MCU failure rate poll; please vote *whether or not* you've had a failure

[KarenRei]

Just doing a rough assessment as to the reliability of Tesla MCUs (regardless of version - there's not enough poll options to break it down by version). If you have multiple Teslas, please vote multiple times, once for each vehicle. Thanks!

If you share this poll, please only do so in threads where there's not likely to be sampling biases. E.g. if you share it in a thread where people are discussing MCU problems, you'll bias the sampling towards people with problems; I'm trying to get an unbiased sampling. Thanks!

 

[Tam]

...MCU problems...

That depends on what problems you are looking for.

Some brand new S & X got a yellow band problem and they got replaced. It's a failure of fidelity display but the car as a whole does not fail as it can drive away fine. There have been quite numerous threads and long discussions about this failure.

Older S & X also got lots of threads about MCU bubbles too!

 

[KarenRei]

That depends on what problems you are looking for.

Some brand new S & X got a yellow band problem and they got replaced. It's a failure of fidelity display but the car as a whole does not fail as it can drive away fine. There have been quite numerous threads and long discussions about this failure.

Older S & X also got lots of threads about MCU bubbles too!

I was mainly thinking of the flash problems (the recent news topic), or any other complete failure mechanisms. They were replacing MCUs for yellow bands? I thought they were just doing a UV cure for those.

Hmm, I would thus encourage people to describe what sort of problems they had.

 

[Az_Rael]

2015 P85D had an MCU fail in 2018 for the flash issue. Replaced under warranty (CPO)

 

[Missile Toad]

At almost 3 hours since this poll was posted, I did the following analysis.
1. I discarded stats for <1 year (infant mortality)
2. I summed the average years for the 'happy campers' (31 people, w/ mid-point in the years, being counted as their 'ownership period');
3. I summed the average years for the 'WTF eMMC failure?' (3 people, as per #2).
4. totaled the good (#2) with the bad (#3) years, to reach 103 car-years of operation.
5. Divided the result of #4 by the 'unhappy campers: 103/3.

Assuming that these accounts are representative of the larger ownership... it looks like there are 3 failures in 103 years of operation. Having said that, if my math and assumptions are correct, we can expect around 1 failure in 34 years of operation.

My assumptions:

1. Each respondent drives a normal amount, annually, and accordingly, exercises the eMMC logging the typical amount;
2. We are not getting unusually high responses from the 'squeaky wheel gets the grease' crowd, i.e., unaffected people are equally as likely to respond;
3. My sample set is big enough <- Hah!, there are only 3 'unhappy campers'... so more data is needed.

 

[aaron0k]

Count my vote as two. I've had MCU failures on two MS. Coincidentally both happened at 84k miles.

 

[Missile Toad]

Count my vote as two. I've had MCU failures on two MS. Coincidentally both happened at 84k miles.

OK, given that the 3 eMMC replacements are closer to 4, I revise my average years between failures to 27-28 years.

 

[mongo]

I was mainly thinking of the flash problems (the recent news topic), or any other complete failure mechanisms. They were replacing MCUs for yellow bands? I thought they were just doing a UV cure for those.

Hmm, I would thus encourage people to describe what sort of problems they had.

I believe flash wear is due to logging and so is proportional to time awake. That may make it more correlated with mileage (though speed dependant) than chronological age.

 

[aaron0k]

OK, given that the 3 eMMC replacements are closer to 4, I revise my average years between failures to 27-28 years.

I'm assuming by your model... you're an MCU1 owner. I think you might be eating crow sooner than you know...

 

[Right_Said_Fred]

If the faulure score started or is starting to go up, it may mean the TESLAQ crowd discovered the poll.

 

[fajitamondays]

Reporting as an outlier here.

MCU on my Model 3 failed at 200 miles! It started rebooting intermittently while driving, and progressively got worse as the issue took hold (I.e. it would freeze/reboot multiple times in a 10 minute drive).

I never did ask for a root cause. Attached is the invoice from that service center visit.

 

[jerry33]

Replaced at 6.5 years 126K miles. Car lives outside in Texas. I think that is pretty good as cabin overheat protection didn't start until a couple of years ago. Display was black, car drove normally.

 

[KarenRei]

Reporting as an outlier here.

MCU on my Model 3 failed at 200 miles! It started rebooting intermittently while driving, and progressively got worse as the issue took hold (I.e. it would freeze/reboot multiple times in a 10 minute drive).

I never did ask for a root cause. Attached is the invoice from that service center visit.

Bathtub curve.

 

[Missile Toad]

9 (or is it 10?) failures in about 300 years. Looks like failure rate is holding steady (notwithstanding TESLAQ folks) at 30-33 Tesla years between failures. Lets see if more data skews things either way.

I'm omitting the < 1 year reports (per bathtub curve, above).

 

[KarenRei]

Poll added: For those who had a failure, what type of MCU failure did you have?

MCU failure poll #2: what *type* of failure did you have?

Thanks for participating! This is really useful data

 

[jerry33]

Poll added: For those who had a failure, what type of MCU failure did you have?

MCU failure poll #2: what *type* of failure did you have?

Thanks for participating! This is really useful data

Might need one more poll for climate (hot and cold). Of course, everyone now has cabin overheat, but that's fairly recent so the older cars have had several summers.

 

[KarenRei]

Might need one more poll for climate (hot and cold). Of course, everyone now has cabin overheat, but that's fairly recent so the older cars have had several summers.

Yeah, there's a lot of nuance that can't be captured by forum polls, unfortuantely. Next time I should just use a Google poll, so multiple questions can be asked at once. But this will have to do for MCU polling.

Also, heat should only affect some sorts of failures, like dead screens. Flash failures, for example, should be write / age related; phantom touch should be manufacturing defects; etc.

 

[aaron0k]

This 'poll' is flawed. It should be by mileage/operational hours not age of the car. MCU failures via eMMC wear aren't going to be an issue until >60-70k (IMO).

 

[KarenRei]

This 'poll' is flawed. It should be by mileage/operational hours not age of the car. MCU failures via eMMC wear aren't going to be an issue until >60-70k (IMO).

There's many different axes of relevance to consider for MCU failures (age, mileage, MCU version, climate, type of failure, etc). It's hard to capture more than a single axis with TMC polls. As age "touches on" several aspects (suggestive of the MCU version, suggestive of warranty coverage, correlates directly with time-sensitive failures and loosely correlates with mileage and climate-related failures (e.g. with a large enough sample size, variations in mileage per year and climates cancel out) ), it's the one that I chose. It's the best you can do with a TMC poll. I'm sorry that it doesn't spark joy. As I said, if I were to do it again, I'd do it as a Google poll and just post a link here.

 

[scaesare]

eMMC flash failed at ~6.5 yrs and 145K miles.

Replaced it myself.

Does the touchscreen leaking goop count as a pending failure?