Mac Pro 1,1 RAM and reboot issues

[Amethyst1]

I received a Mac Pro 1,1 (my first one ) in the mail. The machine looks to be physically in good condition, only three PCI-Express slot covers were loose when I received it and I took care of that before turning it on. It's a four-core 2.66 GHz (Xeon 5150) model with 14 GB RAM comprised of two 1 GB modules and six 2 GB modules, two hard drives and an 8800 GT.

Boot ROM version is MP11.005C.B08; SMC version is 1.7f10.

After cloning a couple of OSes (Leopard, Snow Leopard, Lion – more to come) onto the machine from external hard drives, I noticed the following behaviour in all three OSes:

1. Plugging in the power cable and turning it on results in 14 GB RAM being detected (as expected).

2. If I shut it down from the Apple menu and turn it on again, again 14 GB RAM are detected.

3a. If I reboot from the Apple menu, the machine turns off instead. When I turn it on again without cutting power in between, it shows just 12 GB RAM with System Profiler reporting four 1 GB and four 2 GB modules. Subsequent reboots result in the same behaviour.

3b. If I cut power after the machine has turned off instead of rebooting and press the power button after restoring power, it just shows a solid grey screen and never begins booting an OS.

3c. Turning it off and on again, at this point, makes it boot up again with 14 GB of RAM being detected.

4. Running Temperature Monitor 4.98 (on Leopard, but this shouldn't matter should it?), I see temperatures in the 70~80°C range reported for the memory modules themselves (ambient temp is 32°C right now). Interestingly, only DIMMs A3, A4 and B1 to B4 are listed in the application, A1 and A2 are not.

My "guts" are telling me that my memory modules are running way too hot if TM's readings are correct, possibly causing some to not fully register... The modules have heatsinks and the large fan at the back of the case is running as I can feel warm air coming out. But if this is due to heat, why is the full amount of memory detected after shutting down and booting up again, with only reboots causing these issues?

My plan is to run memtest86(+) over night next – unless there's some better way to check the memory.

Your assistance and hints will be highly appreciated.

 

[swamprock]

Pull the RAM mounting cards and clean out the slots. I would also reseat the RAM on the cards. From personal experience, those cards can be a bit finicky on old Mac Pros. I had similar issues with "missing" RAM and cleaning the cards, slots, and reseating everything fixed my issues.

 

[z970]

Pull the RAM mounting cards and clean out the slots. I would also reseat the RAM on the cards. From personal experience, those cards can be a bit finicky on old Mac Pros. I had similar issues with "missing" RAM and cleaning the cards, slots, and reseating everything fixed my issues.

As a 3,1 owner, I can confirm this. I used q-tips and rubbing alcohol on the contacts and modules, then everything worked perfectly.

Haven't touched them since. No problems to report.

 

[Amethyst1]

Thanks for your hints, will give them a try and report back.

 

[weckart]

RAM modules on the 1,1 do run very hot, especially compared with the later MacPros. 70ºC-80ºC sounds about right. It wasn't for nothing that Apple used heavy heatsinks on those sticks.

 

[bunnspecial]

As said, 1,1 RAM does run hot.

Here's a screen shot from a second ago of my 1,1 at work-it has been idle for a few hours, and I woke it up specifically to show the RAM temps.

These are all 2gb Kingston sticks with Mac Pro heat sinks. Your temperatures are very much what I'd consider normal for a 1,1, but it also is why I prefer to only use memory with the proper heat sinks.

 

[LightBulbFun]

I received a Mac Pro 1,1 (my first one ) in the mail. The machine looks to be physically in good condition, only three PCI-Express slot covers were loose when I received it and I took care of that before turning it on. It's a four-core 2.66 GHz (Xeon 5150) model with 14 GB RAM comprised of two 1 GB modules and six 2 GB modules, two hard drives and an 8800 GT.

Boot ROM version is MP11.005C.B08; SMC version is 1.7f10.

After cloning a couple of OSes (Leopard, Snow Leopard, Lion – more to come) onto the machine from external hard drives, I noticed the following behaviour in all three OSes:

1. Plugging in the power cable and turning it on results in 14 GB RAM being detected (as expected).

2. If I shut it down from the Apple menu and turn it on again, again 14 GB RAM are detected.

3a. If I reboot from the Apple menu, the machine turns off instead. When I turn it on again without cutting power in between, it shows just 12 GB RAM with System Profiler reporting four 1 GB and four 2 GB modules. Subsequent reboots result in the same behaviour.

3b. If I cut power after the machine has turned off instead of rebooting and press the power button after restoring power, it just shows a solid grey screen and never begins booting an OS.

3c. Turning it off and on again, at this point, makes it boot up again with 14 GB of RAM being detected.

4. Running Temperature Monitor 4.98 (on Leopard, but this shouldn't matter should it?), I see temperatures in the 70~80°C range reported for the memory modules themselves (ambient temp is 32°C right now). Interestingly, only DIMMs A3, A4 and B1 to B4 are listed in the application, A1 and A2 are not.

My "guts" are telling me that my memory modules are running way too hot if TM's readings are correct, possibly causing some to not fully register... The modules have heatsinks and the large fan at the back of the case is running as I can feel warm air coming out. But if this is due to heat, why is the full amount of memory detected after shutting down and booting up again, with only reboots causing these issues?

My plan is to run memtest86(+) over night next – unless there's some better way to check the memory.

Your assistance and hints will be highly appreciated.

sounds like you have some potentially bad RAM

as others say you can try give all the contacts a clean

you can also check in the Diagnostic section of System profiler when your RAM vanishes it will prolly say POST failed on memory etc.

the temps you list are fairly normal for FB-DIMMs

if you buy some more modern FB-DIMMs you generally can get away without the need for FB-DIMMs with the special Apple Heatsink

 

[Traace]

Pull the RAM mounting cards and clean out the slots. I would also reseat the RAM on the cards. From personal experience, those cards can be a bit finicky on old Mac Pros. I had similar issues with "missing" RAM and cleaning the cards, slots, and reseating everything fixed my issues.

you can also check in the Diagnostic section of System profiler when your RAM vanishes it will prolly say POST failed on memory etc.

the temps you list are fairly normal for FB-DIMMs

if you buy some more modern FB-DIMMs you generally can get away without the need for FB-DIMMs with the special Apple Heatsink

I can confirm this too

Had the same issue with my first Mac Pro ever. It was filled with dust, dianostic detected a memory error. So I cleaned the PCBs with a can of compressed air, this fixed it.

Later I switched from Apple certificated FB-DIMM to stock HP server FB-DIMM (HP PN: 398708-061). They have way smaller heatsinks but working great, still around 76°C.
With certificated rams the fans running at 800-900rpm, while with HP rams at 1100-1300rpm. However, I dont hear any difference.

 

[raymanster]

This sounds similar to issues I first had with my 1,1 when I picked it up last year. I upped the fan levels using Macs fan control and that cured my issues. Still, would be a good idea to clean contacts as others suggested.