Your PC just died. One minute you're mid-raid or halfway through a spreadsheet, and the next, total blackness. No lights. No fans. Just the cold, silent judgment of a dead monitor. Naturally, your mind goes straight to the most expensive part—the GPU—but more often than not, the culprit is that heavy metal box tucked in the corner of your case. How can I check my PSU without accidentally frying my motherboard or, you know, myself? It's a question that feels daunting because power supplies are essentially the "black boxes" of the computer world. They take high-voltage AC from your wall and turn it into the delicate DC sips your components need. If that process breaks, everything stops.
Honestly, testing a PSU isn't as scary as people make it out to be. You don't need a degree in electrical engineering, though a little healthy respect for electricity is mandatory. You’ve basically got three levels of testing: the "is it even alive" test, the "is it actually working correctly" test, and the "is it dying under pressure" test. We're going to walk through all of them.
The Paperclip Test: A Quick Reality Check
Before you go buying a new unit, you need to know if the thing can even turn itself on. Most modern power supplies won't spin their fans unless they get a specific signal from the motherboard. You can spoof this signal. It's called the paperclip test.
First, flip the switch on the back of the PSU to "O" (off) and unplug it from the wall. You need to disconnect every single cable from your components. Every. Single. One. If you leave the CPU or GPU plugged in while doing this, you're asking for a bad time. Locate the massive 24-pin connector—the one that usually plugs into the right side of your motherboard. Look for the green wire. There’s usually only one. If your cables are all black (common in high-end units like the Corsair RM series), you're looking for pins 16 and 17.
Wait. Don't just start sticking metal into holes.
💡 You might also like: Is Google listening to you? Why your phone seems to read your mind
Take a standard paperclip and bend it into a U-shape. Insert one end into the pin for the green wire (PS_ON) and the other end into any pin with a black wire (Ground). Plug the PSU back into the wall and flip the switch to "I". Does the fan spin? If yes, the PSU is technically "alive." If no, it’s likely e-waste. Keep in mind, some "Eco Mode" power supplies won't spin the fan unless they're hot, so check if your model has a fan-less mode before you panic.
Using a Digital Multimeter for Real Accuracy
The paperclip test is crude. It tells you if the heart is beating, but it doesn't tell you if the blood pressure is right. For that, you need a multimeter. If you're wondering how can I check my PSU with actual precision, this is the gold standard. A PSU might turn on but only output 10V on a 12V rail, which is enough to cause constant blue screens or "random" restarts that drive you insane.
Set your multimeter to DC voltage. With the paperclip jump still in place (so the PSU is running), stick your black probe into any ground pin (black wire) and your red probe into the yellow wire pins. These should read as close to 12.0V as possible.
The industry standard, defined by the ATX specification, allows for a 5% margin of error.
- 12V Rail: Should be between 11.4V and 12.6V.
- 5V Rail: Should be between 4.75V and 5.25V.
- 3.3V Rail: Should be between 3.14V and 3.47V.
If your 12V rail is sitting at 11.2V, your PC might boot, but the second your GPU pulls power, that voltage will sag even further, and your system will crash. I’ve seen cheap "no-name" power supplies show 12.5V at idle but drop to 11.0V the moment a game starts. That’s a recipe for hardware degradation.
The Dedicated PSU Tester Route
Not everyone wants to poke around with probes. If you’re a frequent builder, buy a dedicated digital PSU tester. They cost about $15 to $20. You plug the 24-pin, the 8-pin CPU, and the PCIe cables into the device, and it gives you a backlit LCD readout of every voltage.
The most important value on these testers isn't the voltage, though—it's the PG (Power Good) value. This is a delay measured in milliseconds (ms). It represents the time it takes for the PSU to stabilize its voltages before sending a signal to the motherboard that it's safe to boot. A healthy PG value is typically between 100ms and 500ms. If your tester shows "HH" or "LL" for the PG signal, or if it beeps like a smoke detector, your PSU is failing the internal timing check. Toss it.
Software Monitoring: Seeing the PSU Through the OS
Sometimes the PSU is fine at idle but chokes under load. Since you can't exactly stick a multimeter into your PC while you're playing Cyberpunk 2077, you have to rely on software sensors. Programs like HWInfo64 or OCCT are brilliant for this.
Open HWInfo64 and look for the "Sensors" tab. Scroll down to your motherboard section and find the +12V, +5V, and +3.3V readings. Now, run a stress test like Prime95 or FurMark. Watch those numbers. If you see the 12V rail dipping below 11.5V while your CPU is working hard, your PSU is likely underpowered or aging poorly. Capacitors inside a PSU dry out over time—a process called "capacitor aging"—which reduces their ability to filter noise and maintain steady voltage. If your unit is seven years old and starts sagging, it’s earned its retirement.
Why "Checking" Isn't Always Enough
You have to realize that a PSU can pass every single one of these tests and still be the reason your computer is acting weird. This usually happens because of transient spikes. High-end GPUs, especially the RTX 30 and 40 series, have these tiny millisecond-long bursts where they pull double their rated power. A PSU might be rated for 750W, but if it can't handle a 1000W spike for 20ms, the internal protection (OPP - Over Power Protection) will trip, and the PC will shut down.
No multimeter or cheap tester will catch this. You'd need an oscilloscope to see it. If your voltages look perfect but you’re still getting shut-offs during heavy gaming, try "undervolting" your GPU. If the crashes stop, you’ve confirmed the PSU can't handle the power spikes.
Physical Signs of a Dying Unit
Stop and use your senses. Sometimes you don't need a tool.
- The Smell: If you smell ozone, "hot electronics," or a metallic burning scent, stop immediately. A leaking capacitor has a very distinct, chemical fishy smell.
- The Sound: High-pitched whining (coil whine) is usually harmless, but a "clicking" sound followed by a shutdown means a protection circuit is tripping.
- The Heat: If the casing of the PSU is too hot to touch, the internal fan has likely failed. Without cooling, the components inside will bake until they fail.
Actionable Steps to Take Right Now
If you suspect your power supply is the weak link, don't just keep hitting the power button hoping it'll work. You risk a "catastrophic failure" where the PSU sends a surge through the 12V rail and kills your motherboard, CPU, and GPU in one go.
- Check your warranty. High-quality units from Seasonic, EVGA (SuperNova series), or Corsair (RMX series) often have 7 to 10-year warranties. Don't open the casing yourself; you'll void the warranty and risk a lethal shock from the large capacitors inside, which hold a charge even when unplugged.
- Verify the cables. If you have a modular PSU, ensure the cables are clicked in fully on both ends. A loose 24-pin clip can cause voltage fluctuations that look like a failing unit.
- Clean the dust filter. Most cases have a filter under the PSU. If it's choked with cat hair and dust, the PSU is overheating and throttling its output.
- Use a dedicated outlet. If you’re plugged into a $5 power strip with ten other devices, your PSU might be struggling with "dirty" power. Plug it directly into the wall to see if the stability issues vanish.
- Calculate your headroom. Use a tool like the OuterVision Power Supply Calculator. If your components need 550W and you're using a 600W "White rated" PSU, you're running too close to the edge. You want to stay around 50% to 80% of the PSU's rated capacity for maximum efficiency and longevity.
Testing a power supply is about elimination. If the paperclip test works, the multimeter shows steady voltages, and the software doesn't show massive sags under load, your problem is likely the motherboard or a bad driver. But if any of those steps fail, don't hesitate. Replace it. A $100 power supply is a lot cheaper than a $2,000 system.