2019 3D+ ambient temperature reading 102F while it's 30F outside

[ilyak]

Anyone observe odd ambient temperature readings on 2020.48.12.1?

Yesterday, 30F outside and cloudy, I noticed my 2019 Model 3D+ displaying an ambient temperature reading of 102F: imgur.com

After some time parked with climate running, it climbed to 103F, then went down to 90F. A reboot and power off did not change the reading, so I let the car sleep.

After sleeping for a few hours (still ca. 30F outside), the car showed a much more realistic reading of 42F. However, shortly after I got in, the temperature started climbing, a degree every couple of seconds. After 5 minutes, the car was showing an ambient temperature reading of 85F. I went for a drive at highway speed and it gradually dropped to 38F and then fluctuated between 38 and 40F for the next hour. Will take it for another drive today and see if it recurs.

I realize this is possibly a hardware problem and not necessarily related to the update. However, since my nearest SC is a couple hours away and mobile service doesn't appear as an option for any appointment category at the moment, I wanted to see if this may be affecting other cars. Cursory research pulled up some threads about a blank reading due to rodent damage to the front wiring harness, but nothing on severely inflated readings. The ambient temperature sensor attached to the active grille assembly looks clear of salt and debris, the front wiring harness is plugged in and doesn’t have any signs of damage. The only notable event yesterday was the update to .12.1.

Based on how fast the temperature readings climb when the car is first awoken and climate is enabled and then stabilize, it kinda feels like it’s displaying readings from a different temperature sensor that’s somewhere in the HVAC loop.

Hope this is not what Elon meant by the holiday update being :fire: :fire:

 

[Probllama]

This happened to me a few year back when I owned a BMW. It turned out that a rodent actually chewed some cables while the car was parked at the airport. So I'd say it is probably a hardware issue rather than a software glitch...

 

[ilyak]

This happened to me a few year back when I owned a BMW. It turned out that a rodent actually chewed some cables while the car was parked at the airport. So I'd say it is probably a hardware issue rather than a software glitch...

That was my first guess too, but there are a few arguments against it:
- There is an actual reading. In cases I've read about where there was damage to the front wiring harness, the reading was displayed as "---"
- As ambient temperature fluctuates, the reading also fluctuates. For example, it was 30F yesterday, and when I woke the car up, it showed 32F. It's 50F today, I just went out and woke the car up, and it is showing 55F. So the sensor is at least functioning, the temperature build up starts after a few minutes.
- No traces of damage on the wiring, no rodent feces on any visible surfaces, etc.

 

[ilyak]

Welp, looks like the wrong temperature reading is affecting cabin HVAC - my car is running A/C and blowing freezing air out of the vents while it is 30F outside. Based on the lack of reports on TMC or Reddit, I am guessing a hardware issue with the sensor or the wiring harness is at fault here. I went ahead and scheduled the earliest available appointment for January 19th.

 

[bbell]

Just a data point on this, I had a similar issue about a year ago. It read well above the real temperature, and it seemed to follow the changes in the actual outside temperature, and like you the climate control was also confused.

I was about to call for a mobile service visit, but before I got to doing so a software update came in, and after it was installed the problem went away.

Since I doubt it was a specific problem with the older software (or everyone would be affected) I figure it probably means that the MCU isn't the problem but one of the other modules around in the car. When installing the new software they are rebooted, unlike a steering wheel reboot of the MCU only. Perhaps someone on here knows which module is in charge of temperature readings.

On possibility is it could be resolved by a hard powerdown of the car (IE disconnecting the main battery and 12V battery), but I would only try that only if you are comfortable doing it.

Since you have an appointment set for the 19th, it is possible new software will come along first and so I would install that to see if that helps.

 

[ilyak]

Just a data point on this, I had a similar issue about a year ago. It read well above the real temperature, and it seemed to follow the changes in the actual outside temperature, and like you the climate control was also confused.

I was about to call for a mobile service visit, but before I got to doing so a software update came in, and after it was installed the problem went away.

Since I doubt it was a specific problem with the older software (or everyone would be affected) I figure it probably means that the MCU isn't the problem but one of the other modules around in the car. When installing the new software they are rebooted, unlike a steering wheel reboot of the MCU only. Perhaps someone on here knows which module is in charge of temperature readings.

On possibility is it could be resolved by a hard powerdown of the car (IE disconnecting the main battery and 12V battery), but I would only try that only if you are comfortable doing it.

Since you have an appointment set for the 19th, it is possible new software will come along first and so I would install that to see if that helps.

Thanks, this is very helpful - the description fits. This is why I did ask for a firmware redeploy in the appointment request. Hopefully someone reads it soon and pushes out the current firmware to me. I do some QA as a part of my job so I trust my hunch when it comes to thinking this is software-related. The sensor functioning is too normal to be a hardware failure.

Edit: no sooner than I posted this, I got an update notification. Reinstalling now, will see if this helps.

 

[bbell]

Edit: no sooner than I posted this, I got an update notification. Reinstalling now, will see if this helps.

Good luck!

 

[ilyak]

I *think* redeploying the firmware fixed this. I just went for a 10 minute drive and the ambient temp only climbed by 1 degree. I'll be able to do a better test later today, but - hopefully - this was it. Thanks again, @bbell - I owe you one.

 

[ilyak]

Unfortunately, the problem recurred last night. After the car fell asleep, I got in and watched the ambient temperature reading immediately start climbing again, all the way to a balmy 96 degrees after some 3 minutes. It is, of course, not anywhere near 96F in New HYampshire right now. The odd thing was that I then drove to a grocery store 5 minutes away, spent some 30 minutes there with the car parked and awake (sentry running), and when I got back in, the temp was reading 32F and did not climb for the rest of the drive, some 40 minutes.

So, my theory as of last night was that this is somehow related to the car sleeping and leaving the car awake and parked led to a calibration routine running. I say "was", because when I walked out to the car this morning and got it, the reading did not climb at all. So in addition to everything else, the issue is intermittent.

At this point, I am keeping my appointment to rule out an intermittent hardware fault, but I still feel like this is a software (or, possibly, software/hardware) issue caused by an interaction of several conditions that I have not yet narrowed down. While I wait, I will continue observation.

Oh, and here's the issue in action, albeit at the lower end of the temperature range: imgur.com
It was 26F outside when this was taken.

 

[ilyak]

Just an update that I’ve since had a few instances of blank readings, the issue comes back on and off. Mobile is coming at the end of the month, no estimate yet so don’t know what their plan is.

 

[ilyak]

Another update: mobile finally made it out and replaced the ambient temperature sensor. Now it's reading 30 degrees lower than the actual temperature. They asked I give it a few days and then request another appointment if it does not normalize.

 

[XLR82XS]

3D+ ?

 

[ilyak]

3D+ ?

One of the common ways to refer to a model 3 long range all wheel drive with an acceleration boost, no?

 

[ilyak]

Just another update for posterity.
1. I had the ambient temperature sensor replaced by mobile. It didn’t help, the ambient temp racing to high values returned shortly after the replacement. So it’s not the sensor itself.
2. I removed the number and cleaned the front wiring harness best I could. No effect, but I also confirmed that there’s no obvious damage to the harness itself (I initially suspected rodents since Tesla uses soy-based wire insulation they seem to like).
3. The problem is now more frequent. I cannot pin it to any specific conditions, it seems to start randomly and sometimes goes away after a half hour of driving but other times doesn’t.
4. I got a better video illustrating it. Here’s 71F and climbing while it’s snowing hard in rural NH:

https://imgur.com/a/uZD0Grg

(filmed by a passenger, so no unsafe driving here).
5. I have an upcoming appointment at Deadham. It’s my first visit to this particular SC, but I’ve heard good things. Hopefully with multiple videos I won’t have to fight too much.

 

[danarcha]

One of the common ways to refer to a model 3 long range all wheel drive with an acceleration boost, no?

I've been on here for a year and never seen 3D+ used before. What does the D stand for in this instance?

 

[ilyak]

I've been on here for a year and never seen 3D+ used before. What does the D stand for in this instance?

Dual motor. Does it really matter that much?

 

[ilyak]

Just received an estimate for the upcoming appointment at Deadham. Full heater replacement (likely not going to be needed, heater works except for when the ambient temperature readings are too high for it to do anything) and, interestingly, SB-20-17-010 (check circuits at X-950 and X-952). Suspect the latter will be the ultimate culprit.

 

[ilyak]

Final (hopefully) update: the SC installed an overlay harness from X901F to X527 ("last mile" of the circuit between the front body controller and the ambient temp sensor) and replaced the sensor itself once again. So far everything is back to normal, so this was an electric connectivity issue after all.