Wiki Sudden Loss Of Range With 2019.16.x Software

[AustinP]

I wrote to my SeC about the slower SuC speed and the wrong estimates.
I mentioned this article for reference (
Tesla rolls out improved Model S/X battery thermal management software amid HK fire investigation).
I mentioned that I have the slower charge and am indeed not happy about the situation.
I questioned wether the situation would revert and if not, if a navigation update for charge time estimate was in the pipe.
The answer was that they heard rumors of maybe restoring speed for some VIN, depending on their battery state. But no real information on the topic.
As for the navigation, they say not coming.

I know folks at the SeC do their best to be helpful, and are not always informed properly by their headquarters.

So take it as just another feedback.

 

[Chaserr]

My predicted supercharging times are still accurate

It would be interesting to charge side by side with you at the same SOC.

When the recall notices go out you might not get one right away outside if the US, but when you are eventually forced to update as part of the imminent safety disclosures you should do it at a supercharger... Drive from 100% to 5%, supercharge back to 100%, update, then drive back to 5% and supercharge back to 100% same day. Directly compare pre-gated behavior to post, same day, and making sure battery temps are about equal (ludicrous heating pre charge in both cases?). Also get external audio of parked car noises, as a loud demonstration of pre and post draingate behaviour, and get charging / stationary unplugged voltages at 100% to compare how much is removed by the update after sitting for a specific amount of time.

Treat the update like a science experiment. It should produce interesting results. You could also throw more experimental data in there like 0-60, 0-100, 0-max speed, to both deplete charge faster and to see if performance is downgraded by the update. They obviously fear too much input, but it's not known if they remove how much the battery can output.

I'm also curious if the changes are instant or if they still try to hide it over days/weeks of charging like they did last year.

 

[BigNick]

That reminds me. Recently ((Saturday 10-24) while Supercharging, there was no jet-loud fan action. Late October with ambient temps in the low 50s (F).

 

[Florian500]

According to reports in German TFF Forum, the displayed range rose to the old values, slowly over the last months, a few km every week.

But the reported usable capacity in kWh did not move a single Wh up, according to ScanMyTesla.
Final cell voltage is still at 4,1V, but range shows the value before 2019.16. (minus a few % for degradation which is totally acceptable after 1.5 years).

So it looks like Tesla changed the calculation for available range w/o really increasing range or giving back usable battery capacity.

Akku-Kapazität (70/85-er) gesunken nach Update 2019.16.2

/Edit: there are also reports that usable capacity did rise hand in hand with displayed range to old kWh values gradually over the last few months.

 

[raphy3]

That reminds me. Recently ((Saturday 10-24) while Supercharging, there was no jet-loud fan action. Late October with ambient temps in the low 50s (F).

what firmware are you on? I get pumps going crazy sounding like trying to move the whole car by themselves just with charging on 110V! lol
2018.xx definitely had no such pump craziness. In fact, I heard the pumps so little that the first time I charged my car with 2020.36 above 80%, the car was literally making a new noise for me that I'd never heard of before.

 

[BigNick]

what firmware are you on? I get pumps going crazy sounding like trying to move the whole car by themselves just with charging on 110V! lol
2018.xx definitely had no such pump craziness. In fact, I heard the pumps so little that the first time I charged my car with 2020.36 above 80%, the car was literally making a new noise for me that I'd never heard of before.

I am still on 2020.36.11 despite about half a GB of download about six days ago. Might have been map-related data. Can’t see map version in the app.

 

[Droschke]

I am still on 2020.36.11 despite about half a GB of download about six days ago. Might have been map-related data. Can’t see map version in the app.

I also show huge downloads, but no new software update:

Last day: 154 MB
Last 7 days: 970 MB
Last 30 days: 3.7 GB

 

[Droschke]

According to reports in German TFF Forum, the displayed range rose to the old values, slowly over the last months, a few km every week.

But the reported usable capacity in kWh did not move a single Wh up, according to ScanMyTesla.
Final cell voltage is still at 4,1V, but range shows the value before 2019.16. (minus a few % for degradation which is totally acceptable after 1.5 years).

So it looks like Tesla changed the calculation for available range w/o really increasing range or giving back usable battery capacity.

Akku-Kapazität (70/85-er) gesunken nach Update 2019.16.2

/Edit: there are also reports that usable capacity did rise hand in hand with displayed range to old kWh values gradually over the last few months.

Looks like inconsistent reporting by owners. My numbers have not changed at all.

 

[wk057]

The firmware updates are a max of 1GB on MCU1... but usually they don't push them raw and push a binary difference that's only 50-200MB... then the car uses its CPU horsepower to calculate the updated image in the offline partition using the binary patch data plus the current version.

So anything over 1GB is likely maps. (Edit: I did see a car recently that got stuck in a loop downloading a patch over and over due to eMMC failure. The checksum would keep failing, it would re-download, fail, re-download. But that particular failure mode seems pretty uncommon and I've only seen it happen to one vehicle. eMMC issues are way more common in general, but usually fail in ways that make the MCU unusuable.)

 

[raphy3]

soon after 2020.40 may finally kill off a large mass of MCU1's. The encryption will destroy the EMMC for those chips that are borderline right now. Not to mention, the thing is a dog as is, I can't imagine having an encrypted FS on this.

 

[Droschke]

soon after 2020.40 may finally kill off a large mass of MCU1's. The encryption will destroy the EMMC for those chips that are borderline right now. Not to mention, the thing is a dog as is, I can't imagine having an encrypted FS on this.

I don't have 2020.40, in fact no new software update installed in the last 30 days. Nevertheless, the download size for the last 30 days for my car, which is mostly parked and has been driven less than 50 miles at best, seems excessive. I wonder if this shows a failing new software download that just keeps retrying or just map downloads.

 

[raphy3]

Whoa that's huge. I think maps can be that big from time to time, I've heard.

 

[JustADude]

soon after 2020.40 may finally kill off a large mass of MCU1's. The encryption will destroy the EMMC for those chips that are borderline right now. Not to mention, the thing is a dog as is, I can't imagine having an encrypted FS on this.

The storage controller and soc don't have hardware aes encryption, and the cpu doesn't have the resources to do it in software.

MCU 1 won't be getting encryption

 

[Droschke]

MCU 1 won't be getting encryption

I hope you are right.

 

[wk057]

Maps are ~4-5GB I believe.

As for encryption... have heard conflicting rumors on this one.

Realistically, the dev team would have to be absolutely insane to enable encryption on MCU1 at this point... unless they start offering free replacements for the life of the vehicle or similar.... which seems unlikely to happen. If they do end up enabling encryption on MCU1, within no time at all a ton of MCUs would die... so, this would be pretty bad press.

 

[Sup3rt3sla]

No rasmussen-v-tesla class action updates yet? There should have been an update on October 30, but at least on Stipulation and Proposed Order – #32 in Rasmussen v. Tesla, Inc. (N.D. Cal., 5:19-cv-04596) – CourtListener.com there's nothing.

 

[MP3Mike]

No rasmussen-v-tesla class action updates yet? There should have been an update on October 30, but at least on Stipulation and Proposed Order – #32 in Rasmussen v. Tesla, Inc. (N.D. Cal., 5:19-cv-04596) – CourtListener.com there's nothing.

They have almost always been late on filing the required updates.

 

[MP3Mike]

No rasmussen-v-tesla class action updates yet? There should have been an update on October 30, but at least on Stipulation and Proposed Order – #32 in Rasmussen v. Tesla, Inc. (N.D. Cal., 5:19-cv-04596) – CourtListener.com there's nothing.

"Update" on the class action: They want to continue discussions and are asking the court to stay the case until after their next scheduled update on December 15th.

 

[sorka]

According to reports in German TFF Forum, the displayed range rose to the old values, slowly over the last months, a few km every week.

But the reported usable capacity in kWh did not move a single Wh up, according to ScanMyTesla.
Final cell voltage is still at 4,1V, but range shows the value before 2019.16. (minus a few % for degradation which is totally acceptable after 1.5 years).

So it looks like Tesla changed the calculation for available range w/o really increasing range or giving back usable battery capacity.

Akku-Kapazität (70/85-er) gesunken nach Update 2019.16.2

/Edit: there are also reports that usable capacity did rise hand in hand with displayed range to old kWh values gradually over the last few months.

Any chance they could have dipped a couple of kWh into the 4 kwh buffer?

 

[Florian500]

Any chance they could have dipped a couple of kWh into the 4 kwh buffer?

4.00kWh

3a20839e4c8990178d8ee6337ab26e6dcbbca453.png