Battery charge level drops significantly after every firmware update

[T3SLA90D]

New Tesla S 90D delivered December 2016. After every firmware update, the battery charge level drops significantly about 20-30% compared to before the update while the car is parked and plugged in. An hour or 2 later, the charge level gradually goes back up to baseline. If the car is plugged in, it overrides the charge limit and end up at 100% even if it was set to max out at 80 or 90%. If the car is not plugged in, charge level eventually goes back to what it was before the update.

Service center does not know what's going on. They called Tesla engineers and states that this is a normal behavior with the car trying the re-read battery capacity.

Has anyone experienced a similar issue?

Thank.

 

[smartypnz]

You might want to research how that range is calculated.

 

[Nietschy]

You might want to research how that range is calculated.

why should he?
I never heard about this before. My charge level is exact the same as before the updates (I had about 6 already).
If most/all responses are that way. He should should empower tesla to fix his car because it is not right. period

 

[trayloader]

Yes. Same here.
7.1 was @ 384KM
then
379KM
now @ 370KM
within a mere 7000km of driving.

...

 

[T3SLA90D]

why should he?
I never heard about this before. My charge level is exact the same as before the updates (I had about 6 already).
If most/all responses are that way. He should should empower tesla to fix his car because it is not right. period

Thank you Nietschy

Yes. Same here.
7.1 was @ 384KM
then
379KM
now @ 370KM
within a mere 7000km of driving.

...

Trayloader I think you may have misunderstood the issue. I mean that, for example, before the software update the battery level is at 260 miles. As soon as the update is installed the battery level shows 160 miles and then an hour later it goes to 200 and then another hour later it's back at 260 miles without having driven the car.

 

[trayloader]

Sorry if I was not clear.
384 was at 7.1
379 was right after 8.0 first update
370 after the latest.

I should've stayed at 7.1

 

[ucmndd]

Same here. I dropped 10% after installing the latest update yesterday.

I don't think there's any cause for concern.

 

[Btrflyl8e]

. As soon as the update is installed the battery level shows 160 miles and then an hour later it goes to 200 and then another hour later it's back at 260 miles without having driven the car.

Simple solution, don't check until at least 2 hours after an update!

I don't know that I've ever paid attention to the rated miles showing immediately after a software update...and I've had dozens. Seems odd, for sure

 

[T3SLA90D]

Update on the fix for the record in case others experience a similar issue.

SC fixed the problem with the following description
"Corrections: HV Battery Assembly General Diagnosis

Issue caused b fault in battery management system circuit board. Corrections: Battery Management System (BMS) (HV Battery 2.0)

Replaced BMS circuit board."

I since insltalled a firmware update with no issues

 

[Shock-On-T]

Glad it's resolved.
Just FYI, the miles range is not really battery capacity. You need to change units to "energy" to show percentage. I'd be curious to know if they actual percentage was dropping that much.

 

[TexasEV]

Glad it's resolved.
Just FYI, the miles range is not really battery capacity. You need to change units to "energy" to show percentage. I'd be curious to know if they actual percentage was dropping that much.

No, both the range and the percentage displayed are estimates, as battery capacity can not be measured directly. Toggling the display will just change units from rated miles to percent. One is not more accurate than the other. For example, if a battery with 200 miles rated range shows 100 miles range on the display and you toggle to percentage, the car will display 50%. Of course that doesn't mean you will be able to drive 100 miles, just that the battery is estimated to be 50% full either way.

 

[Shock-On-T]

No, both the range and the percentage displayed are estimates, as battery capacity can not be measured directly. Toggling the display will just change units from rated miles to percent. One is not more accurate than the other. For example, if a battery with 200 miles rated range shows 100 miles range on the display and you toggle to percentage, the car will display 50%. Of course that doesn't mean you will be able to drive 100 miles, just that the battery is estimated to be 50% full either way.

I thought rated miles was one step further extrapolated?
ie. percent is an estimate of coulombs of charge remaining, and rated miles is an estimate of range using percent and driving history.
At least that's how I thought it worked...

Eg, if the battery is 50%, it might show a rated range of 100 miles for someone who drives like a maniac, but the car next door might show 120 miles even though it was also at 50% (with same pack size).

 

[xyeahtony]

The number you see on the dash is the computer estimating how much range is left in the battery. Its a "guess." When you reset a computer (i.e. during a firmware update which installs brand new software) then its perfectly normal to expect these systems that calculate things to start from scratch.

 

[Shock-On-T]

The number you see on the dash is the computer estimating how much range is left in the battery. Its a "guess." When you reset a computer (i.e. during a firmware update which installs brand new software) then its perfectly normal to expect these systems that calculate things to start from scratch.

That's why I was curious to know percentage rather than rated miles.
I can understand rated miles changing after a reset, but I would have thought that percentage would be calculated from some physical (non-resettable) attribute of the battery, such as pack voltage or internal resistance.

 

[Mediocrates]

No, both the range and the percentage displayed are estimates, as battery capacity can not be measured directly. Toggling the display will just change units from rated miles to percent. One is not more accurate than the other. For example, if a battery with 200 miles rated range shows 100 miles range on the display and you toggle to percentage, the car will display 50%. Of course that doesn't mean you will be able to drive 100 miles, just that the battery is estimated to be 50% full either way.

Do you have a source for this hypothesis that explains why a Tesla battery/subsystem would be different in this regard versus every other battery-powered piece of technology that has been in the hands of consumers for decades?

 

[stopcrazypp]

That's why I was curious to know percentage rather than rated miles.
I can understand rated miles changing after a reset, but I would have thought that percentage would be calculated from some physical (non-resettable) attribute of the battery, such as pack voltage or internal resistance.

Instantaneous measurements are not a reliable way of getting battery SOC, since this varies with age and conditions (a lookup table that factors those things in helps somewhat, but can't be completely relied on). Some form of coulomb counting and battery learning/history (to calibrate to prevent drift) is required to get a decently accurate estimation. These factors can be reset (and thus require re-calibration). See links below for details on how its done.

Do you have a source for this hypothesis that explains why a Tesla battery/subsystem would be different in this regard versus every other battery-powered piece of technology that has been in the hands of consumers for decades?

All battery percentage calculations are only an estimate (Tesla's included). The only true way to get accurate capacity is a full charge/discharge cycle. This is not practical in daily use in an EV for obvious reasons. The multi-cell pack (with cells in series) makes things even more complicated.

Ever have a device report 100% SOC for seemingly forever, and then suddenly drop? Ever have a device say it has some battery left and then sudden die? That's because the SOC estimation is not correct or is crude (for example basically just mapping voltage to SOC, sometimes linearly).

Here's some basic primer on battery SOC estimation:
Li-Ion BMS - White Paper - Estimating the State Of Charge of Li-Ion batteries
Measuring State-of-charge - Battery University

 

[Mediocrates]

All battery percentage calculations are only an estimate (Tesla's included). The only true way to get accurate capacity is a full charge/discharge cycle. This is not practical in daily use in an EV for obvious reasons. The multi-cell pack (with cells in series) makes things even more complicated.

I understand that the battery charge state is based on electrical characteristic measurements and an attempt to correlate that to the available capacity. That's the same as it has been done for decades, depending on the battery chemistry and application. We agree on that (although I've never experienced the aberrant behavior you described in terms of wild charge state fluctuations).

I contest the hypothesis that the "rated miles" is as (in)accurate as battery charge state, given they are in different temporal contexts and the former is an inferred prediction based on the latter as input to some function.

 

[stopcrazypp]

I understand that the battery charge state is based on electrical characteristic measurements and an attempt to correlate that to the available capacity. That's the same as it has been done for decades, depending on the battery chemistry and application. We agree on that (although I've never experienced the aberrant behavior you described in terms of wild charge state fluctuations).

I contest the hypothesis that the "rated miles" is as (in)accurate as battery charge state, given they are in different temporal contexts and the former is an inferred prediction based on the latter as input to some function.

The rated miles was always just a direct wh/mi conversion from capacity to miles. It does not vary based on driving conditions. The only thing that changes the mapping (very slightly) is toggling to range mode. This has been a Tesla thing for a while (Tesla uses rated range as a substitute for a kWh number).

@wk057 goes into detail here:
Calculate usable battery capacity based on rated miles values

 

[whttiger25]

I thought rated miles was one step further extrapolated?
ie. percent is an estimate of coulombs of charge remaining, and rated miles is an estimate of range using percent and driving history.
At least that's how I thought it worked...

Eg, if the battery is 50%, it might show a rated range of 100 miles for someone who drives like a maniac, but the car next door might show 120 miles even though it was also at 50% (with same pack size).

Driving history has nothing to do with rated range. Rated range is just a current estimate of kWh remaining in the battery multiplied by a constant which is unique to each battery and drivetrain combo.