MP 1,1-5,1 Mac Pro freezes regularly

[ToborMP51]

I bought the Mac Pro 2010 back in 2021 and at first it worked great. However, in the last year or so, it's been constantly freezing up. Usually it would happen once every 20-30 days or so but lately, I've noticed it only lasts about 2 weeks since the last restart until it freezes up again.
At first, I was told it could be my hard drive since I mentioned it had about 60GB free. So, I replaced it with a fresh HD. But, that did not fix it.
Then, it was suggested to me that I look into reconstructing the BootROM. So, I did but, I must admit, the whole procedure looks way beyond anything I've had experience with. And, to be honest, I don't quite feel comfortable tinkering with something so sensitive.

I've been wondering if it's perhaps related to the GPU I bought and installed late 2024 to replace the stock GPU (ATI Radeon HD 5770). Though, I was assured that there should be no issues with the new GPU.

I hope someone could shed some light on what's causing this issue and if indeed I need to get my BootROM reconstructed.

OS is High Sierra 10.13.06

GPU is a Raedon Pro WX 7100

Fan Temps seem normal (between 37-48ºC/95-122ºF), though this is 2h after having rebooted the MP.
Obviously, it's hard to get a read on Temps before it freezes as it's difficult to predict what causes it or when.

There are no red lights on the MP

 

[tsialex]

Did you run AHT or ASD?

Mac Pro 5,1 powers ON for a second, but quickly shuts off.

AHT/ASD: You need an Apple OEM GPU (HD 2600XT, 8800GT, Quadro FX 5600, GT120, HD 4870/5770/5870) to run AHT/ASD. AHT/ASD can be used with PC flashed GPUs, but will fail mid test with a sensor error when the GPU sensors are tested. Non flashed PC GPUs won't work at all. You need to boot...

forums.macrumors.com

Did you measured the voltage/replaced the RTC battery?

 

[ToborMP51]

I have not run AHT or ASD. Wasn't even aware that existed. I found a link to a mac repository in the thread you pointed to. However, there seems to be three links to different versions of the ASD software for the 2010 Mac Pro:

Apple Service Diagnostics 3S149
Version: 3S149
20 December 2011
Mac Pro (Mid 2010)
ASD_ssps_004-0499-A.001.dmg
ASD_ssps_004-0499-A.002.dmg
ASD_ssps_004-0499-A.003.dmg

I suppose I can just use the latest version (003)? Or should I just go with the link to loremipsum you posted in that same thread: https://www.loremipsum.info/ia-tools-blog/ ?

Also, no, I haven't measured the voltage nor replaced the RTC battery since I received the MP. I'll be sure to do that before running the ASD/AHT. Although, as I understand it, the RTC battery basically keeps the internal clock and I've had no issues with the MP displaying the wrong time.

 

[tsialex]

I have not run AHT or ASD. Wasn't even aware that existed. I found a link to a mac repository in the thread you pointed to. However, there seems to be three links to different versions of the ASD software for the 2010 Mac Pro:

Apple Service Diagnostics 3S149
Version: 3S149
20 December 2011
Mac Pro (Mid 2010)
ASD_ssps_004-0499-A.001.dmg
ASD_ssps_004-0499-A.002.dmg
ASD_ssps_004-0499-A.003.dmg

I suppose I can just use the latest version (003)?

These are multi-part DMGs, not 3 versions, you need all 3.

Or should I just go with the link to loremipsum you posted in that same thread: https://www.loremipsum.info/ia-tools-blog/ ?

Files from loremipsum works, thats why I've linked it.

Btw, the same 3 multi-part DMGs are inside a zip there.

Also, no, I haven't measured the voltage nor replaced the RTC battery since I received the MP. I'll be sure to do that before running the ASD/AHT.

The BR2032 RTC battery is of utmost importance for a desktop Mac, if is below 2,95V in a MacPro5,1, your Mac Pro goes completely crazy - the SMC stops working correctly and all sort of weird things happen.

Below 2,75V you have bootability issues and the NVRAM stops working.

Although, as I understand it, the RTC battery basically keeps the internal clock and I've had no issues with the MP displaying the wrong time.

You are mistaken here.

The time being correct every boot is not the issue, but the several other functions of the RTC that are needed for the Darwin kernel, like timesinceboot counter. Sleep is the first thing that stops working with a spent RTC battery.

There are several high precision time related hardware counters for all sorts of needs inside the RTC and these are powered by the battery.

 

[adib]

The BR2032 RTC battery is of utmost importance for a desktop Mac, if is below 2,95V in a MacPro5,1, your Mac Pro goes completely crazy - the SMC stops working correctly and all sort of weird things happen.

I've recently replaced my Mac Mini M1 BR1632 battery with a CR1632, as I couldn't find the BR variant. It works quite well thus far. The BR variant has a longer shelf life and better tolerance for low temperatures, but I'm in the tropics, so the latter won't matter.

 

[tsialex]

I'm in the tropics, so the latter won't matter.

Matters even more since CR batteries only work up to 65ºC while BR chemistry is up to 85ºC. With Intel Mac minis CR batteries would work weeks instead of half decade for a BR.

 

[ToborMP51]

Btw, the same 3 multi-part DMGs are inside a zip there.

Ok, so please excuse my stupidity but, I'm having trouble installing ASD to my USB drive. I followed the instructions in the PDF telling me to mount the ASD_ssps_004-0499-A.001 disk image from the zip file. Then, I restored both the OS and EFI disk images stored within that ~001.dmg to the USB's OS and EFI partitions.

Where I'm stuck at is how I'm supposed to restore the DMGs within ~002.dmg and ~003.dmg without overwriting what's already been written by ~001.dmg

My concern is that when I mount ASD_ssps_004-0499-A.002.dmg and restore the OS.dmg within it to the USB's OS partition, it'll overwrite what I restored from the ~001.dmg

Also, looking at the contents of the ASD OS and ASD EFI disk images in ~002.dmg and ~003.dmg, they all seem to contain the same four pdfs:
ASD Read Me Service.pdf
ASD Test Results Guide.pdf
ASD User Guide.pdf
Installing ASD on a USB hard drive.pdf

You are mistaken here.

The time being correct every boot is not the issue, but the several other functions of the RTC that are needed for the Darwin kernel, like timesinceboot counter. Sleep is the first thing that stops working with a spent RTC battery.

Apologies for my ignorance. I did not know that the RTC was responsible for all of those functions. Thank you for clearing that up.