pack Balance" or something like that (with a requirement of it being at a higher SOC).

Correct that the pack is only going to run until the weakest brick is at it's shutdown level. However, having a balanced pack helps the car keep calculated out total range & capacity. Correct, you "generally" can't Balance more capacity into them, though it can true up range estimates, and keep the pack healthier. Though, if you start off with a weak brick already considerably lower than the rest of the pack, for whatever reason is causing it (Such as a brick that has cells with a higher self discharge rate causing that brick to bleed off power just from sitting), starting off with a weak brick with less power in it, will reduce range and capacity. If it can be kept closer in balance with the rest of the pack, at the top end, that WILL give you overall more range/capacity, at least in the short term after charging completes.

 

[nicog]

That is not quite making sense. As David said, no way for the car to stop charging specific bricks or modules. What is the kW charge rate when the problem brick's "stop" charging while the rest continue?

yes you are right. this makes no sense to me too, because of this I'm asking what could be causing this problem?!
I monitored the DC charging from 58% of SOC. from the beginning the KW charge rate was 34 and at the and 2-3kw at 100%.
here is the data from 90% to 100%

SOC	Charge KW
90​	17​
91​	16​
92​	16​
93​	14​
94​	11​
95​	10​
96​	9​
97​	7​
98​	5​
99​	3.5​
100​	2.5​

at 58% of SOC the imbalance was about 31 mv while charging but at the and it was 52mv.
At 100% SOC the remaining time was about 15mins 5 min and only after that I got remaining time "calculating". At this time I disconnected the charger.
I could not wait until the end of charging process. because of this I think I have to repeat this process one more time, but, as you wrote to me, I will try it on DC charger.
I agree with everything what @David99 wrote, and because of this I don't understand,
1. why this one brick was charged slowly as others (increasing of imbalance). if it is charging slowly as others, it must be the best capacity brick. am I right?
2. why it stopped charging at 4.144v at 93% of SOC. it was until 99% at 4.144v and at 99% began to fluctuate between 4.146v and back to 4.144v as the BMS (maybe MBM) was working hard to keep this brick to 4.144v.

 

[AAKEE]

Well, you mean to charge the car to 100% and let it sit 100% charged and not discharge down to 90%? will it not hurt the HV Pack?

No.

The hurt by 100% is a myth.

Theres about the same degradation at 80-90% as at 100% for normal temperatures.

You can see the degradation (calendar aging) being about the same from about 70-100%
No one recommends to always have it at 100% but the “must drive asap” is not true.

While this charging - discharging process I found out that the imbalance is going higher at the high SOCs and lower at the low SOCs.

Imbalance comes from the different capacity. If your cells voltage are “perfectly balanced” in the middle SOC, when charging the cells with less capacity will reach a higher SOC (as all cells in series reach get the same amphs), this means they also reach a higher voltage as higher SOC = higher voltage.
Thats the reason.

If you balance the cells perfectly at high SOC the balance will be higher the lower the SOC, for the same reason as above.

For example the best imbalance was at SOC 39% - 19mv and the worst at 99% - 43mv. And while charging I got the max imbalance of 54mv at 100% SOC.

this is the chart of soc - imbalance correlation:
View attachment 1005701

Could anyone tell me please, what it does mean?

The voltage curve goes faster up close to 100%. The cell that is smaller will reach a higher SOC before just like descibed above vill have the voltage go fast up but at the same time the larger cells will be on the flater part of the curve will not have the same voltage raise. This cause the increased imbalance in the top.

Balancing is only about adjusting the SOC (voltage) to the same level.

I will try your (@rowdy) method as well to charge it to 100% and let it sit over 24 hours.

 

[nicog]

The hurt by 100% is a myth.

Theres about the same degradation at 80-90% as at 100% for normal temperatures.

You can see the degradation (calendar aging) being about the same from about 70-100%
No one recommends to always have it at 100% but the “must drive asap” is not true.

that's great.

Imbalance comes from the different capacity. If your cells voltage are “perfectly balanced” in the middle SOC, when charging the cells with less capacity will reach a higher SOC (as all cells in series reach get the same amphs), this means they also reach a higher voltage as higher SOC = higher voltage.
Thats the reason.

If you balance the cells perfectly at high SOC the balance will be higher the lower the SOC, for the same reason as above.

Yes, totally agreed. As you know, I'm trying to balance the pack at 100% SOC. as @islandbayy wrote above, I'd like to have my pack healthier and my BMS calculations more correct.
But as @David99 mentioned, the BMBs and BMS are passively bleeding the high voltage bricks to balance the whole pack. So they could not prevent the problematic brick from charging. Agreed here too, But what if this BMB is defective and is trying all the time to bleed this brick and exactly this prevents this one to get fully charged. I found in other threads out, that the resistors of BMBs can get corroded or something like that.
Also, as I noticed, this last one month the car is going not so often in deep sleep as before.
You can write to me, that not the BMB is defective but there is one cell in the brick which has short and because of this it's bleeding the whole brick.
But, as I monitored the situation, I can say, that if the car goes in deep sleep, I have normal vampire drain as before. So if the cell was short, than the discharge process would continue to be there independent from sleep or awake. Am I right?

The voltage curve goes faster up close to 100%. The cell that is smaller will reach a higher SOC before just like descibed above vill have the voltage go fast up but at the same time the larger cells will be on the flater part of the curve will not have the same voltage raise. This cause the increased imbalance in the top.

This is exactly what I'm talking about. If the fact is, that the smaller cell in the pack will reach a higher SOC faster then larger one (of course it's true), than is in my case the only one brick the larger one? Exactly this case is what I can't understand: Have whole pack's bricks except one suddenly become smaller over the past month or is the BMB not working correctly to balance the Pack at the Top end?

In general my goal is to get as more info as possible to try to eliminate as more problematic points as I can without (or before) opening the pack.
because, in My country we have no Tesla Official SC and I don't like to get to someone, who learned Tesla repair by doing and has no professional knowledge to fix such kind of problems.

One more time thank @AAKEE and all of you for your time and effort to write here so valuable and informative answers.

 

[islandbayy]

It is quite possible the BMS has issues. There are plenty of reports of voltage sensors being inaccurate on BMS Boards, especially the older ones. Still, yours isn't in quite the same way. At this point, it's getting a bit hard to say without some more hands on with the vehicle, and being able to watch TM-SPY and SMT myself over a period of time.
I WAS just going to say, I may not be too far from your location in February, however, I then realized your in the COUNTRY of Georgia, not the US State ;-) So sorry, bit far for me to travel.
At this point, I'd just drive it, keep an eye on it, don't stress over it, as if the imbalance numbers are true, thats still a healthy pack for it's age. Just keep note and check to see if it's getting worse. What will happen will happen, and in your case, when the time comes, the time will come to open the pack or have it serviced.

 

[DR Danny]

Question, I have a Model S 75D from 2018, with 140.000KM on it. But it doesn't feel as fast as I think it should be.

So I connected a dongle and scanmytesla.

This was the result, so it seems there is degradation present, but what surprised me most, was the voltage. Its always just below 300V, shouldn't this be around 400V?

 

[DR Danny]

Cell imbalance is around 70mV

 

[AAKEE]

Question, I have a Model S 75D from 2018, with 140.000KM on it. But it doesn't feel as fast as I think it should be.

So I connected a dongle and scanmytesla.

This was the result, so it seems there is degradation present, but what surprised me most, was the voltage. Its always just below 300V, shouldn't this be around 400V?

I guess the picture are not from the same time?


Not knowing the 75 battery I think they might be less cells in series.
the normal 400V standard is 96 cells innseries, producing 403V at 100% SOC.

There is some batteries with lower number of cells in series, like 84 cells in series (made by using 14 modules instead of 16). These should hold 352 Volts at full charge.

From the pictures, this do not match. 294V and average 3.84V/cell do not match.
Charge to 100% and show pictures with the same data, plus additional from the BMS tab that show all cells in series, then we see the number of cells in series and also the individual cell voltages.
Dont be afraid for charging full. As it seems, your battery would benifit from balancing the cells so you might wanna leave the car parked with no sentry for like 4-5 hours after the charge is full, to let it balance the cells.

 

[DR Danny]

I guess the picture are not from the same time?


Not knowing the 75 battery I think they might be less cells in series.
the normal 400V standard is 96 cells innseries, producing 403V at 100% SOC.

There is some batteries with lower number of cells in series, like 84 cells in series (made by using 14 modules instead of 16). These should hold 352 Volts at full charge.

From the pictures, this do not match. 294V and average 3.84V/cell do not match.
Charge to 100% and show pictures with the same data, plus additional from the BMS tab that show all cells in series, then we see the number of cells in series and also the individual cell voltages.
Dont be afraid for charging full. As it seems, your battery would benifit from balancing the cells so you might wanna leave the car parked with no sentry for like 4-5 hours after the charge is full, to let it balance the cells.

Thnx!
I will do so tomorrow, I'm having problems with the battery heating, so charging it full is a big challenge at temperatures below zero (Celsius).
Service is planned to fix this, but until its fixed... Takes some time.

Anyway, I'll get back to you when I have the possibility to charge it full.

 

[nicog]

It is quite possible the BMS has issues. There are plenty of reports of voltage sensors being inaccurate on BMS Boards, especially the older ones. Still, yours isn't in quite the same way. At this point, it's getting a bit hard to say without some more hands on with the vehicle, and being able to watch TM-SPY and SMT myself over a period of time.

I think this is the BMS (module 4 BMB). I read a lot about it and found out, that the CONDITION X or CONDITION Z could be the problem of my pack. If it is X, than I must always keep an eye on my pack. In other case it may get discharged fully.
But if it is condition Z, than, it could be self correcting issue.
In any case I will continue to monitor the battery.

I WAS just going to say, I may not be too far from your location in February, however, I then realized your in the COUNTRY of Georgia, not the US State ;-) So sorry, bit far for me to travel.

Thank you very much for your helpfulness. But as you already wrote I'm far away from US state of Georgia

 

[MP3Mike]

This was the result, so it seems there is degradation present, but what surprised me most, was the voltage. Its always just below 300V, shouldn't this be around 400V?

No, the 75D packs have two fewer modules, and each module is about 25v. So, the 75D packs top out at about 350v.

 

[islandbayy]

Cell imbalance is around 70mV

75 batteries are a 350v battery as MP3Mike says. That is only when at 100% charge (actually ~353v, which can vary a bit due to age, and balance). Considering the car's battery is low on charge, voltage is correct. Charge to 100% and you will see a voltage near 350ish.

The number of cycles on your pack is fairly low. Both my cars are well over 1000 cycles. I have the same nasty imbalance on my 75D MX that has a dead cell in one of the bricks. I'd imagine you have the same issue with that imbalance. Please post some screenshots of the brick voltages.

Also, charge up to 100% a couple times in a row. As others mentioned, we think your batter will benefit from it given the imbalance. It will help rule out cell level failure VS just being really out of balance.

 

[ran349]

Cell imbalance is around 70mV

Besides what others have said here, you show you were only charged to 20%. Many people have shown their cars to have large cell imbalances at low SOC, but it improves significantly at higher SOC.
I don't know what exactly that means because my car shows about the same cell imbalance regardless of SOC level, now about 20-25 mv, but my battery degradation is fairly high.

 

[DR Danny]

I guess the picture are not from the same time?


Not knowing the 75 battery I think they might be less cells in series.
the normal 400V standard is 96 cells innseries, producing 403V at 100% SOC.

There is some batteries with lower number of cells in series, like 84 cells in series (made by using 14 modules instead of 16). These should hold 352 Volts at full charge.

From the pictures, this do not match. 294V and average 3.84V/cell do not match.
Charge to 100% and show pictures with the same data, plus additional from the BMS tab that show all cells in series, then we see the number of cells in series and also the individual cell voltages.
Dont be afraid for charging full. As it seems, your battery would benifit from balancing the cells so you might wanna leave the car parked with no sentry for like 4-5 hours after the charge is full, to let it balance the cells.

Okay so I charged to 100% and let it be, but it doesn’t get above 97% and says it’s finished charging.

When I check scanmytesla:

Cell information: