Possibly your power source got damaged from overheating.
Possibly your power source got damaged from overheating.
My setup worked perfectly until I connected one more Molex cable to the SATA power adapter. Now my PC won’t turn on at all. Summary: Ryzen 5 3600, RX 2070, Super 2 M.2 NVMe, 4 SSDs, 2 HDDs, 4 fans, 4 DDR4 RAM, 500W quiet D tier gold PSU, non-modular, silent operation, no noise, no smoke, no odor. I removed everything I’d added and restarted, same outcome. After ten minutes of letting the PSU sit, it remained completely unresponsive. Could this be because the extra power draw caused it to fail? Without the final SSD and HDD, the drive setup worked for two days. When I reconnected the power cable into the PSU, I heard some unpleasant buzzing—something you might encounter if it wasn’t fully seated. Any ideas or advice? Recommend a high-end modular PSU rated at least 750W to handle all your components. Also, regarding power consumption: if your system required 550W, it would likely draw close to that value rather than the full 750W unless it was overloaded. In my experience, systems with heavy loads often draw near their maximum output instead of settling around a lower figure, especially when measured with a proper power meter.
a psu only supplies the power required; for example, a 1500 watt unit connected to a 400 watt device would draw just 400 watts plus losses and conversion overhead (assuming 110-120v with 20% loss). this applies to the rtx series which can experience power spikes that may damage the psu if you exceed its limits. choosing a psu close to its maximum capacity improves efficiency because using less of its maximum capacity yields better conversion performance.
Thank you all. I realize the sudden change disrupted my thoughts. I verified everything carefully and decided to manually connect the power pins just in case the power button malfunctioned unexpectedly. By handling all the SATA cables, I accidentally dislodged the PWR_SW jumper. Uh, that’s a bit embarrassing... Please accept my humiliation. On a positive note, I’m relieved the problem is fixed now. After all, @lafrente raises a valid concern. Could the system operate in idle mode but behave differently under CPU and GPU stress? Would a lightweight stress-test program help identify issues with either component? I’m only thinking of simple software that could challenge the GPU or CPU slightly.
It's straightforward and often overlooked, which is nice to know.