Good cable or bad, regardless, if the disk is an IBM or Fujitsu, the later disks had cold-start protection build in. Although the control board is powered up, when it tries to start the spindle motor it first checks the ambient, and will not start if it's out of the disk's spec.

This normally means that the device driver gets an error code back when it tries to spin up the disk (from the kernel). However, in this case, it's the boot loader trying to start the disk up and load the OS, so that's before any "intelligence" in the Linux device drivers kicks in.

What would always have been useful is if the bootloader could print up a message stating the error code returned by the drive controller, if the drive is detected as present but not functioning (in the bad cable case, the drive is not even seen as present).

What I would suggest is that you:

- install HiJack
- record the temperature at startup time by looking at HiJack's "Vital Signs" info and post it here
- post the disk type here, or go find the PDF of the data sheet for the drive
- post a serial log of the boot in the car

Then we can probably help you a bit more.