What is the actual issue?
MCU 1 has 8GB (Hynix) of soldered-on flash / solid-state memory card which is located on a smaller daughterboard – All Tesla cars built before March 2018 have this chip and like all flash memory it only has a certain number of writes before it will slow/fail, effectively Tesla have been writing logs to the chip too quickly which has caused <we don’t have an exact number> of them to slow down or in some cases cause the main screen to completely die.
Tesla can fix the issue but will charge you £1800-£2500+VAT + Labour to replace the whole MCU instead of just soldering on a new £20-£50 chip.
In all cars built after March-May 2018 (if you’re unsure read this or if you have Netflix or games like Cuphead you’re not affected) the chip was upgraded to a higher capacity so it’s much less of an issue.
A few more quotes from experts (from this article) on this:
“The main issue is that this excessive log file writing causes eMMC flash wear. Flash memory is generally only rated for some tens of thousands of write cycles. What happens is that the flash memory starts to fail when writings can no longer be completed. When one block fails, parts of the firmware may also become unreadable, leading to poor operation or failure of the MCU completely.”
“The filesystem in MCUv1 is handled on a NAND-based eMMC flash chip. Although these are solid-state and great for automotive use, there is one pretty serious drawback. Each memory bit on a flash chip can only be written to a limited number of times before data gets corrupt – and that bit can no longer reliably store a 0 or a 1.”
“Tesla selected a flash chip that is unable to handle the constant read/write functions. These chips have since been replaced with a more robust version.”
What have Tesla said so far on this
Within our meeting with Tesla it was stated that:
The highest level of engineering looking into it to stop the problem happening rather than to repair.
Currently, Tesla’s stance seems unchanged, sadly it is a full MCU change if yours fails, we raised the fact this goes against Tesla’s mission as nobody wants to see a computer being scrapped over a ~£20 component, it would seem it’s down to Tesla US to solve this. Cars will continue to be addressed on a case by case basis.
Tesla have also said:
“Tesla continually improve all aspects of the vehicles firmware. Some of these will include life-extending enhancements, customer feature advancements and core level updates. The core level updates along with the life-extending enhancements, will not necessarily be noticeable to the owner.”
Can service departments gauge the status of an eMMC chip?
The service department have access to multiple automated diagnostic tools to investigate issues experienced by the customer however, we do NOT have a tool to check the overall health of the eMMC. Infotainment engineers may have limited access to interpret some data from the memory, but this is not shared with nor is it needed by service technicians.
What has Tesla Owners UK done to help?
A full list of everything we’ve done to date on this can be found on this page (note you will need to be logged in to view).
We keep requesting Tesla UK manages this problem proactively without owners having to fight and request cover under Consumer Law constantly. We will continue to maintain pressure on Tesla to fix these issues during our meetings. We’ve also put together this document which lists 3rd party companies that can do the work for less than Tesla would do it. We’ve increased the information about Consumer Rights Act 2015 and Sales of Goods Act on this page as well as adding all warranty documents for all cars.
What has actually happened to fix this?
Software changes have gone out to reduce the logging of data to these chips, this reduces the damage the software is having on the chip, however, the issue appears to still be a concern for many owners, especially owners that want to keep their cars for 10+ years.
I’m confused can this be explained better?
UK consumer law is pretty clear so worst case this should protect owners but what exactly protects me?
In short, a product should last a reasonable period for the expected length of ownership of the product. We suggest the following action if you’re concerned:
- Request a remote diagnostics from Tesla if you’re seeing slower than previous MCU
- Request to find out the MCU health
- If it’s failed or failing request a replacement/fix
- Cite consumer law if required
- Escalate using these methods if you believe the service manager of your service centre has failed to help you when it should be sorted under consumer law.
- Take Tesla to court over the issue if required or https://www.gov.uk/make-court-claim-for-money or https://www.moneyclaim.gov.uk/web/mcol/welcome
Is there anyway Tesla can give an indication to an owner of MCU health?
We’ve not heard an official answer on this but at least one owner was told the % health of his eMMC/MCU. So if that’s the case with you request a service appointment and escalate it using this method (listed at the bottom).
What are the signs?
It’s hard to give exact answers here as some problems could be simply problems with the latest software update, that said the following are known symptoms from owners that have had complete failures:
- Sluggishness of the systems
- Reboots occurring spontaneously
- Browser slower than ever
- Problems rendering maps
- Reboots start to take longer – several minutes
- After each reboot, some features would be broke
- In the end, reboots would take up to two hours and…
The final sign you have a problem is a main screen dead and no reboots fix it.
Update: You can view the data from our members that had the issue here.
What can be done about it?
- You wait for it to stop working and get it fixed under warranty (if you’re still within warranty)
- You wait for it to stop working and get it fixed under UK consumer law (e.g. Consumer Rights Act 2015, as explained on this page.
- You proactively get a 3rd party to replace the chip (Warning: doing so could void any existing warranty) – Discounts are available for this on our supporter discount page
- You pay Tesla for a replacement MCU1
- You pay Tesla for a replacement MCU2 (when they eventually allow it)
What are the risks with paying for a 3rd party upgrade?
- If the chip has already failed it will be impossible for the 3rd party company to replace as they need a copy of the files/certificates on the chip, if yours has failed the only method is to speak with Tesla unless you’re happy with no app access and no further software updates, ever!
- If the certificates can’t be read from the failing chip and you get a repair from a 3rd party, Tesla won’t help you with sorting new certificates to the new chip (they will however if you buy a brand new MCU from them mind you).
What companies are offering 3rd party replacement chips?
NOTE: We’ve not tested these services so therefore don’t condone their actions by proceeding you most likely will void any warranty and/or any future claims under consumer law. If you wish to seek further advice we still suggest speaking to Tesla first to address the issue, if a service manager doesn’t wish to sort then please escalate using this method.
- Gary at EV LINK [there is a discount for Tesla Owners UK Supporters] in Milton Keynes / Buckingham (expect to pay £500-£650 for a 64GB upgrade with a 12-month guarantee) – Several UK owners have used this service and recommend it
- Laadkabelwinkel (often this is a Swissbit 64GB industrial grade eMMC) – Non UK
The price seems to vary depending on how corrupt/degraded the chip is. Bad chips may take twice as long to fix, which will increase the labour costs. Considering Tesla’s cost to replace the whole unit this should be a cost-effective approach for those outside of warranty and/or outside of the 6 year Consumer Rights Act 2015 protection. It’s not clear what warranty (if any is provided)
We will update once we know of any other UK locations offering this service.
In the past month I've done repairs/replacements on over a dozen @Tesla MCUv1 units for customers suffering from eMMC flash failure.@elonmusk, you really need to tell the engineers to fix the logging wear in /var. It's literally killing a huge percentage of these units. 😐
— Jason Hughes (@wk057) October 9, 2019