Smell of burning during PC startup, possible GPU or power supply issues.
Smell of burning during PC startup, possible GPU or power supply issues.
Yesterday while playing Wonderlands, my PC stopped working suddenly. This has occurred a few times before, likely related to the PSU overcurrent protection from the high power card. I noticed a burning smell of plastic, though I can't pinpoint the source. Since it's happened before, I just unplugged the PSU, plugged it back in, and turned on my PC—everything worked again. If anyone knows what caused this or how to fix it, please let me know. For context, my setup includes a 5800X, 850 Corsair RM PSU, 3080 FTW, 3 Ultra 32GB RAM, and three separate PCIe cables connected to the GPU. I checked all connections and found none burnt or damaged. Updated April 18, 2022 by lapulgaatomika
The system features a specific RAM configuration, a particular brand model, and no indications of CPU or GPU overclocking.
Avoid turning the computer on if it smells like burned plastic.
This setup doesn't function as intended. Running at a high idle clock rate only raises power usage when inactive, leaving actual performance unchanged. It's safe but not particularly practical. The power draw remains minimal—only about half of the boost level—so it won't significantly stress the PSU.
It doesn't seem like a big deal, my temps stay around 44°C with a peak of 75°C during boosts. Consistent heating and cooling help keep things stable, preventing damage to internal wires—like that red ring on Xbox 360. People manage mining rigs by maintaining steady boosts without overworking them, which extends lifespan without causing major worry. Running at full clock speed all day isn't ideal, but it doesn’t necessarily harm performance much. CPUs can also be locked to a fixed rate safely, though it uses a lot of power. My router runs an i5 7200U at 2.4Ghz nonstop and stays cool thanks to passive cooling, yet it still idles at 50°C even when pushed hard—usually only under real stress.
If you detect any burning odor, shut off the device immediately and disconnect the power source. Begin a thorough examination of all connections and parts. A burning scent strongly indicates a problem. It may result in a fire that could ruin your PC completely or cause severe damage to your home if people are inside. Pay close attention to this warning—it's serious and shouldn't be overlooked.
The difference in performance among systems largely comes down to fan settings. The cooling curve is adjusted so it doesn’t activate until temperatures exceed around 60°C on those devices, while others have a more straightforward profile. This isn’t necessarily an indicator of effort—it’s about how the hardware handles heat. Operating at 60°C is normal for silicon chips; my previous NAS/server ran at that temperature continuously and still functioned well, eventually being upgraded to something better. If you pushed it to 85°C, I’d be more worried. Games with varied workloads and constant changes will naturally strain the GPU, causing voltage, current, and clock fluctuations. As you mentioned, sudden spikes can be extreme, especially on the 30x0 series, which is a significant stress load—idle performance doesn’t stay the same even at different clock rates!
Part number: 12345
Version: 2.1
Enabled features: XMP/DOCP
Motherboard details: Intel Z690, brand Micro-ATX, model ATX-MZS-1600