Wiki Sudden Loss Of Range With 2019.16.x Software

[MP3Mike]

I don't KNOW, but that doesn't make sense unless Tesla also confirmed:

You don't need to keep the battery balanced.
Or
You MUST charge above 85% and confirm your battery has balanced.
Or
M3 has completely different system of balancing such that (for example) it can selectively discharge different sections of the battery into the drive motors while driving, then I can see that MAYBE it is less important to bother balancing at lower SOC.

The Model 3 can not draw power from specific bricks, it is just like the S&X in that you just draw from the pack as a whole. Here is the section from the theory of operation in the Model 3 service manual:

Brick Balancing

Note that the capacity of a pack is limited by the brick with the lowest capacity. When that brick is charging, it will gain voltage faster than other bricks. The HVBMS will stop charge when any brick reaches its ceiling voltage (~4.2V). If one brick has a significantly lower capacity that others, the pack will be limited by that brick which will get to 4.2V faster than the other ones. We refer to the brick with the lowest capacity as: min CAC. In periscope, its value can be seen by viewing the signal: 'BMS_cacMin'

Another limitation could come from bricks being imbalanced, or some bricks with a voltage higher/lower than others. This would limit ability to charge the pack as the brick with a higher voltage than others would reach the ceiling voltage early. Same idea when discharging, the brick with the lowest voltage would hit the floor voltage early which would cause the HVBMS to open contactors from low power

To mitigate this imbalance, Batman has some bleed resistor that can be placed and removed in parallel of each brick via a FET relay. Batman can put that resistor across the brick with the highest voltage which would slightly discharge that brick and bring it back to the level of the other bricks. Batman closes a FET which puts that resistor across the brick. The HVBMS will order Batman to put that bleed resistor across the brick with the highest voltage when Delta V is > 5mv MinBrickV > 4.0v (~85% SOC) && HVBMS State == STAND BY.

In the past @wk057 had said that the S&X didn't start balancing until a ~93% SOC, but that could have changed since then.

 

[Battpower]

In the past @wk057 had said that the S&X didn't start balancing until a ~93% SOC, but that could have changed since then.

Exactly why I was treading carefully

But imo it wouldn't make sense.

 

[Battpower]

However, I don't know if that is a specific algorithm of the BMS, or just "natural" behavior of the higher voltage cells providing more energy during discharge.

In some circumstances, (as some cells start to age and therefore store less energy during charging), the voltage of those 'less charged' bricks actually could be higher than that of a more evenly charged brick. So in that case, high voltage would not imply high charge or high energy stored. So if that is the case, those higher voltage bricks actually have less energy to give rather than any intentional process being in control.

 

[Battpower]

The Model 3 can not draw power from specific bricks, it is just like the S&X in that you just draw from the pack as a whole. Here is the section from the theory of operation in the Model 3 service manual:



In the past @wk057 had said that the S&X didn't start balancing until a ~93% SOC, but that could have changed since then.

So that does look just like the S, except to deal with some situations that could be more likely with high charge rates, the balancing system may have been uprated in some way on newer packs / models.

 

[Battpower]

Moving on from Shakespeare and Monkeys, I'm now getting more of an Agatha Christie / Poirot moment as all the suspects gather to debate the clues and make the final accusation!

All of this makes sense for changes in Regen of course. At low temperatures, I believe (possibly even mentionned in this thread) internal resistance is higher, possibly even more so for cells that are aging worse. (Heating to reduce internal resistance would match ludicrous mode etc).

Good cells convert more of the charging current into stored charge for later use while weaker ones presumably dissipate more energy as heat during charging, retaining less stored energy (leaving imbalanced bricks). The high power 'charging' peaks that regen can deliver would be especially hard on the battery, I guess with the potential to push imbalances higher and higher giving the balancing system no chance of rebalancing - especially with batteries showing higher levels of degradation.

So basically with a cold battery - even a new one - you apparently can't dump huge current peaks through the pack without causing problems. At somewhere around 20 Celcius (guess from observations) the cells start being able to absorb energy / charge more evenly and therefore not push the pack too far out of balance.

As a battery gets older, potentially normal deterioration makes all these factors more emphasised.

 

[wk057]

Balancing used to only happen at a higher SoC, but some years ago Tesla revamped the BMS and it balances as needed across almost the whole SoC range based on accurate data calculated for each brick. (As in, if it calculates that a brick will go out of balance, it proactively balances it when there is a load/charge and such to keep it in line, as it has enough data to calculate out how much use of the balance bleeders is needed).

(Edit: I did post this info in various places over the past few years, but seems the old info is what gets found more often.)

 

[David99]

Balancing used to only happen at a higher SoC, but some years ago Tesla revamped the BMS and it balances as needed across almost the whole SoC range based on accurate data calculated for each brick. (As in, if it calculates that a brick will go out of balance, it proactively balances it when there is a load/charge and such to keep it in line, as it has enough data to calculate out how much use of the balance bleeders is needed).

(Edit: I did post this info in various places over the past few years, but seems the old info is what gets found more often.)

Yes I remember you making that updated post but it didn't get as much attention. I assume the BMS just can't bring my cells back in balance because those weak bricks are 'leaking' (weakShort message). Funny enough, that weak short message was there for months. I saw a while ago just never posted it here. 2 days after I posted it here publicly, it disappeared. I wonder if Tesla did that.

 

[Battpower]

@wk057, do you know if it is a continuous drain due to an internal shorting effect, or a self balancing effect as charge balances between stronger and weaker cells that is more likely responsible? The voltage going higher during charging for a given net charge into a brick doesn't sound primarily 'shorty'.

 

[JustADude]

Balancing used to only happen at a higher SoC, but some years ago Tesla revamped the BMS and it balances as needed across almost the whole SoC range based on accurate data calculated for each brick. (As in, if it calculates that a brick will go out of balance, it proactively balances it when there is a load/charge and such to keep it in line, as it has enough data to calculate out how much use of the balance bleeders is needed).

(Edit: I did post this info in various places over the past few years, but seems the old info is what gets found more often.)

Does this only apply new newer packs with a more advanced bms?

Or does this affect all vehicles? Even the earliest ones?

 

[Dave EV]

Balancing used to only happen at a higher SoC, but some years ago Tesla revamped the BMS and it balances as needed across almost the whole SoC range based on accurate data calculated for each brick. (As in, if it calculates that a brick will go out of balance, it proactively balances it when there is a load/charge and such to keep it in line, as it has enough data to calculate out how much use of the balance bleeders is needed).

(Edit: I did post this info in various places over the past few years, but seems the old info is what gets found more often.)

Yes, I recall you saying this, but I even though it can do this, exposing the pack to a wider range of SOC and module/pack voltages should give the BMS better information since lithium cell voltages are relatively flat from 20-80% compared to 0-20% and 80-100%. Maybe it is that good, but then why are some packs getting out of balance?

The remaining question that remains unanswered, is does Tesla still top-balance so that all it attempts to make all modules hit the maximum voltage at the same time, even as the pack ages and modules hold varying amounts of energy?

Logically, I can't think of any reason not to top-balance in this manner compared to say balancing at 50% (at 50% SOC all modules voltages are the same) or bottom balancing (at 0% SOC all module voltages are the same).

 

[DILIE]

Yes, I recall you saying this, but I even though it can do this, exposing the pack to a wider range of SOC and module/pack voltages should give the BMS better information since lithium cell voltages are relatively flat from 20-80% compared to 0-20% and 80-100%. Maybe it is that good, but then why are some packs getting out of balance?

The remaining question that remains unanswered, is does Tesla still top-balance so that all it attempts to make all modules hit the maximum voltage at the same time, even as the pack ages and modules hold varying amounts of energy?

Logically, I can't think of any reason not to top-balance in this manner compared to say balancing at 50% (at 50% SOC all modules voltages are the same) or bottom balancing (at 0% SOC all module voltages are the same).

Ideally both top and bottom, right? To get the maximum power into the battery and then get it all back out, hitting 0 with all modules essentially "empty" at whatever the minimum power limit is.

 

[gasisnasty]

This post seems to have wandered off course.
It was originally about the sudden loss of range, about 25-30 miles, in Model S Teslas.There was a class action lawsuit filed last year by Attorney Edward Chen. However all his business phone numbers have been disconnected. Can anyone provide an update on the lawsuit and what happened to Edward Chen?

 

[Battpower]

This post seems to have wandered off course.
It was originally about the sudden loss of range, about 25-30 miles, in Model S Teslas.

It does keep wanderring, but partly because the thread is a mess and how anyone wanting to actually understand anything is supposed to dive into nearly 13000 posts (and countimg....) of which may be 10% are solid facts supported with evidence and the rest made up of others trying to address their own personal issues / perspectives given what they can understand of the technical info.

The sudden range loss appears to be a response by Tesla to deal with changes that have taken place in some batteries with use. Without getting into semantics, the battery (as a unit, including electronic controls etc) was found to be degrading in a manner that required changes to its operating parameters to maintain optimal performance, given the status of a given battery. IMO some of the same lessons learned were also applied to the fleet at large as they would potentially extend the life of the battery. (regen & charging profiles)

I suppose you could view this as trying to coax batteries past end of warranty periods, or making sure every battery delivers the longest possible service life.

 

[Battpower]

exposing the pack to a wider range of SOC and module/pack voltages should give the BMS better information since lithium cell voltages are relatively flat from 20-80% compared to 0-20% and 80-100%. Maybe it is that good, but then why are some packs getting out of balance?

I believe I am starting to appreciate more of @wk057's more recent posts, esp regarding charge-counting vs voltage limits etc.

Is there any suggestion or evidence that 'weak bricks' are any better or worse at absorbing charge when they are at certain states of charge? I would suggest that brick performance data gained over the more linear regions would be easier to process and maybe more reliable, so why bother about what goes on at the extremes (other than to protect the weakest bricks). I suppose you do get to see the weakest bricks at their very worst at those extremes, but my guess is that by then you are well beyond what the balancing system can deal with so it would just be data without a possible corrective measure.

.... question that remains unanswered, is does Tesla still top-balance so that all it attempts to make all modules hit the maximum voltage at the same time, even as the pack ages and modules hold varying amounts of energy?

I feel we do have an answer to that. Yes, as @wk0057 just clarified, balancing is over a wide range of SOC. While voltage readings can under certain conditions suggest the SOC, they are not what balancing the battery is about. While charging, high voltages are probably not telling you which bricks are most charged. It could be the exact opposite.

"does Tesla still top-balance" - yes, it keeps trying to maintain balance whenever it can - including at higher SOC's. While charging, that will be right up to the point that the weakest brick reaches max permitted voltage. ANY charge that continues to flow through that brick - (The BMS can't take a brick out of the charging circuit completely) - will keep nudging that brick voltage even higher. How long you can keep charging all the bricks is dictated by how long your weakest cell voltage will allow charging to continue. You need the weak cell voltage to increase because that has to happen as part of charging. What the balancing would be doing is leaving the weak cells to do the best they are capable of (can't improve on that, it is what it is) but reduce the charging current through strong cells so that they get less charge.

In this way, the balancing dissipation is actually not a function of just the one / few resistors balancing the weak cells, but of all the other resistors shunting current away from the better cells and giving the weaker ones longer to charge.

@wk057 reminded recently that (if I understood correctly) balancing goes on (most) of the time as long as there is current flowing - in or out of the pack.

The maximum amount of energy you can bleed off is dictated by time (charging or discharging - but not just sitting in deep sleep) and the dissipation capability of the balancing circuit (FETs and resistors), and the only one of those we can control is the time.

"as the pack ages and modules hold varying amounts of energy" So the two things we can do to help are:

1) Reduce the amount of imbalance we create. Any situation with 'High power for short period of time' gives the balancing system the hardest job as it then has to (over extended time at relatively low power) redress the inevitable imbalace that will have been caused. I can't test this idea on a 'problem' battery, but if running the battery to low SOC creates greater imbalance - which it is likely to do imo - then it is NOT a good idea as it's just making more work for the balancing process.

2) Charge at a slow rate for as long as needed (or for some who hit max volts on a weak brick early, as long as possible) - to give balancing the longest time to dissipate 'imbalance energy'.


"attempts to make all modules hit the maximum voltage at the same time"

No, I do not see that anything is trying to do this. Infact, while charging, unless every cell in your battery was perfect and identical, seeing every brick voltage the same would imply that stored energy was NOT balanced between bricks. What balancing is trying to achieve is even distribution of charge / energy throughout the battery. It has to give up trying once the weakest brick either reaches max permitted voltage at the top end or lowest permitted voltage as the bottom. But all the time the battery is in use, there is evidence / data being produced to predict which bricks will need most balancing so the BMS can keep working on that whenever possible.

 

[Battpower]

@Chaserr, if you are still tuned in, as the picture has (for me at least) clarified to a point that I feel I have a reasonable handle on most (though probably not all) that's going on, it does look more like 'life with lithium batteries' rather than some specific dangerous situation that took everyone by surprise. With gasoline, how long did it take for issues and compromises to work their way through the system? And even knowing far more and further reaching ones today, most cars driving round are still burning oil based fuels.

Tesla have evidently kept their eye on the ball and adusted settings & monitoring to respond to evidence from the fleet about battery performance. That is still pretty cool. Imagine how it would be if they had gone the route of trusting they would get everything perfect from day one and just left everything on our cars the way it was when they rolled out the factory? Then there would be grounds for concern about the safety of your car imo. They could have gone the route of only charging to 4.1V or lower and limited acceleration / regen to shopping car levels from new, but then we wouldn't have the cars we own today.

There are a lot of questions this raises - right down to the accepted model of car ownership and liability - which doesn't just stop at the battery technology. FSD is pushing just as hard in its arena. It does make for interesting contemplation.

One thing that's clear to me is that with such a high value component as the battery, its warranty cover and terms must be way clearer than at present. I mean, just look at these posts, all trying to understand how the battery works and why some are drgrading and if Tesla have an obligation to fix something.... or even if anything is broken! Or just 80% broken.....

The earlier warranty needs urgent challenge and clarification. The fact that it doesn't specify any sensible measure for what might constitute a failure requiring remedy, or even what a suitable remedy might look like - is ridiculous.

Even with the 2020 terms, how can a typical owner be expected to know how to determine if their battery is 'bad' and even 'bad enough' to justify replacement. Then there is the whole issue of if the battery has been abused by the present or previous owners. Tesla have so much wiggle-room that the only time they would HAVE to do anything would be in the most extreme cases.

Every car should be legally obliged to have a battery condition readout that provides mandated and verifiable evidence to harmonised standards.

My Renault EV cut down equivalent to SMT does something along those lines.

 

[Battpower]

This is a screen from CanZE for my 2017 Zoe 40kwh.

My dad has similar model, 2015, 22kwh and his health shows 72%.

Seems like this kind of info should be standard on every EV and accessible without buying 3rd party tools.

 

[Florian500]

Lots of cell voltage plots here: Teslalogger mit Raspberry Pi & MySql & Grafana & OSM

 

[Battpower]

Lots of cell voltage plots here: Teslalogger mit Raspberry Pi & MySql & Grafana & OSM

Awesome! (Genuinely).

What, now I have to learn German too?!

 

[Florian500]

Awesome! (Genuinely).

What, now I have to learn German too?!

DeepL is your friend.
Or just ask me in case DeepL translates to gibberish

 

[Battpower]

Lots of cell voltage plots here: Teslalogger mit Raspberry Pi & MySql & Grafana & OSM

If the German thread is half a compromised as this one, between what's lost in translation and original misinformation and ambiguity, we could soon sink to new depths.


Any way, @Florian500, are you prepared to mediate between threads?