[weblacky]

Okay, clearly this is a very unusual situation. And I’m going to just put this out there even though I have no basis for this and that is I think you have a power or VRM problem.

In my mind unless you have a cracked solder joint or lifted trace or damaged board, that is by scratches from improper handling things like that, that local power problems would leave ICs unresponsive or otherwise not operational.

Now I fully realize that a lot of the Qlogic stuff is BGA so it’s not easy to just find the right pin put a multimeter on it and say yes I’m getting correct VCC on this chip.

But since you’ve gone through multiple cards I think you have a power issue either on mainboard or on socket.

Here’s what I’d like you to try. Before you start swapping things can you get to an L1 prompt to issue the ENV command?

Because you’re, maybe, not booting far enough to know if the ENV system is on or off or working at all when it comes to the monitoring.

Let’s start with what your monitoring says for everything. If you’ve already posted this information then I just don’t remember seeing it. If you could re-post it for me, let’s start there.

I haven’t taken a good look at an I09 but mine is in fact removed right now so I can take glance at one easy enough and see if there any voltage conversion circuits on it at all that convert Power from the slot to power for the IO9 itself. If I don’t see such a thing then I’m going to assume that power comes from the slot and it’s directly used by the ICs on the IO9.

I’m going to propose that you have a missing or low voltage that is being sent to your SCSI controller. And basically the chip is simply not turning on and that’s why it’s gone. I would hope opposite isn’t true. That it’s being sent too much power and is actually damaging it.

But either way this is where I would start.

Let’s also see if we can find the Qlogic chip on the IO9 and find its data sheet. That’ll tell us exactly what voltage it’s taking and potentially what VRM were using. I do know that some of the removable or semi removal VRM‘s are used to produce the lower voltages. But I’d normally believe that the Qlogic chip would use the normal power rails on the power supply.

Why don’t we proceed under this assumption and see if we can find ourselves an errant or missing voltage that might explain loss of power on part of the IO9 board?