View Single Post
Old 02-12-2019, 10:20 PM   #213
JasonACT
Away on leave
 
Join Date: Apr 2019
Location: ACT
Posts: 1,732
Tech Writer: Recognition for the technical writers of AFF - Issue reason: Outstanding work on the FG ICC issues. Technical Contributor: For members who share their technical expertise. - Issue reason: The insane amount of work he has put into the Falcon FG ICC is unbelievable. He has shared everything he has done and made a great deal of it available to us all. He has definitely helped a great deal of us with no personal gains to himself. 
Default Re: FORD technical service bulletin : ICC touch screen display

Quote:
Originally Posted by Ansith View Post
Awesome work! Looks like you've finally go the answer on what actually fails on these.
Thanks. But while I'm sure both USB and Bluetooth chips stopped working properly, I'm not sure that is the end of it...

Yesterday and tonight, I worked on programming up a microcontroller to simulate the Bluetooth chip (because, while the brand I need were very popular 10 years ago, they don't seem to be in favour anymore - so I can't get one in a module I can use - I can get a 2.0 version but it initialises at 9600 baud - the 2.1 versions I need do 115200 on startup). The simulation will perform the initial connection, and respond with recorded data from the working unit, but while I can get it to connect ok - the main CPU never requests anything more than when the broken chip was still installed.

Importantly, the "break" signal (which I've also simulated, but was definitely not working with the real chip) isn't responded to in the same way as the working unit. But if I restart the connection routine from the "Bluetooth side" in the simulation, using a short timer, then the main CPU does go through the connection process again to completion.

From what I can tell, the iThing authentication chip is the next item on the list to be accessed when Bluetooth starts up, but it never gets accessed after the BT initialisation - and in fact the working unit has two "glitches" showing on the analyser (I posted a picture about the iThing) before a major read happens and I only get one glitch on the broken unit.

I had pulled two level shifting chips off the board (U408 & U1003) because the errata for the chips says "don't use them for new designs" as they have some serious faults. Testing tonight, I still didn't see any activity on the two wire bus they are meant to convert... Except once, when it did start to be very active - but I tried again quite a few times and nothing ever happened again. I've since put the chips back, but still no activity on any other attempts.

I suspect a 3rd casualty in this messy ongoing war. I wonder if the boards that die on their own are simpler to fix? But having said that, I am glad I've got hold of one that has the plethora of issues.
JasonACT is offline   Reply With Quote
4 users like this post: