Turn Signal intermittently doesn't work

[metroplex]

So I thought it was my imagination at first, but I can confirm my 22 M3LR turn signal intermittently doesn't want to work, whether quick tapping or full activation. It's during the oddest times and I haven't found any correlation.

Sometimes its leaving my driveway and going to the first intersection of the morning. Others, it was to pass someone at near WOT and pushing the stalk like 3-5 times yielded no signaling whatsoever... but the lane assist didn't kick in for any of these situations.

I'm guessing the turn signal stalk sent a signal to the computer saying I was signaling, but I didn't hear the turn signal indicator nor did the blind spot camera come on. But I wasn't fighting the lane assist. Normally if I change lanes without touching the turn signal, it would try to prevent me from steering into the lane.

As a test I was cruising down my street and was able to use my turn signals for left/right whether quick tap or full activation so obviously it's working, but in the 1 out of 20 times it just doesn't want to work.

Has anyone else encountered this?

 

[Gauss Guzzler]

Yes, there is a defect that affects all years of Model 3 and Model Y. You have to hold the stalk for a fraction of a second in order for the input to register. It doesn't matter whether it's a partial or full press, whether your initiating or canceling, or how the signal function is configured in the settings. You just have to hold the stalk for long enough to register. Oddly, the gearshift does not have this same flaw.

 

[metroplex]

I don’t always have to hold it. When I drove slowly down my street I was gently tapping it and it would work.

During the times it didn’t work I’d would quickly press it like 5 times rapidly to try and make it work but no dice. The lane assist doesn’t kick in so the stalk is providing input but something isn’t activating the indicator lights and camera etc.

 

[Gauss Guzzler]

No, you have to hold it for at least a fraction of a second. Try it out and you'll see that it's easy to duplicate. Tap and release very quickly and it fails 100% of the time, tap and release slowly and it works 100% of the time. This holds true for both directions.

 

[metroplex]

No, you have to hold it for at least a fraction of a second. Try it out and you'll see that it's easy to duplicate. Tap and release very quickly and it fails 100% of the time, tap and release slowly and it works 100% of the time. This holds true for both directions.

I'm trying to explain that this is not the case for me. I can tap and release quickly and it works almost all of the time. I repeated this down my street. Both directions. I was clicking it gently, rapidly, held it, etc just to experiment. It worked fine.

The times where it doesn't work, it doesn't matter what I do. The turn signal just doesn't work. Both directions.

 

[Soul Surfer]

I'm trying to explain that this is not the case for me. I can tap and release quickly and it works almost all of the time. I repeated this down my street. Both directions. I was clicking it gently, rapidly, held it, etc just to experiment. It worked fine.

The times where it doesn't work, it doesn't matter what I do. The turn signal just doesn't work. Both directions.

This process is like driving a European car. @Gauss Guzzler is correct. You have to just not rush and do it. Even my Corvette had this feature and a tap didn’t work. I had to hold for just a little. Perhaps your flicking rapidly sets off an error or timeout. I’ve never experienced this on mine as I’ve been used to driving cars in Europe and most of my cars had this functionality as well.

Lastly, have you put in a service request? Be sure to include date and times when this happens. It makes their debug process, so much easier to isolate the events in the car’s logs.

 

[FOOLSLFDRVNG]

I have this issue, they want me to get a video.

 

[andyaycw]

Yes, I have the same issue. I've just chalked it up to the fact that I have to re-train my brain to get used to holding the turn signal about a 1/2 second longer than normal to make sure the car registers the fact that I'm trying to signal.

 

[metroplex]

Everyone is missing the fact I don't have to hold it for half a second or more. I can rapidly tap it and release and it works. Until it doesn't work and then nothing makes it work. This ain't my first car with the half tap quick indicator. Ford has had this since the 2010s and I'm well aware of the functionality. I can even adjust it the number of blinks on my Fords for each tap.

The problem with my Tesla is that it intermittently doesn't accept any type of tapping or holding. Nothing. No indicator lights. But lane assist doesn't kick in, so it is reading a turn signal input.

Like I said a few times already. I purposely tapped the stalk rapidly like a split second. Sometimes not even fully tapped it, and it works left and right.

The times it doesn't work is when it really doesn't want to work no matter if I hold it down or whatever.

It's not just quick tap when it doesn't work. A full tap doesn't even turn the indicators on. Again this is when it doesn't want to work.

During my testing to try and reproduce this issue, it'd blink as soon as I tapped it or even moved the stalk. It's sensitive enough not to have to hold it down half a second.

 

[stopcrazypp]

Turn signal stalk sometimes does not seem to do anything?

Read through the whole thread and see if any conditions apply to you. Changing the direction rapidly can cause this (signaling one direction and immediately signaling other). Some also mention a third mode (#1 being half detent for 3 blink, #2 full for continuous blink, #3 hold stalk for it to continually blink until you release it). There is also a TACC signal function, where holding the stalk accelerates the car until you release it.

 

[bbell]

Call for mobile service, explain the stalk is flakey and doesn't always register an input properly. I had mine replaced due to this, left turns only. It's a quick swap. Problem solved after that. If you can video it even better when you set up the service call.

Invoice says "Replaced SCCM".

 

[EVRider-FL]

You’re under warranty, no reason not to make a service appointment.

 

[metroplex]

You’re under warranty, no reason not to make a service appointment.

Looking at the other thread people have had the stalk replaced number of times with the same result. Seems like a software issue.

 

[EVRider-FL]

Looking at the other thread people have had the stalk replaced number of times with the same result. Seems like a software issue.

Maybe, but you won’t know unless you make a service appointment.

 

[Gauss Guzzler]

This ain't my first car with the half tap quick indicator...It's not just quick tap when it doesn't work. A full tap doesn't even turn the indicators on.

It doesn't matter how far you push it, only how long it takes. There's a narrow window of acceptable durations - too short and it's ignored, too long and it triggers the "third mode" temporary signaling that @stopcrazypp mentioned, which will blink but immediately self-cancel upon release of the stalk.

Try it out in your garage. You can easily press the stalk all the way past the detent, even all the way against the hard stop quickly enough to be ignored. Or slowly enough to trigger the temporary mode. You'll quickly see how confusing the signal response is when messing with it like this. Then try it with the gearshift and you'll see that it works as expected, never missing a shift no matter how brief or long the press is.

In addition to this egregious defect, there's also very poor hysteresis logic which fails to resolve ambiguity when making multiple lane changes or rapidly changing directions. This flaw is more pronounced with the new "auto canceling" feature which stupidly disables the ability to latch the signal by pressing beyond the detent. But it doesn't sound like this is the OP's situation.

It's very interesting that you seem to have found a condition where the lane departure warning is averted without actually blinking the signal. As I noted earlier, the gearshift uses identical hardware but does not have this time-sensitive flaw, so I figured they must be using a lower priority polling interval for the turn signal which could be a hardware limitation of some microcontroller somewhere in the system. But your anecdote suggests that the stalk is being read properly, but only partially processed on the software side. That'd be good news since Tesla is much more likely to fix a software flaw on their own accord, whereas a hardware flaw would likely require a NHTSA recall and/or class action suit.

Meanwhile, NHTSA is obligated to force a recall for this obvious and clearly demonstrable safety defect. And anyone who has suffered damages in a collision where lack of signaling was a factor can cite this defect to make a slam-dunk legal case against Tesla. I hate to promote America's toxic litigiousness, but Tesla has known about this safety issue for over 6 years and they've done nothing to address it.

 

[ochompsky]

i made a service appointment and they just said "cant reproduce" which i understand because it's intermittent like yours OP.

the internet is littered with unresolved complaints about this. you can count on tesla to not be able to implement something this basic without bugs.

 

[metroplex]

It doesn't matter how far you push it, only how long it takes. There's a narrow window of acceptable durations - too short and it's ignored, too long and it triggers the "third mode" temporary signaling that @stopcrazypp mentioned, which will blink but immediately self-cancel upon release of the stalk.

Try it out in your garage. You can easily press the stalk all the way past the detent, even all the way against the hard stop quickly enough to be ignored. Or slowly enough to trigger the temporary mode. You'll quickly see how confusing the signal response is when messing with it like this. Then try it with the gearshift and you'll see that it works as expected, never missing a shift no matter how brief or long the press is.

In addition to this egregious defect, there's also very poor hysteresis logic which fails to resolve ambiguity when making multiple lane changes or rapidly changing directions. This flaw is more pronounced with the new "auto canceling" feature which stupidly disables the ability to latch the signal by pressing beyond the detent. But it doesn't sound like this is the OP's situation.

It's very interesting that you seem to have found a condition where the lane departure warning is averted without actually blinking the signal. As I noted earlier, the gearshift uses identical hardware but does not have this time-sensitive flaw, so I figured they must be using a lower priority polling interval for the turn signal which could be a hardware limitation of some microcontroller somewhere in the system. But your anecdote suggests that the stalk is being read properly, but only partially processed on the software side. That'd be good news since Tesla is much more likely to fix a software flaw on their own accord, whereas a hardware flaw would likely require a NHTSA recall and/or class action suit.

Meanwhile, NHTSA is obligated to force a recall for this obvious and clearly demonstrable safety defect. And anyone who has suffered damages in a collision where lack of signaling was a factor can cite this defect to make a slam-dunk legal case against Tesla. I hate to promote America's toxic litigiousness, but Tesla has known about this safety issue for over 6 years and they've done nothing to address it.

It works fine in my garage and on the street. I can rapidly tap it however I want and it works. It's the 1 out of 10 times where it doesn't work that frustrates me and its difficult to reproduce. It has 0% to do with how long I hold the stalk because I have confirmed the stalk is sensitive enough to register the input.

And in each of the instances where the turn signal doesn't engage, I can change lanes without the lane assist fighting my steering wheel. So the computer/brains obviously recognizes I am trying to signal a lane change. If I drive like a BMW driver (just kidding) and change lanes without signaling and cross the lanes, it will try to fight me based on my setting (lane assist and not warning).

 

[metroplex]

i made a service appointment and they just said "cant reproduce" which i understand because it's intermittent like yours OP.

the internet is littered with unresolved complaints about this. you can count on tesla to not be able to implement something this basic without bugs.

It's frustrating indeed! Lots of similar complaints. I heard someone suggest filing a bug report as soon as an issue is encountered. I am not sure how effective this is but I've done it a few times and the screen says it was submitted. I'll keep doing that for phantom braking events as well. My M3 has a tough time with the construction barrels and traffic lights around SE MI. Loves to tap the "brakes" and piss people off behind me.

 

[beachsideev]

My last m3 did this too. Sometimes just wouldn't register the turn signal. It was random and annoying.

 

[Gauss Guzzler]

I can rapidly tap it however I want and it works. It has 0% to do with how long I hold the stalk because I have confirmed the stalk is sensitive enough to register the input.

Interesting. I found it easy to get it to fail on an early Ryzen 2022 3LR. Perhaps they actually do have a fix? Or it somehow only affects some cars?