Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

MacInquiry

macrumors newbie
Original poster
Aug 25, 2023
5
0
I recently got an amazing deal on a 2009 MacPro flashed to 5,1 firmware from a professional music producer who upgraded to a more modern Mac and just wanted to sell his older one fast. The body of this MacPro is literally in near mint condition and even the inside of it only had a moderate to minimal amount of dust and it even came with the original Mac Pro box--this guy knew how to take care of his gear to say the least!

I tested all the ports on it were working ok, ran Geekbench 5 test to see that it ran it without crashing or other issues and finally ran the EXTREMELY LONG 3+ hour EXTENDED Apple Hardware Test DVD boot disc and after 3.5 hours testing the machine it reported back that there were NO ERRORS or problems found with the system or the memory.

I even downloaded trial software of Adobe Photoshop 2023 and Topaz Photo AI to see this Mac Pro run them because I read in a YouTube thread on upgrading these 5,1 MacPros that one commenter essentially said that his MacPro 5,1 after upgraded to Monterey with OCPL could no longer run PhotoShop 2023 or Topaz Photo AI at normal speeds as this software was just crawling along at such a oddly slow rate that this software was nearly unusable. Both Photoshop 2023 and Topaz Photo AI ran at good speeds in this MacPro with 64GB RAM while working on multiple very high res files.

Yet I have this very odd problem with this 5,1 Mac Pro in that every so often, and more recently more often than before, when I push the power button to start the machine up I get a flashing power button light and it doesn't start. The thing is though that it always starts up again fine after I turn the machine off with the power button, wait 5 seconds or so then press the power button again. It's not the end of the world to have to deal with this annoying problem but it is somewhat annoying and a waste of time but I'm also worried that it might be an indicator of a greater problem lurking in the system. This problem occured/occurs both when I was running this machine on Mojave from an HDD (the HD set up this machine came with) and on Monterey via OCLP on a Sabrent nVMe PCIe card with a Samsung 2TB nVMe SSD.

The other odd problem I have with it is the GPU a Radeon RX580 which has very clean fans and otherwise looks brand new at times makes a loud clicking sound at startup that then fades away after I log in to the machine. Sometimes however this loud clicking sound it makes doesn't fade away after start up and this is just unbearable to have to listen to while using this machine. It has only happened on 2 occasions that this loud sound continued on this machine but this is the second odd problem on this machine.

Machine specs:
2009 Mac Pro flashed to 5,1
12-Core 3.46 GHz
64 GB of 1333 MHz RAM
Samsung 2TB nVMe SSD on a Sabrent nVMe PCIe card in the 16x PCI card slot
Radeon RX580 8GB

Since I know from reading other threads here that a flashing start up light is a sign that there is something wrong with the RAM in the MacPro I checked the memory LED lights on the CPU tray and they aren't lit up. I also tried putting in other known working RAM from from my other MacPro that has never had this flashing start up light issue and the intermittent flashing startup light problem continued so I just put the original RAM back in. Also note that I ran the EXTENDED extra long 3+ hour system hardware and memory test on this machine and the hardware and RAM passed 100% ok. Also the machine starts up fine with just one attempt after these failed start up attempts with the flashing light and when its running I get no errors or problems of any kind except for the loud GPU but that issue is very rare and seems to be some kind of problem with the physical fans on the GPU even though they look very clean.

Anyone have any idea of what to try next to solve these intermittent problems?
 

tsialex

Contributor
Jun 13, 2016
13,457
13,607
Start to diagnose each component with your sure working Mac Pro.

I'd start checking the RTC battery voltage, then installing the PSU of the sure working Mac Pro to the problematic one and vice-versa.

It's not easy to diagnose intermittent problems, but with patience you can track it down.
 
  • Like
Reactions: MacInquiry

tsialex

Contributor
Jun 13, 2016
13,457
13,607
Also, when the Mac Pro does not start, which DIAG LEDs are not working when you press the DIAG button?
 
  • Like
Reactions: MacInquiry

Macschrauber

macrumors 68030
Dec 27, 2015
2,985
1,494
Germany
Can you see how the CPUs were mounted?

When looking from the sides under the heat sinks you will see if there is the PCB of the processors or you see a ca. 2mm thick metal plate on the top of the CPUs.

If you see that lid the processors were mounted the wrong way. Maybe there is the wrong amount of pressure.

Also a clear sign of a CPU upgrade with the lid not taken off is when there are two layers of thermal pads between the heat sink and the voltage regulator chips.

The machine had could work perfect until the mechanical stress of transportation.

As Alex told, cross test the components before, the fans will run wild if a 4,1 CPU Board is built in a 5,1 but that works in both directions (as the 4,1 is crossflashed).

Also check the temperatures with MacsFanControl or another tool. The Northbridge cooler pins tend to break.
 
  • Like
Reactions: MacInquiry

MacInquiry

macrumors newbie
Original poster
Aug 25, 2023
5
0
Also, when the Mac Pro does not start, which DIAG LEDs are not working when you press the DIAG button?
Thanks for your thoughtful reply! Where is the "DIAG button" on the backplane board or the CPU tray board? Is there a MacRummors thread you can point me to with a pic/diagram and the meanings of what the lights mean when on or off?
 

MacInquiry

macrumors newbie
Original poster
Aug 25, 2023
5
0
Start to diagnose each component with your sure working Mac Pro.

I'd start checking the RTC battery voltage, then installing the PSU of the sure working Mac Pro to the problematic one and vice-versa.

It's not easy to diagnose intermittent problems, but with patience you can track it down.
Thanks for your thoughtful reply! Can you point me to a thread here on MacRummors or an external site that describes how to check the RTC voltage and why a user would want to do this?
 

tsialex

Contributor
Jun 13, 2016
13,457
13,607
Thanks for your thoughtful reply! Where is the "DIAG button" on the backplane board or the CPU tray board? Is there a MacRummors thread you can point me to with a pic/diagram and the meanings of what the lights mean when on or off?

Apple Technician Guide, pages 27 to 36.


Thanks for your thoughtful reply! Can you point me to a thread here on MacRummors or an external site that describes how to check the RTC voltage and why a user would want to do this?

RTC battery keeps the Real Time Clock and several time related counters that if not working correctly (anything below 2.8V) cause several issues and kernel panics.

If the BR2032 battery is spent - nominal voltage is 3.0V, but from factory is usually 3.2V - your Mac Pro goes crazy.

You can check the BR2032 battery with a voltmeter set to DC voltage. Remove it from the socket with care to not damage the socket, the top of the battery is +, bottom is -.

Replace it with another BR2032 if it's below 3.0V. Do not install a CR2032, different chemistry and not capable to resist the high temperatures below the GPU heatsink.
 

tsialex

Contributor
Jun 13, 2016
13,457
13,607
There are two banks of LEDs, the first bank LEDs are:

  • OVTMP CPUA (RED)
  • OVTMP CPUB (RED)
  • PSU PWROK (AMBER)
  • 5V STBY (AMBER)
  • 3.3V (AMBER)
  • 1.5V (AMBER)
  • 1.05V (AMBER)

The second bank LEDs (not all PCBs pads have LEDs installed, just the three below) are:
  • SYS_PG, (AMBER)
  • EFI_DONE (GREEN)
  • GPU_OK (GREEN)

If you don't have EFI_DONE lit, you can't boot at all, but even if EFI_DONE LED is lit, the BootROM could be corrupted in a way that you can't boot.
 

MacInquiry

macrumors newbie
Original poster
Aug 25, 2023
5
0
Can you see how the CPUs were mounted?

When looking from the sides under the heat sinks you will see if there is the PCB of the processors or you see a ca. 2mm thick metal plate on the top of the CPUs.

If you see that lid the processors were mounted the wrong way. Maybe there is the wrong amount of pressure.

Also a clear sign of a CPU upgrade with the lid not taken off is when there are two layers of thermal pads between the heat sink and the voltage regulator chips.

The machine had could work perfect until the mechanical stress of transportation.

As Alex told, cross test the components before, the fans will run wild if a 4,1 CPU Board is built in a 5,1 but that works in both directions (as the 4,1 is crossflashed).

Also check the temperatures with MacsFanControl or another tool. The Northbridge cooler pins tend to break.
Thanks for your thoughtful reply! I tend to think that the upgraded CPUs were installed correctly in this machine given that the guy who owned it before me was so meticulous in how he took care of this machine and he used it for professional work but I'll put this on my list of things to check.

One thing you said dovetailed with something I was thinking might be the cause of the problem:

You said: "The machine had could work perfect until the mechanical stress of transportation."

This MacPro came with the original Apple box that it shipped with including the foam form fitting pads and it was transported about 40 miles from the original owners home to mine in my car in a secure spot in the trunk of the car BUT it was a pretty hot day that day not crazy hot but like maybe 88 degrees out and it was getting very hot in my car as I drove back plus I made 2 stops on the way home plus there was a lot of traffic driving back home so all in all this MacPro was sitting in my car in very hot (but not extremely hot) tempertures for like maybe 4 hours. I didn't start it up until the next morning so it had plenty of time to cool off however but the first start up of it when I got it home gave me the flashing start up light but like I said before the next start up attempt went fine and it started right up and functioned perfectly for hours and hours all day long once it starts up fine which always happens 1 or 2 attempts after any failed attemps at start up with the flashing start up light.

Could being exposed to this heat have caused the problem? If so what would that source of the problem point to in terms of failing components or settings in the machine?
 

Macschrauber

macrumors 68030
Dec 27, 2015
2,985
1,494
Germany
Well, testing the cpu tray in another Mac Pro would give the most accurate answer.

I dont think the thermal stress and transporting in the original box in a car would damage something.

However, if the cpu upgrade was done with the lids left on it is very delicate. If you can nail it down to the CPU tray (and the ram sticks are not faulty) I would check that.

I had a lot of 4,1 in my hands with bad cpu upgrades so I made my experiences. The memory controllers are in the cpus and all sort of memory faults can happen with bad contacting.
 
  • Like
Reactions: MacInquiry

MacInquiry

macrumors newbie
Original poster
Aug 25, 2023
5
0
Well, testing the cpu tray in another Mac Pro would give the most accurate answer.

I dont think the thermal stress and transporting in the original box in a car would damage something.

However, if the cpu upgrade was done with the lids left on it is very delicate. If you can nail it down to the CPU tray (and the ram sticks are not faulty) I would check that.

I had a lot of 4,1 in my hands with bad cpu upgrades so I made my experiences. The memory controllers are in the cpus and all sort of memory faults can happen with bad contacting.
I put the 12-Core CPU tray from the MacPro with the intermittent flashing start up light into my other one that never had any flashing startup light problems and sure enough on the first start up my MacPro that previously had no flashing start up light problem all of a sudden DID have the flashing start up light problem.

So it seems the issue IS with the CPU tray and maybe as you suggested an issue with the mounting of the upgraded 3.46 GHz CPU chips. I didn't get anything specific to look for from you in terms of mistakes in how the upgraded CPU chips are installed. So if the upgraded chips are lidded from what I understand there needs to be a secondary additional thermal pad between the CPU heatsink and the temperature sensors and the CPU fan connector needs to be pulled down a little to be in a position to connect to the female CPU fan connector input. So I imagine that these are the aspects of the install that I should be looking for to make sure they were done correctly? That is if the upgraded CPUs are in fact lidded. If the upgraded CPUs turn out to be delidded what should I be looking for in terms of non functional install attributes?

Keep in mind this 12-Core MacPro and CPU tray went through the extensive extra long Apple Hardware Test from the original DVDs that came with the Mac and 100% passed this 3+ hour test of the entire system and memory. So it seems to me if there was some kind of issue with how the upgraded CPU chips were installed some kind of error would have shown up via this test but maybe I could be wrong about that I don't know...

I'm ccing Alex on this message too.

Many many thanks for your advise...

@tsialex
 

tsialex

Contributor
Jun 13, 2016
13,457
13,607
I put the 12-Core CPU tray from the MacPro with the intermittent flashing start up light into my other one that never had any flashing startup light problems and sure enough on the first start up my MacPro that previously had no flashing start up light problem all of a sudden DID have the flashing start up light problem.

So it seems the issue IS with the CPU tray and maybe as you suggested an issue with the mounting of the upgraded 3.46 GHz CPU chips. I didn't get anything specific to look for from you in terms of mistakes in how the upgraded CPU chips are installed. So if the upgraded chips are lidded from what I understand there needs to be a secondary additional thermal pad between the CPU heatsink and the temperature sensors and the CPU fan connector needs to be pulled down a little to be in a position to connect to the female CPU fan connector input. So I imagine that these are the aspects of the install that I should be looking for to make sure they were done correctly? That is if the upgraded CPUs are in fact lidded. If the upgraded CPUs turn out to be delidded what should I be looking for in terms of non functional install attributes?

Keep in mind this 12-Core MacPro and CPU tray went through the extensive extra long Apple Hardware Test from the original DVDs that came with the Mac and 100% passed this 3+ hour test of the entire system and memory. So it seems to me if there was some kind of issue with how the upgraded CPU chips were installed some kind of error would have shown up via this test but maybe I could be wrong about that I don't know...

I'm ccing Alex on this message too.

Many many thanks for your advise...

@tsialex

Inspect the Xeons, if the processors are still lided, change to de-lidded ones or de-lid yourself.

It´s a waste of time to try to debug lided Xeons with an early-2009 dual CPU tray nowadays, the most probable outcome is that you will damage the sockets.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.