Meltdown here: Hardware test says what?!

[IceMacMac]

I'm having massive problems with my MacPro...and they've been worse since I upgraded to 48Gigs of RAM, a new RAID card and RAID. Any help from any party would be greatly appreciated.

No kernel panics but increasingly frequent system freezes (requiring hard reboots)...that reached a crescendo yesterday.

So I run the Apple hardware diagnostic...the short version and it reports this:
Alert! Apple Hardware Test has detected an Error!
4MOT/1/40000002:PCI-0

How do I do decipher this? Does this meant the card in slot 4 is bad? The port? Or is it PCI slot 0? (first?)

Then I ran the deeper test and it FROZE up during the test...about 10% of the way into it. (I let it go all night but it was clearly froze up in the morning...no mouse movement, etc)

Crap. Crap. Crap.

[beto2k7]

either the pci fan or the heat sensor

[IceMacMac]

either the pci fan or the heat sensor

That would make sense given some of the crazy fan noise I've been getting lately...but then what would explain the freeze-up of the hardware test when I ran the deeper version.

[Big-TDI-Guy]

Bad temp sensor = over temp condition = freeze?

Could you have some bad memory? I had bad memory cause failures on the extended test. But with that temp sensor / fan bit - I'd look at this first.

[matspekkie]

The thing is the more you increase your memory the more you increase the changes some of it is bad. another you could try is swap pci-e cards/slots run the test again see if the error remains. and of course try first an pram zap (hold alt+cmd+p+r) at startup. this will reset hardware fan snesors etc etc.

[IceMacMac]

The thing is the more you increase your memory the more you increase the changes some of it is bad. another you could try is swap pci-e cards/slots run the test again see if the error remains. and of course try first an pram zap (hold alt+cmd+p+r) at startup. this will reset hardware fan snesors etc etc.

Good ideas, matspekkie. Tonight I'll try what you said and then give Apple Hardware Test another go.

[IceMacMac]

Bad temp sensor = over temp condition = freeze?
That would make a lot of sense because the system's fan was absolutely freaking-loud when I went to bed. So either it got too hot...or the sensor falsely reported it getting too hot.

Could you have some bad memory? I had bad memory cause failures on the extended test. But with that temp sensor / fan bit - I'd look at this first.



I wonder how much of a repair cost for sensor and/or fan (Apple Care expired).

$150?

[IceMacMac]

It's looking very much like my core problems have nothing to do with RAM and everything to do with my SATA card or the PCI port it's seated in.

Today I ran my main Data drive via Firewire and had zero problems. Massive contrast.

[bizzle]

Might want to run disk utility or DiskWarrior on that drive while you can now. I am sure the FS is a little corrupt after all those freezes and hard shut downs. Repair now before it's unrepairable.

[IceMacMac]

Might want to run disk utility or DiskWarrior on that drive while you can now. I am sure the FS is a little corrupt after all those freezes and hard shut downs. Repair now before it's unrepairable.

Yes...I've run Disk Warrior. It took five hours to repair my boot volume!

[Joshuarocks]

Yes...I've run Disk Warrior. It took five hours to repair my boot volume!

Could also be your hard drive might be dying.. Have you checked the smart on it?

[IceMacMac]

Could also be your hard drive might be dying.. Have you checked the smart on it?

Yes, checked that.

My current conclusion is that I've got two issues:
-A bad fan or thermal sensor
AND
-Something up with my RAID card or the slot its seated in

Could the two be related? I'm not smart enough to figure that out yet...

[Cindori]

Apple doesnt repair specific parts on motherboard, they just replace the whole thing.
So you are looking at ~400$ if you want to fix a thermal sensor.

However, the message you are getting, deciphers to: "PCI fan not working". Go and install iStat menus, a app that tells you your fan speeds. Check if any of the fans actually are 0 rpm.
Check by putting your head in the machine too.

It could explain why you get fan noise (mac pro compensating for one broken fan).

Check the graphics card too. PCI 0 might refer to either PCI bay fan or the fan of the PCI unit at slot 0.

If all fans are running perfectly fine, then yes, there is a thermal sensor going crazy somewhere in that area. Check the temps with istat too.


You won't die from a broken thermal sensor. You can install apps to control the fans manually instead of letting them react to the broken thermal sensor.

[IceMacMac]

Apple doesnt repair specific parts on motherboard, they just replace the whole thing.
So you are looking at ~400$ if you want to fix a thermal sensor.

However, the message you are getting, deciphers to: "PCI fan not working". Go and install iStat menus, a app that tells you your fan speeds. Check if any of the fans actually are 0 rpm.
Check by putting your head in the machine too.

It could explain why you get fan noise (mac pro compensating for one broken fan).

Check the graphics card too. PCI 0 might refer to either PCI bay fan or the fan of the PCI unit at slot 0.

If all fans are running perfectly fine, then yes, there is a thermal sensor going crazy somewhere in that area. Check the temps with istat too.


You won't die from a broken thermal sensor. You can install apps to control the fans manually instead of letting them react to the broken thermal sensor.


Massively helpful post, Cindori. Thanks!!! I'm already using a paid-for version of iStat, but I had not known/noticed you can check for fan speeds. What a great app!

I see that the Expansion Slot fan is at 0 rpm...so I guess that explains it...

The next question I'll have to decide...to fix myself...or have the Apple shop fix the fan....

[Big-TDI-Guy]

If the fan isn't soldered to the main-board (very, very rarely they are), find the part # via iFixit or equivalent - and this should be a fairly straightforward job, unplug one connector, swap 4 screws / elastomers and you're up and running again.

At the expense of being captain obvious - have you tried moving the fan by hand to see if it starts spinning? Excessive dust that can be blown out? Could be a simple as this. I can't tell you how many stalled fans just had a ball of pet hair sucked into a corner - a little air duster and they're good as new. (the fan, not the dogs)

[IceMacMac]

If the fan isn't soldered to the main-board (very, very rarely they are), find the part # via iFixit or equivalent - and this should be a fairly straightforward job, unplug one connector, swap 4 screws / elastomers and you're up and running again.

At the expense of being captain obvious - have you tried moving the fan by hand to see if it starts spinning? Excessive dust that can be blown out? Could be a simple as this. I can't tell you how many stalled fans just had a ball of pet hair sucked into a corner - a little air duster and they're good as new. (the fan, not the dogs)


Big-TDI-Guy, I am hopeful that it is a dust issue...and when I recently installed RAM I noticed I'm due for a good deep cleaning.

But if the fan isn't soldered...then perhaps it would be a simple fix, as you suggest.

[beto2k7]

Big-TDI-Guy, I am hopeful that it is a dust issue...and when I recently installed RAM I noticed I'm due for a good deep cleaning.

But if the fan isn't soldered...then perhaps it would be a simple fix, as you suggest.

I have a MacPro4,1 as well and the fan is NOT soldered just attached by two phillips screws