Wiki Sudden Loss Of Range With 2019.16.x Software

[MP3Mike]

What you're saying is is that there was never an actual range reduction with the 2019.16.x firmware and that it was just a reporting issue. If the BMS hadn't been capping cell voltage to 4.1 after that update, that assertion might be true but capping to 4.1 volts means that after the update, there's already 10% that is being locked out on that top that would be available if the BMS was allowed to finish charging to 100%.

Assuming a pack suffers from Condition Z, and was reporting 260 miles of range when fully charged:

• Pre 2019.16.x it would shut down prematurely, at ~30 miles reported remaining. So you only had 230 miles of capacity that could be used.
• Post 2019.16.x the car only charges to ~230 miles of reported range, all of which can be used.

So in both cases you had ~230 miles of energy capacity you could use. In the first case it was shutting down before it hit the bottom, in the second case the bottom capacity is unlocked and the top is locked. Why is that so hard for you to understand?

Using your percentage: Before the update the car lets you use 10-100%, after the update the car lets you use 0-90% until it finishes its analysis at which point you should regain access to most of the top 10% that was locked out.

Show me where Jason claims there wasn't an actual range reduction with the detection the mistaken detection of condition X when it's actually condition Z.

Explaining Changes post-firmware 2019.16 Regarding Range Loss | wk057's SkieNET (I can't quote his specific text because of his copyright, but it is in there. And you can see that he rated my post describing the situation with a like, so I assume that means I am describing his analysis correctly.)

 

[sorka]

Explaining Changes post-firmware 2019.16 Regarding Range Loss | wk057's SkieNET (I can't quote his specific text because of his copyright, but it is in there. And you can see that he rated my post describing the situation with a like, so I assume that means I am describing his analysis correctly.)

Can you just PM it to me?

 

[sorka]

Assuming a pack suffers from Condition Z, and was reporting 260 miles of range when fully charged:

• Pre 2019.16.x it would shut down prematurely, at ~30 miles reported remaining. So you only had 230 miles of capacity that could be used.
• Post 2019.16.x the car only charges to ~230 miles of reported range, all of which can be used.

So in both cases you had ~230 miles of energy capacity you could use. In the first case it was shutting down before it hit the bottom, in the second case the bottom capacity is unlocked and the top is locked. Why is that so hard for you to understand?

Using your percentage: Before the update the car lets you use 10-100%, after the update the car lets you use 0-90% until it finishes its analysis at which point you should regain access to most of the top 10% that was locked out.

Nobody who's been effected by this condition has provided any data that would show this assertion is true.
Regardless, locking out the top 10% of the battery by capping to 4.1 volts is still a range reduction on the top that exists but is being locked out.

 

[oaito]

...drive around in a supercharger parking lot to less than 3% SOC to find out if your pack has condition Z?

How is that related?

That is funny, because, yes, you (@dark cloud) are right! At least partly... Do not get me wrong, I am laughing because of the madness of this situation! AFAIK...


The Tesla-Mantra:
- Tesla stated through their minions ...errh meant to say SeC-employees on several occasions to numerous owners around the world: "We, the mighty SeC-Siblinghood are the advanced ones to be chosen to handle TOOLBOX2! Do not believe in FUD! Your battery is safe and all normal wear! - There is no thing as Zzz'ed!
- Tesla stated in all their release notes so far no indication that it is necessary for safety to upgrade. (*1)

The investigative fact-driven 3rd party:
All is save. (I actually believe him!) Even if your battery is Zzz'ed :
- you're 85 is running on a firmware older than 2019.16.12? Then you do not have the energy/range shown onboard. (*2)
- you're 85 is running on a firmware older than 2020.30.x? Then you do have the energy/range shown onboard and stays that way. (*3)
- you're 85 is running on a firmware newer than 2020.30.x or even better, at least 2020.36.x? Then you do have the energy/range shown onboard and throughout a couple of month and 1000s of miles you should gain some range. (*4)

The Owner has freedom of choice, until you need to fix your Tesla by a SeC:
- Downgrade? Impossible!
- You want to get CCS upgrade? Uh, need to upgrade before the appointment, otherwise will charge you extra. (EU luxury )
- You want to get your MCU/EMMC fixed? Uh, need to upgrade before the appointment, otherwise will charge you extra.
- You want to get your car diagnosed to fix XYZ? Uh, need to upgrade before the appointment, otherwise will charge you extra.
- ....

(*1) Okay, not in the release notes, they tried to push us to upgrade for remote/app functions and most recently for basic functionality when the MCU (emmc) fails. Yet, there is a handful of indomitable owners refusing to upgrade and enjoying there decision...

(*2) And this could be the case "to drive around in a SuC parking lot down to 3%, 2%, 1% and even 0% or 0mi and below, to check if you're Zzz'ed! Not necessarily at a SuC, you could choose any charger. Alternatively you buy a AC Emergency Supply You could check wk057's list of possibly affected battery part#, then again, according to your SeC of choice: All is fine!

(*3) Did you saw a drop in range when charging to a certain soc? Or do you have trouble driving the same route under similar circumstances? Then it could be you're Zzz'ed! Only way to mitigate the effects would be further upgrading.

(*4) Already all good again? See, Tesla SeC was telling you all along, with OTA everything is going to be better. You're our little beloved Guinea pig and we will not tell you that you (again) saved Tesla-as-a-company a couple BTC with your lifetime, but as we said: As a good customer you do not need to worry, simply believe in us keeping your best interest in mind... and saving Tesla lots of money is in your best interest, isn't it?

BR! Oaito.

PS: Sorry if you think too much irony authored the better part of my message, perhaps I'm utilizing irony as a coping mechanism in this bizarre play.

PS2: My S85 is affected. I was able to drive regularly down to 5%, 3%, 0% according to my logs before installing 2019.16.x Yet the nominal thus usable capacity dropped after installing 2019.16 from being displayed in diagnostic mode as 72,3kWh to 67,1kWh, the car does not charge beyond 4.1V Max Brick Voltage (as displayed in diagnostic mode) and to make an awful situation even worse charging stops at 96% soc regardless of AC or DC. To charge the same amount of energy/range I need to charge significantly more delta-%-soc while charging speeds are overall reduced. I installed 2020.36.11 at 11th November, got the CCS upgrade, no change in range or Max Brick Voltage. Had the chance to co-pilot in another 2015 S85 with similar limitations than mine for a long distance, roughly 800km in one trip. At our destination the emmc was changed and afterwards we drove back with 2019.12.1.1 - the trip back was under nearly identical ambient situation and we took the same roads, same max speed and same SuC, cross checked with ABRP, little to no difference in elevation: 70min less charging time spent! So as much as I appreciate all the explanations so far regarding batteries being Zzz'ed and Teslas ingenuity to counter this through BMS during 2000mi and cycles and duration - I bought this car for long distance driving, twice a week each time 420km, the next project I've accepted brought me onto the road twice a week 680km each. The last project was 980km away, so I planned to see my family only every other week... So the next 250000km will add 14days worth of lifetime spent additionally at SuCs. It took Tesla roughly two years for putting a mitigation of the range issue into place without even acknowledging there is an issue in the first place.

So anyone out there who holds back the explanation for Condition F(lash - the "fastest" three toad sloth) and possible mitigation - maybe it is similar easy, you just have to stop charging higher than 63% and not discharge lower than 41% for, say, 405,3 cycles and keep in mind not to charge faster than 2kW... damn, there it is again, the irony!

 

[Dave EV]

This part still has me a little confused. Why would it do this? Supposedly the issue is the BMS will erroneously read lower than it actually is, and cutoff because of that. The update in 2020.30 is supposed to fix the erroneous reading so that it doesn't cutoff.

I am making some assumptions here based on wk057's posted information on Condition Z:
I am assuming that the BMS is relying less on the direct module level voltage readings and more on indirect module level voltage readings obtained by reading voltages across all other modules simultaneously - perhaps completely relying on indirect voltage readings. The indirect readings have higher margins of error. So to be safe, and to make sure it isn't overcharging a module until it can trust those readings, it limits the peak voltage of the pack. At the low end, there's no real harm to be done if voltages end up being slightly lower than expected. Over time as the BMS gains trust in the indirect voltage readings, it raises the maximum voltage threshold and you get nearly all your range back.

A brand new car does not take 30 days and 30 cycles for the BMS to figure out its range.

That's because if you have a new battery, you know what condition it is in and there's no learning that has to be done.

Nobody who's been effected by this condition has provided any data that would show this assertion is true.

wk057 has seen numerous packs with this issue and has confirmed it himself.

 

[AmpedRealtor]

Yay for this issue being resolved.

Nay for how @wk057 has been treated by other members and the moderators. I have complained several times relating to disinformation and trolling by a handful of folks yet no action has been taken. It appears those who like to cause problems are protected while those who are fact seekers and problem solvers are marginalized.

What the hell, TMC?

 

[wk057]

*grumbles something about still being pulled back here*

If you are on pre-2019.16.x and you were able to drive down to 0% your pack does not currently suffer from Condition Z.

Correct, with the only clarification being that you can't be 100% certain on this by going to 0% SOC. You're rolling the dice at low SOCs. If you have Condition Z, and the safety system catches a Condition Z bad voltage reading, the car will shutdown. The chance increases the closer to 0 you go. Depending on how badly Condition Z is affecting readings, the chance can be higher or lower.

It's also possible (and probable) that you don't have Condition Z.

So, you might be able to drive at low SOCs with pre-2019.16 + Condition Z in some cases, but your odds are not great.

...

All correct.

 

[fbitz777]

Pre 2019.16.x update if your pack suffers from Condition Z it will shut down when you hit ~30 miles remaining. (i.e. you can't access all of the range that is reported as available.) Post 2019.16.x update it corrects the range reporting, while locking out the top, but you can then use down to ~0%.



If you are on pre-2019.16.x and you were able to drive down to 0% your pack does not currently suffer from Condition Z.

Nobody who's been effected by this condition has provided any data that would show this assertion is true.
Regardless, locking out the top 10% of the battery by capping to 4.1 volts is still a range reduction on the top that exists but is being locked out.

there is ONE guy who was affected 2 years ago...got his battery replaced however...not sure this was a condition Z but car would shut down with 16miles of range left.
Infinite Mile Battery Warranty [Now] Being Honored By Tesla [Issue Resolved]

 

[BPeter]

*grumbles something about still being pulled back here*



Correct, with the only clarification being that you can't be 100% certain on this by going to 0% SOC. You're rolling the dice at low SOCs. If you have Condition Z, and the safety system catches a Condition Z bad voltage reading, the car will shutdown. The chance increases the closer to 0 you go. Depending on how badly Condition Z is affecting readings, the chance can be higher or lower.

It's also possible (and probable) that you don't have Condition Z.

So, you might be able to drive at low SOCs with pre-2019.16 + Condition Z in some cases, but your odds are not great.

All correct.

Interesting. So they knew about condition Z in 2019.16 and immediately unlocked the bottom of the battery (ignoring erroneous low readings) but locked the top until they could figure out a way to reliably read the voltage?

Is it ok to do this because it's safe (as in no fires, not necessarily for battery health) to let the battery over-discharge (in the case of the low reading was not erroneous) but it's unsafe to over-charge in case the lower reading is the correct one?

 

[stopcrazypp]

A brand new car does not take 30 days and 30 cycles for the BMS to figure out its range.

A brand new car has mostly balanced cells from the factory, but unless I'm mistaken the people affected don't have brand new cars, so we aren't talking about brand new. There's a specific batch of BMSes from cars much older that are mainly affected by this. The capacity estimates drift with use and cycling.

Until I see any of the many people affected on this thread having recovered to 4.2V I remain skeptical. I certainly have more deep cycles since 2020.30 and I have not seen any recovery. (however my battery is not on the list but has all the symptoms)

Your issue may not necessarily be the same as those cars. But if you read the FAQ link, with the mitigations placed, affected cars may never hit 4.2V. There's going to be a bit lost as part of the mitigation (a few miles / a few mV). In order to respect the copyright wishes laid out by the author however, I don't dive deeper than that though. I encourage people affected to read the whole thing end to end.

 

[stopcrazypp]

Interesting. So they knew about condition Z in 2019.16 and immediately unlocked the bottom of the battery (ignoring erroneous low readings) but locked the top until they could figure out a way to reliably read the voltage?

Is it ok to do this because it's safe (as in no fires, not necessarily for battery health) to let the battery over-discharge (in the case of the low reading was not erroneous) but it's unsafe to over-charge in case the lower reading is the correct one?

That's not my take. They didn't unlock the bottom, just that after the update the range estimate will reflect that the bottom is locked out (just like in the back end estimate that user wasn't able to see). Prior to the update it was also locked out, but just that the range estimate didn't reflect it.

Also from the voltages posted for Vmin, Tesla even when discharged lower than expected is still far from battery damaging or even extra degradation as that is 2.5V. And even if you discharge below that, it's bad for the battery (especially leaving it there for a long time) but not inherently unsafe. However overcharging can lead to fire, so that is definitely something Tesla needs to make sure to avoid.

 

[Ostrichsak]

there is ONE guy who was affected 2 years ago...got his battery replaced however...not sure this was a condition Z but car would shut down with 16miles of range left.
Infinite Mile Battery Warranty [Now] Being Honored By Tesla [Issue Resolved]

Actually, this nearly exact thing happened to me (P85D that shut down w/16 miles of range left w/o any advance warning) about two years ago as well. I actually thought you were referring to me until I clicked the link and saw someone else with a similar story. I was told that the 12v battery was the problem (ROFL) so I let them replace it under warranty. I no longer own that car so I can't say what the current status is.

 

[MP3Mike]

That's not my take. They didn't unlock the bottom, just that after the update the range estimate will reflect that the bottom is locked out (just like in the back end estimate that user wasn't able to see). Prior to the update it was also locked out, but just that the range estimate didn't reflect it.

We aren't talking about the normal bottom anti-bricking buffer, is that what you think wk057 was describing?

We are talking about the last few percent that are normally usable. If your battery suffers from Condition Z it is likely that you won't be able to use the last few percent if you are on a pre 2019.16.x firmware. If you are lucky and don't have any sporadic false voltage readings you can use it, but if you get one of the false voltage readings at a low SoC it will shut down on you with no warning. Apparently 2019.16.x, and newer, fixed that issue.

 

[stopcrazypp]

We aren't talking about the normal bottom anti-bricking buffer, is that what you think wk057 was describing?

We are talking about the last few percent that are normally usable. If your battery suffers from Condition Z it is likely that you won't be able to use the last few percent if you are on a pre 2019.16.x firmware. If you are lucky and don't have any sporadic false voltage readings you can use it, but if you get one of the false voltage readings at a low SoC it will shut down on you with no warning. Apparently 2019.16.x, and newer, fixed that issue.

No I'm not talking about the anti-bricking buffer. To clarify, my read is 2019.16.x made the back end estimate of capacity come to the surface (reflecting how the last few percentage may not be usable due to Condition Z), but didn't have mitigations to fix the issue yet. In my comment above I was talking about 2019.16.x.

It was 2020.30 and afterwards that did have the mitigation. It's a bit confusing I guess to talk about pre/post update without specifying which version, so I guess I should have clarified.

 

[bhzmark]

Can you just PM it to me?

Will you update now?

 

[sorka]

Will you update now?

Do you mean firmware? Getting MCU2 on Tuesday.

 

[BPeter]

That's not my take. They didn't unlock the bottom, just that after the update the range estimate will reflect that the bottom is locked out (just like in the back end estimate that user wasn't able to see). Prior to the update it was also locked out, but just that the range estimate didn't reflect it.

Also from the voltages posted for Vmin, Tesla even when discharged lower than expected is still far from battery damaging or even extra degradation as that is 2.5V. And even if you discharge below that, it's bad for the battery (especially leaving it there for a long time) but not inherently unsafe. However overcharging can lead to fire, so that is definitely something Tesla needs to make sure to avoid.

If that was the case then what is the deal with people reporting being locked to 4.1V max charge voltage? Is the BMS now reporting the lower voltage over CAN instead of the correct voltage?

 

[stopcrazypp]

If that was the case then what is the deal with people reporting being locked to 4.1V max charge voltage? Is the BMS now reporting the lower voltage over CAN instead of the correct voltage?

From the article, there is no hard 4.1V voltage cap (was there evidence there was, I didn't follow this issue?). It may just look that way from the BMS thinking that some cells might be at higher voltage than being reported, which if condition Z was the cause, the updates after 2020.30 supposedly will address as the system gathers more data and it is able to figure out the true voltage of the affected cell group (and thus would allow higher peak voltages). See the last section in the article.

Keep in mind due to how BMS works (in terms of trying to keep cells balanced) it takes just one cell group (out of 96) being out of whack to drag the entire pack down. Don't assume just because it seems like it is limiting all groups that there is some sort of universal limit being applied.

 

[hpartsch]

Well, I still think the lawsuit has merit. You can't sell something to someone, take it away without disclosure, then fix it years later. They may have very well only fixed it due to all of the publicity... There was and still is damage to customers.

However, I'm skeptical it's all truly fixed. Supercharging speeds and time left is still way off.... This change/slowdown happened to lots of people. At least for myself -- I've updated to all the latest software versions and drove over 8k miles. I've not seen this supercharging issue/fake time left issue resolved.

 

[stopcrazypp]

Well, I still think the lawsuit has merit. You can't sell something to someone, take it away without disclosure, then fix it years later. They may have very well only fixed it due to all of the publicity... There was and still is damage to customers.

However, I'm skeptical it's all truly fixed. Supercharging speeds and time left is still way off.... This change/slowdown happened to lots of people. At least for myself -- I've updated to all the latest software versions and drove over 8k miles. I've not seen this supercharging issue/fake time left issue resolved.

Your issue seems to be supercharging speeds, which the lawsuit for this issue would not cover (it only covers range loss from the reports I read). Is there another lawsuit having to do with supercharging speeds?