Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Wiki Consolidated eMMC Thread (MCU repair) (Black Center Screen)

This site may earn commission on affiliate links.
Car supercharging today. Went in to use the restroom. When we came back out the center screen was black. Instrument cluster worked.

Did notice that the radio volume buttons worked we hit the right area on the black screen.

Rebooted. Came back and the audio didn't work. Rebooted again and all was OK.

Tesla keeps telling me its all ok. Had frequent black screens for a while now.
 
  • Like
Reactions: Mucall
You might be 6 days, 6 weeks, or 6 months for complete failure to boot. Good luck!

So is complete failure the official threshold? I mean clearly, they will replace it at that point, but not before? I'm looking for something concrete, like "25% of blocks in the eMMC have failed" or "Only 4096mb remain usable". Something precise that I can run a diagnostic on and have a definitive answer before I even schedule the visit.
 
So is complete failure the official threshold? I mean clearly, they will replace it at that point, but not before? I'm looking for something concrete, like "25% of blocks in the eMMC have failed" or "Only 4096mb remain usable". Something precise that I can run a diagnostic on and have a definitive answer before I even schedule the visit.
Some people have reported that the service centers do a memory check which does look for the number of bad blocks and if it's above some threshold, they will replace it.
 
So is complete failure the official threshold? I mean clearly, they will replace it at that point, but not before? I'm looking for something concrete, like "25% of blocks in the eMMC have failed" or "Only 4096mb remain usable". Something precise that I can run a diagnostic on and have a definitive answer before I even schedule the visit.

The latest firmware has their checks built in and will warn you when you need to schedule an appointment. (Or if it gets really bad it goes into a "limp" mode that only displays the backup camera, and sets other options to "safe" defaults, like HVAC on auto...)
 
  • Like
Reactions: Frankman60
The latest firmware has their checks built in and will warn you when you need to schedule an appointment. (Or if it gets really bad it goes into a "limp" mode that only displays the backup camera, and sets other options to "safe" defaults, like HVAC on auto...)

Is this a reliable message which will always display when the eMMC is past the replacement threshold? Are the conditions which lead to this message being displayed the only such conditions which warrant a replacement?
 
Now my FM radio stopped. Can still stream Bluetooth through the phone.

I’ll schedule another appt. but I’m past the warranty miles so they will probably point at something else.

one would think the savings in eventual towing and the lost goodwill would make it worthwhile to just swap boards.
 
From the official service bulletin these are the criteria to be eligible:

For the older Hynix eMMC chip (Model S VIN up to about 50000):
3. If the eMMC name is “H8G2d”, determine if the filesystem malfunctioned due to wear.
  • If alerts “MCU_w088_eMMCNeedsReplacement” or “MCU_w086_eMMCEOL” are actively asserted, recently asserted, or present in vehicle logs, continue to “Provide Rectification”.
  • If alert “MCU_w031_cidSquashfsError” is actively asserted, recently asserted, or present in vehicle logs, AND there are fewer than 65 reserve blocks remaining, continue to “Provide Rectification”.
  • If neither of the above is true, the eMMC filesystem has not malfunctioned due to wear, continue to “Vehicle Ineligible”.

For the newer Hynix eMMC chip (Model S VIN > 50.000 and all Model X):
If the eMMC name is “Hynix”, determine if the filesystem malfunctioned.
  • If alert “MCU_w031_cidSquashfsError” is actively asserted, recently asserted, or present in vehicle logs,continue to “Provide Rectification”.
  • If alert “MCU_w031_cidSquashfsError” is not present, the eMMC filesystem has not malfunctioned, continue to “Vehicle Ineligible”.

Basically this means the system already has developed extensive eMMC damage, and is currently inoperable/blackscreened.
Note Tesla does not look at the write cycles (which are only reported in the newer Hynix chip model by the way) but instead they depend almost fully on the precense of the dreaded squashfs errors in /var/log/syslog. Which looks like this:

IMG_6250.jpg
 
I had my P85D looked at by the Honolulu Tesla SC a few weeks ago to see about eMMC / daughter board replacement.

BL - They kept the car for several days, gave me a rental car, and "erased some data" to improve eMMC functionality.

I actually received a phone call from Tesla the day of my appointment. The caller must of been in Fremont, because he said some things the SC hadn't heard before and admitted he wasn't on the island. Because the daughterboard was out of stock locally, he tried to reschedule my appointment.

I told him my MCU also had a bubbling screen that acted up around the same time. Maybe they could just replace the whole MCU?

He said keep the appointment, take the car in later that day, they could "probably work something out."

Instead, the SC gets the car and says they won't replace the daughterboard because its not failed. They do some of the software work and provide me a rental car, for which I'm very grateful.

I explained the misleading phone call from Fremont (or somewhere!?) and he's frustrated. Jonathan, the SC CS agent, tries researching who is calling his customers. He didn't like that I was made to believe I would get a new daughter board and possibly a whole new MCU.

He resolved the issue by offering my 25% off the touchscreen replacement. I paid a little over $600 for them to swap the screens, but no changes to the MCU or eMMC. Good news though, the car is working better for now.
 
My eMMC was replaced a couple weeks ago under the warranty adjustment. However, it doesn't seem to have helped my problem much. Though the MCU does seem to boot a little faster and is a little more responsive, I keep getting intermittent lockups that require a reboot. Normally what happens is after some time of driving (usually over an hour), I see the LTE resetting and the outside temp going to ---. Then audio effects sometimes don't play (blinkers, etc.). Then the map gets jittery and the interface starts to lag more and more until it doesn't respond at all, and there is no audio and no further updates to the screen. Sometimes the MCU will reboot itself after a couple minutes of that final state, but sometimes I have to reboot it. :( I was previously on 2020.36.11 and now I'm on 2020.40.9.2 and it's still happening. This never happened before 2020.36.11, not that that means anything though.
 
My eMMC was replaced a couple weeks ago under the warranty adjustment. However, it doesn't seem to have helped my problem much. Though the MCU does seem to boot a little faster and is a little more responsive, I keep getting intermittent lockups that require a reboot. Normally what happens is after some time of driving (usually over an hour), I see the LTE resetting and the outside temp going to ---. Then audio effects sometimes don't play (blinkers, etc.). Then the map gets jittery and the interface starts to lag more and more until it doesn't respond at all, and there is no audio and no further updates to the screen. Sometimes the MCU will reboot itself after a couple minutes of that final state, but sometimes I have to reboot it. :( I was previously on 2020.36.11 and now I'm on 2020.40.9.2 and it's still happening. This never happened before 2020.36.11, not that that means anything though.

You likely need some factory resets, and possibly an additional update. This often happens if some of the corrupted data gets copied to the new eMMC.
 
  • Informative
Reactions: dark cloud
try 5-6. Note the fine print

Thanks Tony. Not sure I found the fine print, but one upgrade (2020.40.9.2 to 2020.48.12.1) and 6 factory resets later, we'll see what happens. I won't be doing an extended drive for another week or so.

I did notice that after each of the first 3 factory resets, my LTE would continuously reset until I did an MCU reboot. After the 4th factory reset, LTE came up immediately with no reboots, and so did the 5th and 6th resets.