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

Liger Megatron

macrumors newbie
Original poster
Hello there,

I'm reaching out to this community seeking some good advice and recommendations for restoring my 2012 Mac Pro to working condition. I've pulled it apart with the exception of removing the back-plane board and have given it a thorough clean.

Have also removed the heat sink from the processor board and cleaned out all the dust therein too.

Essentially - it won't power on. At first I thought it was a dead/dying PSU. But I've replaced it recently and there's been no change. When I press the diagnostic button on the board the 3rd or maybe the 4th? LED lights up amber. When I remove the power cable and plug it back in the first 2 led's briefly flash red.

I've also tried starting it without HDD's, Optical drives, video card, processor board + RAM to try and identify the cause. Have also replaced the CR2032 battery too.

Is it possible for a component to be dying - rather than dead? I kinda feel thats the case here - like something isn't quite dead and still is trying to work.....

I'm also coming to the conclusion that its possibly the logic/back-plane board....

But guesswork is expensive - as I found from just assuming it was the PSU. Hence my message to this community...

I know back-plane boards can still be purchased readily in tested and untested condition. Gotta feeling no one actually stocks a brand new board anywhere anymore though? And part of my dilemma too is freight to Australia.

Hoping someone here can offer some wisdom and good counsel please?

Kind regards to you all,
Liger Megatron
 
Last edited:
Get another working Mac Pro, will be cheaper than buying piecemeal, to test everything one at a time and buy exactly what is defective, doesn't make sense to spend with guesswork.

You can even buy an early-2009 for the diagnostics, but is better to stick to mid-2010/mid-2012.

This sounds insane to read, but to test each part with a working Mac Pro is what even the most experienced people do even when we have all the diagnostic tools that an end-user doesn't have access, like MATT cards and loggers. The Mac Pro is too complex and sometimes the defect is intermittent and difficult to track, especially with the CPU tray and PSU, so you really need to test overtime.
 
  • Like
Reactions: ddhhddhh2
Before doing anything that involves spending money, check the USB ports for damage or debris. I had a Mac Pro 2010 to fix that would not start and I found bits of some small USB device (maybe Bluetooth) in one of the USB ports. I removed the bits, checked that the contacts were not shorting and the Mac Pro booted straight up. Prior to that, when the power button was pressed, the fans would just barely move before the PSU detected the short circuit and shut off.

Maybe a damaged mouse or keyboard cable would do the same, try without them.

Finally, get a copy of the official Apple Technician guide and go through the fault finding section, details here:

If it is the logic board, it will likely be cheaper to buy a new Mac Pro 2010/2012 and strip the old Mac pro for usable parts to sell on ebay or similar, or keep for spares.
 
Thank you to those who've responded to my query. I'm currently weighing up all options on what would be best for me moving forward. As @tsialex has said, the problem seems to be intermittent. I bought this Mac brand new back in 2013. In the years that have followed its been in to Apple and had the PSU and logic/backplane board replaced.

Since my original post, I'm tried starting it with and without the processor board in. I still have the original RAM sticks and video card it came with out of the box.

I've really given it a good clean/dusting internally. Have also blown out and dust in the USB and Firewire ports. I've read somewhere the power button can sometimes be the cause. Is it possible to take that apart to clean if you suspect its sticking?

For as long as I can remember, one of the processor tray levers is really, really tight to loosen. It has nothing to do with the lever itself but one of the 5 pins directly underneath.

I've swapped them around but it makes no difference. I don't believe this is related to my current problem. And there's no sign of bent pins or ports on either the processor board or the back plane board. Just a nuisance.
 
Before you drop money on a backplane — on a 4,1/5,1 the specific LED pattern you're describing (3rd/4th amber, first two flashing red on cable insert) tends to point at the SMC losing its handshake with the CPU tray rather than the whole board being dead. That brief red flash on plug-in is the SMC trying to come alive on the 12V standby rail and giving up when it can't validate something downstream — intermittent because the standby regulator on the backplane is marginal, not open-circuit.

Two cheap tests worth doing before ordering parts:

Each memory riser in isolation, one DIMM in slot A only, then watch whether the amber LED pattern shifts — a marginal riser regulator will move which LED lights up. Then a standby-rail check with the CPU tray pulled: DMM across the 12V standby test points on the backplane (top-right area near the CPU tray connector), should hold a solid 12.0V. If it's sitting under 11.4V or bouncing, that's the backplane's own standby regulator or the tantalum caps around it, which are the usual failure mode at this age and a $30 recap job rather than a full board.

tsialex is right that a working donor is the ultimate diagnostic rig, but on a 2012 you're often chasing capacitor aging on the standby rail before anything more exotic. Freight to Australia for a whole backplane hurts enough that it's worth an hour with a multimeter first.
 
  • Like
Reactions: Borowski
Before you drop money on a backplane — on a 4,1/5,1 the specific LED pattern you're describing (3rd/4th amber, first two flashing red on cable insert) tends to point at the SMC losing its handshake with the CPU tray rather than the whole board being dead. That brief red flash on plug-in is the SMC trying to come alive on the 12V standby rail and giving up when it can't validate something downstream — intermittent because the standby regulator on the backplane is marginal, not open-circuit.

Two cheap tests worth doing before ordering parts:

Each memory riser in isolation, one DIMM in slot A only, then watch whether the amber LED pattern shifts — a marginal riser regulator will move which LED lights up. Then a standby-rail check with the CPU tray pulled: DMM across the 12V standby test points on the backplane (top-right area near the CPU tray connector), should hold a solid 12.0V. If it's sitting under 11.4V or bouncing, that's the backplane's own standby regulator or the tantalum caps around it, which are the usual failure mode at this age and a $30 recap job rather than a full board.

tsialex is right that a working donor is the ultimate diagnostic rig, but on a 2012 you're often chasing capacitor aging on the standby rail before anything more exotic. Freight to Australia for a whole backplane hurts enough that it's worth an hour with a multimeter first.
I really do appreciate the time people have taken to respond and offer advice. But I'm an amateur at some of the terminology and possible remedies for a partial backplane board fix. Just some other info that comes to mind:

The top 2 LED's flash red briefly when power is applied. I've tested this with and without the processor board inserted. The 3rd or 4th amber light only flashes when I press the small diagnostic button. Otherwise nothing seems to be active or working

So irrespective of whether the processor tray is in or not - the only visible activity is the top 2 red LED's briefly.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.