Wiki Sudden Loss Of Range With 2019.16.x Software

[sorka]

49 miles, 102kW
75 miles, 87 kW
119 miles, 65 kW
159 miles, 48 kW
177 miles, 43 kW
193 miles, 39 kW

Amazingly well written! We should all join (DAVID RASMUSSEN, an individual, on behalf of himself and all others similarly situated)

This is from Teslafi. You pay $50 per year for them to automatically collect this data (and more).

So glad I signed up with them years ago.

 

[sorka]

There are only two plausible explanations for the observed drop. Either the battery packs are faulty, leading to fires at normal states of charge, or the battery management software was faulty, overcharging cells that had already experienced some degradation.

Except that 4.2 volts is not overcharged. It's 100%.

 

[Chaserr]

You realize this thread was linked in the lawsuit, right? And many TMC members were quoted. Anyone who's interested in Tesla doing well is going to visit this thread. I just bought my 3 in March and I really don't want the people who delight in harming Tesla to use this as another piece of ammunition against them.

And forgive me if I'm misunderstanding, but how is this not about battery degradation? The plaintiff is suing for the value of their lost range. All Tesla needs to prove is that the reported loss in range is within normal parameters. It doesn't matter if it was sudden or not, the software changed how the hardware behaved and that led to the sudden drop. 4.2v could have been a dangerous level due to normal degradation, and Tesla has now adjusted the firmware to bring it back in line.

Degradation doesn't limit charge voltage, artificially lowered capacity using a software limit does. Every impacted battery is voltage limited, not degraded.

 

[DJRas]

I am just learning more about this issue today.

Someone measured their voltage at 100% charge before 2019.16 as 4.2v, applied the update, and then measured it again at 4.07v? Or is 4.2v the measurement from brand new?

There are only two plausible explanations for the observed drop. Either the battery packs are faulty, leading to fires at normal states of charge, or the battery management software was faulty, overcharging cells that had already experienced some degradation.

If the cells themselves were faulty, would an artificially imposed software lock really fix them? Or is it more likely that the battery management software was underestimating the degradation of the batteries and charging them beyond a safe top buffer?

4.2 volts is what ALL Li batteries charge to at 100%.EVEN degraded ones.
4.07 was measured after the update on affected cars at 100% charge. Thus, software limited to 85% of full charge.

So, does that make a faulty battery safe?
GOOD QUESTION!

Only very few cars were affected by the update... so, it certainly points to faulty batteries were detected and they HOPE limiting the charge will protect the cars.

The BMS doesn't estimate degradation. It measures how much power going in is required to get to 4.2 volts. That is relative to how much power can then go out.

 

[Blu Zap]

In case you missed the earlier link to the Reuters article.Tesla owner lawsuit claims software update fraudulently cut battery capacity - Reuters
Thank you David Rasmussen! We are with you all the way. Nice to see our little forum mentioned as the focal point to discuss this issue.

I have found focusing on kWh as the best metric. So in my case I had 73kWh capacity. Overnight with the "upgrade" to 2019.16.1 the battery capacity was at 62kWh the following morning.

 

[DJRas]

Interestingly, Tesla replied to Fred from Electrek.

They told him that new releases will fix this range issue. That sure seems like Tesla still believes the batteries are fine. It was a bug. Not Lithium dendrites or lithium plating.

I have an update from last night. Let's see how much capacity (range) comes back over the next 2 weeks.

Now on 2019.18.2

 

[Chaserr]

Their statement to Electrek directly contradicts what WK057 has said and I trust him more than Tesla, especially since a lot of what he said is from his Tesla contacts. I hope he dumps his info if they try to blow smoke at us again.

@DJRas 2019.18.2? Do you mean 2019.28? There is no .18 on TeslaFi.com Firmware Tracker but I wouldn't put it past Tesla to give you a custom firmware.

 

[MP3Mike]

Their statement to Electrek directly contradicts what WK057 has said and I trust him more than Tesla, especially since a lot of what he said is from his Tesla contacts. I hope he dumps his info if they try to blow smoke at us again.

Does it really? He said they were looking for X and found Z that they didn't previously know about. They didn't want to just reverse the changes. Maybe now that more time has passed and they have finished properly investigating Z they now aren't worried about it and are in the process of reversing the changes for cars with Z. (Which is what it sounds like Tesla told Electrek.)

 

[Chaserr]

He said that both X and Z would need to be physically fixed eventually. He even said it was important enough that people that haven't updated should despite the negative side effects.

 

[MP3Mike]

He said that both X and Z would need to be physically fixed eventually.

Nope, he never said that, at least as far as I can see.

He did say:

Basically they went looking for X and found Z instead. X is pretty bad, but doesn't seem to have happened anywhere. Detecting X is definitely a good thing. Z is not good, but not as bad as X. The process of looking for X's ended up finding a bunch of Z's as well. Z was not being looked for and wasn't known. Detecting Z is still a good thing. The people with a rapid range loss have condition Z.

The mitigation in place meant to combat X does also effectively mitigate Z. However, Z should get a more refined mitigation than is required for X.

I personally wouldn't consider X or Z to be "normal degradation." Tesla seems to recognizes that X would not be normal. Jury is still out on Z at the moment, on Tesla's side, with a proper detection correction. A proper mitigation of Z shouldn't be nearly as noticeable, in any case, so may be a non-issue... not really sure how that'll eventually pan out.

He even goes as far as to say that he thinks the mitigation for Z may be a non-issue.

 

[Blu Zap]

Hi everyone. Link to Electrek's article responding to this issue and the class action suit.
https://electrek.co/2019/08/08/tesla-owner-range-slashed-software-update-class-action-lawsuit/

As you recall, they had an early article on this topic.Tesla owners see battery range drop after software update to 'improve longevity' - Electrek

 

[Matias]

From electrek

"
Update: Tesla sent us the following response following the report:

“Delivering the best possible customer experience with the highest regard for safety has always been our priority, and we do not disregard either of these things as this complaint suggests. A very small percentage of owners of older Model S and Model X vehicles may have noticed a small reduction in range when charging to a maximum state of charge following a software update designed to improve battery longevity. As previously noted, we have been working to mitigate the impact on range for these owners and have been rolling out over-the-air updates to address this issue since last week.”

AFAIK no one has seen this update??

 

[DaveBC]

The way I understood wk057’s explanation is that the software detection of X also detected Z and applied the urgent fix for X to those vehicles.

Presumably if a pack has condition X it is in danger of catching fire. They don’t want to roll back their code intended to detect X (that is also flagging Z condition packs) and have a fire that could have been prevented.

They need to refine their detection of X and Z and apply the appropriate action to each, which hopefully would be to make safe by software and schedule a pack refurbishment or replacement for those packs that actually have X.

(I am not a battery chemistry or BMS expert, this is just what I concluded after spending weeks of my spare time reading this thread)

 

[Guillaume]

Beware: The update 2019.28.2, which is rolling out atm, further reduces the charging speed.

Left curve is 2019.28.2, right curve is 2019.24.x

Source: German Tesla Forum: 85er Akku : Reduzierung der Ladeleistung seit 2019.20.1 • TFF Forum - Tesla Fahrer & Freunde

 

[Blu Zap]

Beware: The update 2019.28.2, which is rolling out atm, further reduces the charging speed.

Yikes! Thanks for the heads up!

 

[sorka]

I am sure Tesla will say it only affects 10 cars, all on the forum

As funny as that is, a certified class will result in Tesla having to turn over all customer info for every

Nothing about how a ranges read out is some inherent measure of what a battery holds. Battery charges are measured via coulomb counting - e.g. tracking electrons in vs. electrons out. This drifts over time and has to be recentered, by a variety of means. The driver in theory never charges / discharges it across its entire limits (because parts of the SoC range are locked off), and in practice, rarely if ever charges / discharges it across the portion of the cell that they have access to. So there's a lot of guesswork.

It's Tesla's BMS software that determines what to tell you your range is. If they change their software, it can readily tell you a different figure. And have various new driving constraints based on that figure. Even though nothing physically changed with the cells. That doesn't mean that the changes are wrong. They can be for accuracy. They can be to make sure that you don't run out of charge above 0%. They can be to make sure that you're keeping enough of the upper end of the charge range locked off to minimize degradation. There's all sorts of reasons.

Clearly someone who hasn't read the thread and isn't aware of the facts.

 

[DaveBC]

Beware: The update 2019.28.2, which is rolling out atm, further reduces the charging speed.

Left curve is 2019.28.2, right curve is 2019.24.x

I don’t like to repeat something without having the source at hand but I will anyway. I heard somewhere that the SC power limits are tied to new thermal parameters.

I’d want to know the ambient temperature of each Supercharger session to be compared.

 

[xveys]

Beware: The update 2019.28.2, which is rolling out atm, further reduces the charging speed.

Left curve is 2019.28.2, right curve is 2019.24.x

Source: German Tesla Forum: 85er Akku : Reduzierung der Ladeleistung seit 2019.20.1 • TFF Forum - Tesla Fahrer & Freunde

That's very bad... The first decrease was already very noticeable with reductions up to 30% in speed for me, now it will be even lower?
I really can't believe they just do this on a whim without communication or explanation. But hey, at least our cars got 'better' with a Chess app...

 

[RobPet]

From electrek

ave been rolling out over-the-air updates to address this issue since last week.”

AFAIK no one has seen this update??

So, 21 cars so far with the 2019.28.2.5 revision:

All 85s.

However, even though 28.2 is installed on 85s, all of the .5 were updated from .24.
So this could also be nothing, but interesting nonetheless. Hopefully, someone reading this thread will be/has been updated and could tell us if it changes anything.

 

[KarenRei]

Beware: The update 2019.28.2, which is rolling out atm, further reduces the charging speed.

Left curve is 2019.28.2, right curve is 2019.24.x

Source: German Tesla Forum: 85er Akku : Reduzierung der Ladeleistung seit 2019.20.1 • TFF Forum - Tesla Fahrer & Freunde

That could be true. It could also just be a random charging session difference. Have multiple people confirmed this without any denials of it?

Just from (Google) translating that thread, I see the comment in which that graph was posted reads "OK last week, the battery was warm but the curve before the update was slightly better than now with the 28er straight to the end."