Welcome to My World, DEAD Jag ECU and What To Do About It
I am still learning electronics. There is so much to learn. I always like a challenge. A shop had a programming event on a Jaguar using SDD and sent me the original and a donor to see if I could make one or both work. The last time I did one I learned that the security built into SDD will not allow you to program an engine module alone on the bench. No warning, it just does not work. After three days of experimenting I put SDD away with a few curse words. So for this project, I decided to use the Autologic. I currently own the Assist Plus. It does work on the bench. In order to take advantage of programming you have to own the PRO version software for Jaguar.
The first thing I did was obtain a pinout of the ECU. I decided to mark all the critical grounds and powers on the print making connections easier. Once you can think like an Englader, you can follow their diagrams quite well. I connected up the K-Line, CAN leads, Grounds, B+ and Switched Powers. It is important that the switched leads be separate from B+ with this ECU. Part of the process involves key on and key off. I used wires meant for use with a breadboard to connect to the ECU pins. If I get a few more of the Jaguar jobs, I may try to get a pigtail so that I can connect to the ECU easier.
Here is an example of the leads used. You can just strip wires as you need them. I use Red for B+, Orange for Switched Power, Black or Green for Ground, Yellow and Blue for CAN. Whatever available after that for other leads like the K Line. I connect them through a home made breakout box that has switches to simulate battery disconnect and key switch.
After taking the ECU apart, I tried to read the EEPROM memory in circuit. This did not work. I had to desolder the EEPROM Chips and read them. This is a common problem. When you power up the EEPROM it also powers the system clock which starts up and wreaks havoc. Some circuits you can just use a chip clip and read. I like those. After reading, I soldered them back on the boards. I used VVDI Programmer to read the EEPROM. There are many solutions out there. So far, I like it the best. Sometimes you have to try several programmers to find one that works well with a particular EEPROM. The AR32A is a solid programmer. It requires a parallel port and 32 bit operating system to function correctly.
It appears that even in 2004, Denso has encrypted the HEX data. During programming with SDD, the VID block was corrupted. It is stored in one EEPROM or both. Both were damaged but for this article I am only showing one. The first picture is that of the donor. You can see that there is a VIN at the top of the page but the numbers are broken up and the order of the digits changed. I tried to determine a pattern but it is too complex for my brain.
The picture below contains the corrupted VID block. The EEPROM also contains vehicle order and immobilizer data. You can see much of the data is gone.
Fortunately Autologic has a Create VID Block feature that can get you back on the road. It will rewrite the VIN basic features like tire size, transmission, 2 or 4WD, ect... Much of the data is still missing. Theoretically it is enough to allow the shop to learn keys and get back on the road. The rebuilt VID block / EEPROM below.
Programming with the Autologic is simple. Just follow the prompts. If there are multiple calibrations look at the vehicle VIN and current calibration. Usually it counts up alphabetically. A to B, C, D ect.. at the end of the calibration ID. There is a warning about the air temp sensor. since the ECU is at room temp and the sensor disconnected, I ignored and moved on. The process went smoothly, no problems. I read the created VID block before programming and then wrote it back after programming. As far as I can tell on the bench, all is Golden. VIN reads correctly, VID block Built, Latest Calibration installed.
So now the ECU is updated to the latest calibration and there is a VID block. I am currently looking for a way to deal with the encrypted EEPROM. I suspect the encryption is native to the Renesas MCU running the board. If so, maybe by reading the EEPROM through the MCU the data will be clear. This would allow me to rewrite the VIN and identify vehicle features in the HEX data. The caveat to that is if the main processor is toast, I would still need to decipher the EEPROM with an algorithm.
I have done a Jaguar ECU before. It that case, the EEPROM was intact but the MCU (main processor) was fried. I was able to write the information from the EEPROM to a donor unit and reprogram with the correct updated calibration. In this case the EEPROM was corrupted and I had to go to plan B.
From my conversation with the customer, there are still steps that need to be taken to get the vehicle to run. This one is not plug and play. Key learning and reflashing a few of the other modules is required. Maybe that helps fill in some of the missing data in EEPROM. Lots of experiments to be done, little time to do them.
If this is all gibberish to you, I was there once too. It took hundreds of hours of research to get to this point. I am fascinated by it so it draws me in time and time again. A big jigsaw puzzle that you can't get out of your head until it is done.
Thanks for sharing this Mike! I'm definitely not on that level but I find it fascinating as well. Look forward to learning more and hearing how this turns me out.