MP 1,1-5,1 Mac Pro 4,1 (2009) - No Power On, Replaced Multiple Parts, Need Guidance

[AphoticD]

Hi all, this is a long one, so thank you in advance to anyone who takes the time to read and may be interested in helping me solve this...

I am troubleshooting a stubborn Mac Pro 4,1 (Early 2009 base-model Quad Core 2.66GHz single CPU) that refuses to power on. I bought it as a "For Parts" machine for around AU$50 (Australia), during late 2023. The seller described the machine as problematic/intermittent. At first it did not appear to respond to power button presses or shorting the power-on pads. In my initial few days with it and after some perseverance of trying SMC resets, reseating components, swapping out RAM and GPU configurations, replacing PRAM batteries, and eventually pulling the main board and PSU out to clean, and open-air test, among other various techniques, I put it all back together and did eventually see it succeed in powering on and booting twice using the front power button. So, I thought I had made progress.. I shut down to swap out drives, and then went to power it on again, and it has never powered on again.

Pressing the DIAG button while connected to power shows 5V STBY light as expected. When I initially connect power, I see a quick flash of red at the CPU lights. If I hold the front power button (or hold the short of the SYS_PWR pads) and connect power, I get sustained red lights on the 2 CPU LEDs (indicating the front power button works and/or logic board is wanting to send a POWER ON state). There are no other signs of life otherwise in terms of fan activity or other diagnostic LEDs.

Initially, without any spare parts on hand to troubleshoot, my first thought was potential power supply failure, so I hunted around for a working replacement and imported a PSU from the USA for around US$90 inc international shipping.

Once the replacement PSU arrived, I promptly installed it and observed the same behavior as before (no response to power button or pads), so I tentatively ruled out the PSU as being the culprit (I still thought maybe it was possible both PSUs were faulty), and then tried hunting for a backplane/ logic board next.

The only nationally sourced replacement I could find at a non-exorbitant price was around AU$80. The seller sold it as untested, but claimed that it once worked, and they no longer had parts to test it. I took a gamble and ordered it. The replacement backplane arrived quickly and was a match in terms of part numbers and EEE code (AFAIK). I installed the backplane and put the machine back together again, to witness the same no-power behavior: Zero progress to date. I then tested combinations of both PSUs and both backplanes to attempt to rule out one thing or another.

My next logical step was the CPU tray, so I hunted around for a little while and did not find anything of a decent price which I could consider investing in, so I shelved the project for over a year. Then last month I spotted a 6-core 2.93GHz (single CPU) tray complete with RAM, claiming to be "Pulled from a working MacPro4,1 In perfect working condition". The seller accepted my offer and I got it shipped for AU$110 with 12GB RAM pre-installed, which I considered to be not-so-bad considering how much I had invested in total so far.

The replacement CPU tray arrived and was installed, but as you may have guessed, the Mac Pro is still a no-go and does not respond to power in any manner.

I then proceeded to test each combination of Backplane, PSU and CPU on hand, with and without GPU, and inside and outside of the chassis to see if I could potentially see any sign of life. Continuing to dig deeper, I studied the PSU pinouts and bench-tested the 12V and 5V lines while in PS_ON short. Both PSUs responded with solid, reliable looking readings, so I feel quite confident the PSUs are not at fault. Leading me to attempt to record my progress to document and share here for anyone who might suggest ideas...

Long story short, I’ve invested significant time and a fair amount of money seeking out and testing parts to attempt to isolate the variables, and I’m now trying to determine next steps, and/or how I can better determine exactly where the fault lies.

Are there potentials in my case where EEE code mismatches could cause the no-power state? Or could I have multiple marginally functional or failed parts?

System Overview​

I've tested combinations of the following:

2x Logic Boards / Backplanes​

• 820-2337-A / Serial: J592900YZ1LTC / 630-9399 (original - I have witnessed it boot)
• 820-2337-A / Serial: J591401AW1LTC / 630-9399 (sold to me as untested, but appears OK)

2x CPU Trays (Processor Boards)​

• 820-2482-A / Serial: J5928021U4MFE / 630-9752 2.66GHz (original - 2.66GHz Quad Core)
• 820-2482-A / Serial: J590802104MFB / 630-9752 2.26GHz (2.93GHz 6-Core - Pulled from a working system)

2x Power Supplies (both tested working via bench test)​

• Delta DPS-980BB-1 A / Apple PN: 614-0435

All components are clean, contacts are isopropyl-alcohol cleaned, and the system is tested both in-chassis and open-air configs.

To recap what I’ve tried:​

• Confirmed both PSUs deliver 12.3V (blade) and 5.1V (pin 13) are correct on rails while using jumper wire between pin 4 and pin 14
• Confirmed front power button is functional (can be held while connecting power resulting in sustained DIAG LEDs)
• Used SYS_PWR power on pads instead of button
• Tried powering on with:
  • No RAM
  • No GPU
  • No HDD
  • 3 different known good GPUs
  • With/without front panel I/O board connected
  • With/without bluetooth and airport cards installed
  • Tried with/without PRAM battery
• Tested with both CPU trays individually on both backplanes
• Tried resetting SMC
• Tried resetting RTC
• Repeated cleaning of tray slot and tray contacts
• Inspected pins / contacts at CPU trays OK. Clean, no corrosion, no bent pins.
• Inspected all capacitors on backplanes and CPU trays - no evidence of bulging or leaking
• Tried power-on using CPU tray without CPU, RAM or heatsink
  • As per Apple service manual troubleshooting steps, this should allow the fans to spin during power on but it won't POST – this could potentially bypass EEPROM or VRM issue or incompatibility to help troubleshoot backplane + CPU tray communication and POWER ON state.

Observations​

• 5V STBY LED illuminates when power is connected and DIAG button pressed.
• Shorting front panel power-on pads does not trigger any fans, LEDs, or appears to show any voltage rail activity
• No PSU clicks/clunks when attempting to power on, which are clearly audible from the PSUs when I bench-tested shorting PS_ON and connected to power.
• I have seen the original config boot twice in the past, but never again – so surely something here has the potential to work.

---

Questions​

1. Are any of these parts potentially incompatible from an EEE code pairing perspective?
2. Is it possible that both PSUs and CPU trays are fine but both backplanes have failed?
   1. Should I take another gamble and try purchasing another backplane?
   2. From the experience of others is this the most likely issue?

---

Goal​

I'm just trying to get one known-working pairing that will reliably power on, and ultimately POST, so I can then move forward with this machine, and put it to work.

At this point, my (limited) experience with the 4,1 is making my noisy old 3,1 seem rock-solid in comparison.

Any advice, verification, or compatible EEE code lists would be hugely appreciated.

Thanks for your help!

 

[tsialex]

Questions​

1. Are any of these parts potentially incompatible from an EEE code pairing perspective?

No, you bought the right parts.

1. Is it possible that both PSUs and CPU trays are fine but both backplanes have failed?

Yes, very possible, but right now you have to assume that the defect could be elsewhere.

The best way to identify is to borrow/buy a working Mac Pro and test one part at a time on the working one.

Since you probably have an intermitent defect, more on the side that once in a while works, you probably have to test for some extended time with each part, like running ASD extended test and powering on and off (including removing mains power for like 1 hour).

1. 1. Should I take another gamble and try purchasing another backplane?

Like I wrote above, get a working 2009 Mac Pro, buying piecemeal is more expensive and is not getting you anywhere.

1. 1. From the experience of others is this the most likely issue?

People that repairs only one Mac Pro have a different experience from people that does that for a living and already have a test mule for testing all the parts.

The defect could be anywhere, not your case but for example a short circuited Superdrive made me completely tear down a Mac Pro twice and took me hours and hours of diagnose. Same thing with a leaked capacitor on the front IO board that made me crazy for an hour - sometimes is not the big items that go defective.

Btw some times even removing the shorted component, you still have to power it off from the mains for a long time until the fuses reset, sometimes takes a day.

 

[tsialex]

Also, do you know if the backplanes are already cross-flashed to MacPro5,1?

You can't run a Westmere Xeon with a backplane still with MP4,1 firmware, the Mac Pro won't power on.

This is frequently overlooked.