42.3 Bug Report: Charging stops 1% short of target

[remlemasi]

Here's a screenshot of my last 9 charges:

The last five sessions that were allowed to go to completion, which happen coincide with the 42.3 update, all stop 1% short of the target.

I know there are some precision issues with setting the limit via the screen or the app, but the targets were all set and/or confirmed to be multiples of 5 via TeslaFi.

Not a big deal and doesn't practically impact range or usability, but it's a bug nonetheless (likely with 42.3 and not TeslaFi) and is kind of annoying for those who like to see nice, round numbers.

My complete charge count on TeslaFi made much more convoluted and difficult to read, although I admit a lot of the prior mess was my own fault. Well, only kind of, since the precision/alignment of limit slider is a UI issue.

 

[BizJet]

I have experienced this, too, on 2018.42.3.

Last several charges:

 

[Sophias_dad]

Me too... sounds like a pattern.

 

[ibdb]

I was directed here thanks to TeslaFi. 13 of 19 charges since 2018.42.2 19e7e44 installed on 10/27 have stopped at 89%, despite being set to 90%.

No similar issue with my S.

 

[GregRF]

Same on 2018.42.2.1

 

[ibdb]

Should have been more clear in my earlier post. I first noticed it in the release listed above, and it's persisted in 2018.42.4 ccb9715 and 2018.42.8 d1a49f2.

None of the 55 charges before that point stopped short of the set value.

 

[Trevor B]

All of !% short!!!

OMG ... what are we going to do? Doomsday is fast upon us

 

[ucmndd]

FWIW, I've noticed similar behavior, but I think it’s an off-by-one error in the API. For example, the mobile app says the car stops charging at 89%, but the display in the car itself says 90%. Charging amps in the app says 31 amps, car says 32 amps.

 

[ibdb]

All of !% short!!!

OMG ... what are we going to do? Doomsday is fast upon us

I'm not concerned about the 1%. I'm concerned about setting a target charge state and the car not reaching that target. If there's something else behind the bug that could affect whether a car reaches a set state of charge, I'd rather it be noticed and fixed while it's just 1% than if the faulty code continues to be used and it turns into something else.

I used to work with developers all the time, troubleshooting problems with their code and testing for errors. Often I'd find very minor things, but things that could have turned into bigger problems in the right (wrong) situation. In one case, an error reading a particular date on a particular type of credit card caused all the cash registers in a store to freeze and crash. 99.9%+ of the time, things were fine. When they weren't, they REALLY weren't.

 

[ibdb]

FWIW, I've noticed similar behavior, but I think it’s an off-by-one error in the API. For example, the mobile app says the car stops charging at 89%, but the display in the car itself says 90%. Charging amps in the app says 31 amps, car says 32 amps.

I'm seeing the car still say 89% and the Tesla app still say 89%.

 

[ibdb]

Updated to 2018.44.2 3b2a5c3 last night, and my first charge this morning stopped at the correct %.

 

[remlemasi]

Here's a screenshot of my last 9 charges:

View attachment 350388


The last five sessions that were allowed to go to completion, which happen coincide with the 42.3 update, all stop 1% short of the target.

I know there are some precision issues with setting the limit via the screen or the app, but the targets were all set and/or confirmed to be multiples of 5 via TeslaFi.

Not a big deal and doesn't practically impact range or usability, but it's a bug nonetheless (likely with 42.3 and not TeslaFi) and is kind of annoying for those who like to see nice, round numbers.

My complete charge count on TeslaFi made much more convoluted and difficult to read, although I admit a lot of the prior mess was my own fault. Well, only kind of, since the precision/alignment of limit slider is a UI issue.

View attachment 350390

Nearly all of my "1% short" charge sessions were corrected by TeslaFi. See below and compare to my original post:

All the "spikes" of extra range in the chart have also been adjusted to be relatively flat.

 

[ibdb]

From the TeslaFi Change log:

3 days ago

Improvements:

• Battery degradation report will now use the charge_limit_soc field instead of the chargePercent field if the charge percentage is 1% below the set charge limit when a charge completes. Additional information has temporarily been added to the 'Included Data Points' table.

Data Logger 11.30.18.1:

• Added better error handling for a new timeout occurring during schedules

James yesterday at 9:23 a.m.

Logger 12.3.18.1:

• Reminder schedules will only run once if the reminder is false.

Improvements:

• Integrated battery level fix from 11/30/18 into charge calculations directly to display throughout site. All charge records have been reset.

 

[remlemasi]

From the TeslaFi Change log:

3 days ago

Improvements:

• Battery degradation report will now use the charge_limit_soc field instead of the chargePercent field if the charge percentage is 1% below the set charge limit when a charge completes. Additional information has temporarily been added to the 'Included Data Points' table.

Data Logger 11.30.18.1:

• Added better error handling for a new timeout occurring during schedules

James yesterday at 9:23 a.m.

Logger 12.3.18.1:

• Reminder schedules will only run once if the reminder is false.

Improvements:

• Integrated battery level fix from 11/30/18 into charge calculations directly to display throughout site. All charge records have been reset.

Thanks for posting. I should've checked the change logs first!