What does it mean to interpret my ezdebug indicators?
What does it mean to interpret my ezdebug indicators?
I'm dealing with a boot failure issue. My system has operated smoothly for several months. Recently, I restarted it and encountered a persistent looping problem.
The sequence of events:
- The ezdebug CPU indicator lights up and then turns off intermittently while the power supply clicks.
- The ezdebug DRAM LED flashes similarly, briefly before shutting down.
- The ezdebug VGA LED stays illuminated for a moment before switching off at the next step.
- The ezdebug CPU indicator gradually fades while the power supply clicks again.
- The BIOS display remains inaccessible, and the cycle continues.
I consulted the user manual for my motherboard online, but the ezdebug features are described in just one sentence per LED. It appears the issue might point to a CPU fault, though I'm still uncertain.
If this is indeed a CPU-related problem, possible fixes include replacing the CPU, checking or replacing the PSU CPU cable, or swapping the motherboard.
Other attempts:
- Disconnected all storage drives from both the motherboard and power supply.
- Switched my GPU to an older model I had available.
- Used a different PSU with separate cables (though I'm unsure if the previous unit still functions).
- Removed the CPU fan, reinserted the CPU, and checked for physical damage.
- Disconnected each RAM chip.
- Motherboard model: MSI PRO B760M-P DDR4
- Power supply: Coolermaster V850
- CPU: Intel i5-13400F (possibly excessive thermal paste)
- GPU: RTX 3090
- Current GPU: A variant of Gigabyte 1060
I also changed between Linux and Windows frequently, adjusting the boot sequence often because GRUB wasn't managing it properly.
Welcome to the forums, new member!
CPU: Intel i5-13400F, possibly too much thermal paste applied, but I don't recall any overheating issues
You may wish to skip sharing the amount of thermal paste used between the processor and cooler. It would also be a good opportunity to inspect the socket for any bent or damaged pins.
How are you managing to cool the processor now?
If the debug LED remains stuck on a specific component, that component is likely the issue. If it stays on DRAM or CPU, it might point to the memory controller on your processor.
PSU: Coolermaster V850
Coolermaster offers several V series units; which one do you have? If unsure, feel free to share a link. If the unit is old, how long has it been in use and what was its previous power consumption?
Typical GPU: RTX 3090
You might consider obtaining a PSU that is robustly built with at least 1.2KW for the entire system.
The ezdebug VGA LED lights up briefly and then turns off during the next step.
This should inform the memory module that the GPU is faulty here.
Thank you for the introduction and your reply. You may wish to share a picture showing the amount of thermal paste applied between the processor and the cooler. I've already removed it, but it appears the entire surface was coated with paste. I checked the CPU temperature when I first assembled the system; it was running quite high because the stock fan wasn't seated correctly. After purchasing an aftermarket heat sink and fan, the temperatures dropped significantly. This suggests that excess paste wasn't the main problem, just for completeness.
How is the processor being cooled now?
I own a DeepCool AK400. If the debug LED remains on for a certain component, that component is likely the issue.
Would you consider it a problem if the GPU is the cause?
I find it confusing that the last indicator is the CPU light, even though the VGA light stays on briefly. It would not be unexpected if the GPU or its motherboard connector were the source of trouble, since those parts are frequently used. The most recent boot didn't involve any demanding tasks on the GPU, though I did browse the internet and transfer files to a thumb drive.
Coolermaster offers several V series units; which one do you have?
Let me check... MPY-8501-SFHAGV
80 plus gold
850W
It seems this might be too low for a 3090, possibly contributing to the problem. However, it has worked well for months with heavy AI use. I also switched to a 1060 and still faced the same issue.
The PSU was new, and I used PCPartpicker to guide my choice.
You might want to consider obtaining a PSU that reliably provides at least 1.2KW for the entire system.
I’ll have one ready by tomorrow.
Additional observations indicate the GPU fan operates normally, activating as anticipated. Deleting a memory stick triggers a notification on startup stating "CPU or memory has changed," prompting the need to press f1 or f2 for setup. Immediately after, the CPU indicator lights up and the power supply disconnects. Given the GPU-related signal on the monitor, this confirms the GPU is functioning to some extent.
I've got a 1200w PSU (RM1200x shift 80 gold plus etc.) and it's giving the same outcome. I've also changed between the 1060 and 3090 and still got the same result. This suggests a motherboard issue connected to the GPU, doesn't it?
Swapping out the motherboard appears to have resolved the issue.