My PC has become unstable whenever any RAM stick is installed in slot A1, while the same sticks work normally in slot B1. With a stick in A1, the system may run for hours and then begin crashing, fail to recognize the memory, or crash even in the BIOS or bootloader. Waiting for a cooldown sometimes lets it work longer, but the behavior is inconsistent.
This started after transporting the PC by bus and later discovering that my cat had urinated on top of the case. I cleaned the system, but the machine has previously suffered from coastal humidity, oxidation, and PSU problems. I also accidentally tried inserting a RAM stick the wrong way and broke part of the plastic on slot A1, although the instability had already begun and the slot still latches the module.
I have tested multiple known-good RAM sticks, including matched pairs, and the problem follows slot A1 rather than a particular module. I cleaned the RAM and slots, reset the CMOS, tried disabling XMP, and underclocked the memory, but none of this made A1 reliable. Using only slot B1 is stable, but it forces single-channel operation and causes noticeable performance loss in games, screen sharing, and 3D work.
My hardware is a Ryzen 5 5500, Gigabyte A520M K V2 motherboard, 2x16GB DDR4-2133 HyperX Fury memory, Radeon RX 6600, and XPG Core Reactor 650W PSU. Should I replace the motherboard, or could the CPU's memory controller, cooler pressure, corrosion, or another issue be responsible?
4 Answers
Transport may have worsened an existing problem, but the urine and oxidation are serious clues even if no component looked visibly wet. Clean contamination properly with high-purity isopropyl alcohol, allow everything to dry completely, and inspect the A1 slot, memory contacts, and nearby motherboard traces for residue or corrosion. If A1 still fails after a bench test and the CPU socket looks good, replacing the motherboard is the most sensible repair. If a replacement board has the same issue, then investigate the CPU or its socket contact.
Before buying parts, remove the motherboard from the case and test it on a nonconductive surface with only the CPU, cooler, one known-good stick, and graphics output. Check whether the cooler is overtightened or causing board flex, and inspect the CPU socket and pins under strong light for bent, dirty, or corroded contacts. On this platform, a bad CPU socket contact can disable or destabilize one memory channel, so the CPU cannot be ruled out completely.
Because multiple known-good sticks work in B1 but fail in A1, the RAM modules themselves are unlikely to be the cause. The fault is probably in the A1 slot, its solder joints or traces, the CPU socket/contact pins, or the CPU's memory channel. The damaged slot and possible exposure to urine, moisture, and coastal corrosion make the motherboard especially suspect. Do not keep powering it if liquid residue or corrosion may remain, since it can continue eating away at contacts and traces.
Run a memory test with one stick at a time in B1 and then A1, but do not rely only on tests that boot into the operating system. The fact that it can crash in BIOS points to hardware rather than Linux or drivers. Also avoid mixing unmatched kits and leave XMP disabled while troubleshooting. A stable single stick in B1 is useful as a temporary workaround, but it will not repair the failed memory channel.
That makes sense. The behavior follows A1 with every tested stick, so I am treating the motherboard and CPU socket as the main suspects rather than buying more RAM.