I built a system with a new ASUS TUF Gaming B450-PLUS II motherboard, a Ryzen 3 2200G, one XPG DDR4 module in slot A2, and a nearly new be quiet! System Power 10 650W PSU. The CPU and memory were already confirmed working in another system. With the stock cooler, integrated graphics, the 24-pin ATX cable, and the full 8-pin CPU EPS cable connected, the computer initially reached BIOS and booted Windows several times.
The system then became intermittent. On three occasions, unplugging and reseating the 24-pin connector brought it back to life, although the fix was not completely consistent. While the machine was sitting in BIOS, I pressed near the motherboard mounting and 24-pin area, and it immediately shut down. I also discovered that the motherboard had not been aligned correctly with the case standoffs, and the PCB flexed noticeably when the 24-pin connector was inserted.
After removing the board from the case, I tested it on its cardboard box with only the CPU, stock cooler, one DIMM, integrated graphics, the 24-pin cable, and the 8-pin EPS cable connected. Shorting the PWR_SW pins directly produces no response. The board still shows standby power and Aura lighting, indicating that 5VSB is present, but there is no fan movement, PSU startup, or power to SATA devices.
The symptoms make me suspect a damaged or poorly soldered 24-pin socket, a cracked motherboard trace or joint, a bad terminal in the ATX connector, or possibly a fault in the PSU or cable. What is the best electrical test to determine whether the motherboard is failing to pull PS_ON low or whether the PSU and 24-pin power path are at fault?
3 Answers
Before declaring the board dead, inspect the AM4 socket and the 24-pin connector closely with magnification. Look for bent or damaged CPU socket pins, debris, a pushed-back ATX terminal, scorching, cracked solder around the connector, or a lifted trace. Also make sure the 8-pin CPU power connector is the EPS cable rather than a PCIe cable. Since the board now fails even when breadboarded and the power switch has been bypassed, a case short alone no longer explains the behavior; an RMA or replacement motherboard is reasonable if the PSU passes its independent test.
The repeated improvement after reseating the 24-pin plug, combined with the immediate shutdown when the PCB was flexed, strongly points toward a physical fault around the motherboard's ATX connector or nearby solder joints and traces. A cracked joint or damaged contact could make and break connection as the board moves. The case alignment and excessive flex may have caused or worsened the damage, so the motherboard is the main suspect rather than the power switch.
Test the PSU separately by disconnecting it from the motherboard and briefly connecting the green PS_ON wire on the 24-pin plug to a black ground wire. If the PSU starts, then check its output voltages with a multimeter under no load or with a suitable load. This does not prove every PSU rail is healthy, but it will show whether the supply can leave standby. If it works independently while the board has 5VSB but never starts the PSU when the power pins are shorted, the motherboard may not be asserting PS_ON.
The CPU and RAM have already been reseated and tested, and the processor works in another AM4 system. Since the board is only about two weeks old, returning or exchanging it seems safer than trying to repair a suspected cracked connector or solder joint.