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

Daughter Board: - HomeLink & Tune-in issues

This site may earn commission on affiliate links.
It's been almost 4 glorious years of ownership that I've been relatively free from serious problems. Although I must admit year-on-year, as the various updated screen layouts have been installed the screen functionality has gradually deteriorated! As in making a phone call & finding keypad to select dept. options - unsafe!

In October, my Model S85 really started playing up, to the extent; regularly certain basic functions started ceasing to work, needing either a Thumb-Wheel reset, or intervention from a higher Tesla Technical source, somewhere in the ether.

After a long drawn out series of diagnostic tests, I was told there was no option but I /they had to replace my 'Daughter' board (DB), expecting this to be F.O.C. - fat chance!

I'd no option but to pay well over £500.00 ($675+) & lose my beloved car for 2 & 1/2 days! Thereby involving me in the cost of personal taxi collection from Tesla & return 2 days later to their service centre some 45 minutes’ drive to retrieve my S85.

On getting my car back there were a few minor problems, e.g. accessing the vehicle, getting the ECU or equiv. to initiate, opening the charging flap & getting some of the screen buttons to respond.

The biggest issue at that point was my ‘HomeLink’ garage door system. This now failed miserably! I spent several hours (& eventually with a Tesla mobile Ranger) trying to diagnose why the 'Chamberlain' PCB receiver, which acts as the interface (necessary in Europe to communicate with my German ‘Hörmann’ garage door) to recommence opening & closing, when leaving, or returning to the integral garage in my home.

Eventually, I was forced to shell out the cost of a replacement ‘Chamberlain’ PCB , not forgetting the cost of a professional electrician to install it! Thereby racking up nearly another £200.00! Now at last, I can report it’s finally functioning as it did with my old/ previous 'DB'

However, it was barely 10 days after this replacement 'DB,' had been fitted, when my 'Tune-In' radio was failing with intermittent signal reception. It then completely stopped working! Subsequently I’ve had numerous conversations with different Tesla technicians, all promising to rectify this fault; eventually the last one admitting failure to diagnose the greyed-out screen function buttons. Incidentally, I normally only listen to internet radio, preferring retro-radio stations; as these sadly are not available on normal terrestrial broadcasting.

So I’m now having to wait, yet again, not days, but weeks! This is until in the new year when I’m promised a further visit from another Tesla Mobile Ranger to hopefully fix it?

What’s really upsetting, when chatting to one of their techies in confidence, he advised the reason why my original DB failed, was most likely caused by Tesla Tech-HQ transmitting too many updates, without deleting the old, or previous versions, thereby blowing up my original board!

I wonder who else is, or has suffered similar indignities?
 
...my 'Tune-In' radio was failing with intermittent signal reception. It then completely stopped working! Subsequently I’ve had numerous conversations with different Tesla technicians, all promising to rectify this fault; eventually the last one admitting failure to diagnose the greyed-out screen function buttons. Incidentally, I normally only listen to internet radio, preferring retro-radio stations; as these sadly are not available on normal terrestrial broadcasting.

What’s really upsetting, when chatting to one of their techies in confidence, he advised the reason why my original DB failed, was most likely caused by Tesla Tech-HQ transmitting too many updates, without deleting the old, or previous versions, thereby blowing up my original board!

First, Tune In has been acting poorly for a couple of weeks here too. I heard someone else complaining, they asked me if mine was sketchy too. So, not your car.

Updates, without deleting the old, or previous would not be the cause of a problem. The storage chip on the DB is divided into four partitions. Two are large and where firmware resides. One is active and one is inactive. the active one is where the current working version of firmware resides and loads to the CPU from. When you get a new OTA update, it copies down, and installs on the inactive one. Then the car reboots itself and the inactive one becomes the active one, the previously active one becomes the inactive one. Flip-flop. And when it downloads for installation, it overwrites the previous on that was sitting there. It takes care of it - itself, so that statement is just a bunch of poppy-cock.
 
First, Tune In has been acting poorly for a couple of weeks here too. I heard someone else complaining, they asked me if mine was sketchy too. So, not your car.

Updates, without deleting the old, or previous would not be the cause of a problem. The storage chip on the DB is divided into four partitions. Two are large and where firmware resides. One is active and one is inactive. the active one is where the current working version of firmware resides and loads to the CPU from. When you get a new OTA update, it copies down, and installs on the inactive one. Then the car reboots itself and the inactive one becomes the active one, the previously active one becomes the inactive one. Flip-flop. And when it downloads for installation, it overwrites the previous on that was sitting there. It takes care of it - itself, so that statement is just a bunch of poppy-cock.

I've just read your 'Poppy-cock' comment & wondered with natural scepticism, how you know so much about the way/ method of OTA updates & how they are handled within the Daughter Board.

I'm keen to know your background & qualifications to advise this info? Lastly, to enquire if you're directly, or indirectly employed by Tesla?
 
Hi I read and I ask questions. And I have been studying from these sources for a couple of years now. Here's one that is easy to find. Unofficial Tesla Tech from a regular poster here on TMC.

Thanks for the link & info. It's concerning to read some of the comments on this 'UTT' Wiki site, that all Teslas MCUs have built-in obsolence, due to the poor quality of the eMMC (Hynix) that Tesla have chosen to use. It seems I've been lucky so far, that the entire MCU has yet to fail! I guess I'll need to budget for this, as I plan to I keep the car.
 
Me too. I plan to keep this one.

We can reflect accurately on "how could this have happened" - now. But its also easy, I think, to believe that engineers at Tesla (at the time, if they are still there) could not have forseen this problem. This Tegra board was an off-the-shelf component from nVidia in 2007. It was used in an Audio A7's infotainment system. Tesla started designing the MCU and its screen components in 2010, and it was still current tech at the time. Tesla asked nVidia to mod the power connections, but it was pretty much a product that nVidia was already making. I mean, you stop and consider, Tesla was a fledgling company. Those of us that put a $5,000 deposit down in 2012 didn't know for sure that we would even see our cars. I believe Tesla took any/all the closest to off-the-shelf stuff that they could to build those first years of cars. I guess you know the turn signal stalk for example and the window switches came from MB. Who knows what else. Yes, by 2015 they should have recognized there was a problem. Yes, we've heard from many 3rd party fixers that they replace the Hynix chips with Swissbit 32's or 64's that are what they call industrial strength too.

There's some interesting data that I have collected tracking MCU1/eMMC failures reported on TM and TMC for the last year+. I don't have information on every failed MCU1/eMMC. But I have tried to fairly collect every TM/TMC reported repair. I believe I have a representative picture of the total. I don't know what number X is to multiply my numbers against to get a total worldwide. But I believe this is an equal fraction of the real count.

The Tegra board uses three different versions of a Hynix 8 GB chip. 2012 & 2013 use one. 2014 uses one. And 2015-2018Q1 uses the third during all those years. The least failed year from 2012- thru 2015 is 2015. That's strange considering the youngest of those years has the most failures. Then the same chip between 2015 thru 2018Q1 - 2015 has the most failures. What makes 2015 failure more than 2016-2018Q1? No one knows that has reported. But what is most different about 2015 - 2018Q1? AP1. All of 2015 and until Sep of 2016Q3 are AP1. So why is 2016 failures for 3 quarters of 2016 44% of the 2015 failures? These are mysteries. (BTW, AP1 came out in late Sep 2014, so we can't say AP1 is the cause of 2014 failures.)