Custom PC issues at one month old – need some guidance on fixing it
Custom PC issues at one month old – need some guidance on fixing it
It seems my new motherboard didn’t work at all. I obtained another one and everything is functioning correctly. It appears the first board I got was really faulty. THE ORIGINAL POST: I recently assembled a fresh PC, and it operated smoothly for a week. Then last week, while playing Overwatch—a game that doesn’t heavily tax CPU or GPU—I experienced a sudden black screen. Although fans were still running and my motherboard LEDs displayed an error code 00, indicating a potential CPU or board issue. My top heatsink felt unusually hot, yet I had Corsair iCUE configured to alert me when CPU temperature exceeded 80°C, automatically shutting down at 90°C. I read that the Ryzen 5900X can reach up to 95°C and should be safe under those conditions. Although it was quite warm, it shouldn’t have overheated or caused damage. After the initial failure, I swapped out the motherboard for a new identical model and experimented with a different power supply to check for failures. Now I’m facing even more problems. Once I filled up my custom loop and connected all PSU cables, I attempted to boot but the system remained unresponsive. At least before it failed to power on, it would have shown a POST. Still, the motherboard LED code 00 persists. Since then, I’ve replaced the board again, tried a fresh PSU, but the issue continues. Using a jumper cable, fans spin normally and most LEDs illuminate, though side fans’ LEDs stay off despite being on the same core as the top fans. I also noticed RAM LEDs are not lighting up at all, even though they used to. It seems the board still receives power when the PSU is active, but pressing the power button doesn’t work. I can’t even activate the piezo speaker on the motherboard. Replacing the motherboard didn’t help and I’ve already returned it. My best guess now is that the CPU might be faulty or damaged, possibly due to ESD. I’m seeking advice on possible causes and steps to resolve this. I’ve attached my parts list with photos of the new setup under these conditions, plus a picture showing the system still won’t power on. AMD Ryzen 9 5900X 3.7 GHz 12-Core Processor MSI MEG X570 UNIFY ATX AM4 Motherboard Corsair Dominator Platinum RGB 32 GB (4 x 8 GB) DDR4-3600 CL16 Memory Sabrent Rocket 4 Plus 4 TB M.2-2280 NVME SSD Zotac GeForce RTX 3080 10GB 10 GB GAMING Trinity OC Video Card Lian Li PC-O11 Dynamic Razer ATX Full Tower Case EVGA P2 1000 W 80+ Platinum Certified Fully Modular ATX Power Supply 3x Corsair iCUE QL120 41.8 CFM 120 mm Fans 3-Pack 2 corsair commander pros Corsair Hydro X products for custom loop CableMod E-Series Pro ModFlex Sleeved Cable Kit for EVGA G5 / G3 / G2 / P2 / T2 (Black + Silver).
The 5600X operates safely up to 95°C. 5800X, 5900X, and 5950X remain safe at 90°C under full load. iCUE provides its own temperature readings which may vary slightly from actual conditions—this has happened on my air-cooled 5900X. Verify your RAM placement; I lost mine after moving it during cable installation and experienced similar issues three hours later. Liquid cooling isn't essential; my 5900X reaches a peak of 75°C without PBO and 86°C with PBO using a Scythe Mugen 5 and six additional 120s. I haven't liked MSI much since the B450M Bazooka caused a shutdown during a GPU upgrade, requiring an RMA. They seemed uncertain about supporting the 5900X at that time, and their website later changed its stance. Not advisable for my biggest rival. Updated April 29, 2022 by An0maly_76 Added more details
It’s really tough to fix a custom loop this way. I’d disassemble the whole system and test it on the motherboard box. This helps pinpoint the issue. Skip the riser, pumps, or custom PSU cables—just focus on the core problem. Replacing the board before testing outside the loop doesn’t make sense; it’s better to isolate everything first.
It's not the issue itself, but a strict loop prevents you from resolving it completely. If you can't share or figure it out, you should remove it.
I lack any extra parts or a dedicated workspace. That means I can't fully verify the system without assembling it all myself. Are you asking about cooling the CPU and checking components without a dedicated cooler besides your waterblock?
Ok I still don't think my CPU got past the 70s unless my sensors were displaying the wrong data. Also I checked the RAM and it appears to be secure. I can't tell if it's the complete failure is in the mobo or not now. It's a brand new one and Idk if they'll send me a new one since this is already the replacement.
The setup seemed stable too. My concern disappeared after I released the latch, pulled it out, and repositioned it. The VRM cooling looks decent but not great in the photos. A 5900X on a budget board without VRM support isn’t ideal. Chipset temperatures matter a lot. If this board has problems, it’s a good chance to replace it with something better, since I’ve seen MSI often have issues and their support is unreliable. I used an Asus Tuf B550 Plus; the Pro model works even better. You might want to think about switching to air cooling instead. Liquid systems can give uneven temperature readings and are prone to pump failures. Leaks through porous hoses are common, and cheap connections can cause this too. Many people prefer liquid cooling, but it’s not essential and can lead to frequent troubleshooting. Every day, at least a few folks discuss problems with liquid cooling, some boards have recurring recalls. Linus compared them and found liquid cooling only slightly better (but more stable) than air cooling when properly configured. He mentioned liquid cooling is fine if you enjoy customizing your rig, but it often causes downtime and extra headaches. Most people criticize air cooling, yet it’s just as effective and dependable if you set it up correctly. It’s not complicated—even I documented my experience. Updated April 29, 2022 by An0maly_76 Added more details