I winded up with a Roxor cluster cause i got a call from a desperate Mahindra dealer that can't get them repaired under warranty, for months and for many clusters. I repair instruments clusters as a lucrative hobby.All of the Roxio clusters suffer from the FFF issue. Seems the software got corrupted on the chip or the chip itself is internaly damaged. I mostly suspect only corrupted. When you says the data access port is disabled from factory.
Do you have a bench set up so you can test them? More than just power up? Would it be possible ripping the code from the chip? I know some get trough with gm cars and trucks. After researching on the subject seems evereybody has a cluster replacement/ fix problem. My client said he had more than 2 dozen waiting for repair and no solution. Eager to hear your opinion. Tony Cluster.
The MCU is an older Freescale S12 unit. They are programmed via the background debug pin. There is a programming option to disable the pin after initial programming, these seem to be configured that way. No matter what hacking tool I use, I get no response from the pin and so cannot get any data from the MCU. I don't think the on board data is corrupted, because then I would think that none of the other functions would work. These S12s don't actually have a separate EEPROM, and instead use a part of the flash storage as a sort of virtual EEPROM. Along with this, they also have flexible addressing ranges. I think the buffer that stores the virtual EEPROM address is what is getting corrupted and so the MCU no longer reads the correct address. Hexadecimal F after all indicates essentially no data, as opposed to the random garbage you would get from corrupted data.
At this point, until someone can secure a binary for the MCU, any units with a burned out MCU or the F issue are kind of out of luck. Aside from those two issues, I've been able to repair all others.