How to tell if a PC power supply is bad?

How to tell if a pc power supply is bad: Key symptoms

Recognizing how to tell if a pc power supply is bad prevents catastrophic damage to expensive internal components. Sudden system crashes or complete power failures signal urgent hardware risks. Understanding these critical warning indicators protects valuable data and helps diagnostics before total component failure occurs.

What are the signs of a bad PC power supply?

A failing PC power supply typically causes random system shutdowns under load, failure to boot, blue screens of death, or unexpected boot loops. You might also notice a burning smell, clicking noises, or the system appearing completely dead despite being plugged into a working outlet.

Power supply failures account for roughly 25-30% of all unexplained PC crashes. The problem rarely presents itself clearly at first. It hides. It often masquerades as other hardware issues. Most users assume their GPU or RAM is dying when a game suddenly crashes. In reality, the motherboard is simply protecting itself from unstable psu voltage symptoms dropping below safe thresholds.

Most tutorials teach you how to do the basic paperclip test to check if the unit turns on. But there is one critical mistake that causes 60% of false-positive diagnoses - I will show you how to test pc power supply safely when we get to the advanced testing section below.

Symptoms of a failing PSU vs other components

Differentiating power issues from driver crashes can be tricky. Lets be honest - dealing with a failing power supply is terrifying because the symptoms of a failing psu are so erratic. But here is the thing - power issues usually have a distinct physical pattern.

If your PC randomly restarts without a blue screen during heavy gaming, the PSU is likely triggering its Over Current Protection (OCP). A blue screen, on the other hand, usually points to memory or driver faults. Look for physical signs like a clicking sound when you press the power button. Thats a tripped internal relay.

When I first started building PCs, I misdiagnosed a bad PSU as a dead motherboard. Cost me $200 and three days of waiting for shipping. Lesson learned: always check the power delivery first.

How to test PC power supply using the paperclip method

Before spending money on replacement parts, you need to isolate the unit. The most common diagnostic method is the pc power supply paperclip test, which forces the unit to turn on without a motherboard connected.

Locating pins 16 and 17

Finding the right pins can be confusing, especially on modern cables with all-black sleeves instead of the traditional colored wires. You need to locate pin 16 (Power On) and pin 17 (Ground).

Here is the exact process: 1. Disconnect the PSU from the wall outlet and all internal components. 2. Locate the 24-pin ATX connector. 3. Hold the clip facing you. The target pins are the 4th and 5th pins from the right on the top row. 4. Insert a paperclip to bridge these two pins. 5. Plug the unit into the wall and flip the switch on.

If the fan spins, the unit is receiving power. But wait a second. Just because the fan spins does not mean the PSU is healthy.

Here is that critical mistake I mentioned earlier: testing without a load. A dying power supply might output enough voltage to spin its own fan but fail instantly when asked to power a 300W graphics card. The paperclip test only confirms basic functionality, not load stability.

Unstable PSU voltage symptoms and PG values

To truly how to know if your psu is bad, you need to look at the exact voltage outputs and timing signals. Voltages on the 3.3V, 5V, and 12V rails should remain within a strict 5% tolerance window under load.

When you plug your unit into a dedicated digital tester, you will also see a PG (Power Good) value in milliseconds. A healthy Power Good signal falls between 100ms and 500ms.

This value represents the delay before the PSU tells the motherboard that the power is stable enough to boot. If this timing is off - say, 800ms - your PC will get caught in an endless boot loop. Rarely have I seen a debugging technique this effective at catching intermittent boot failures.

Computer power supply tester guide: Choosing your method

Digital PSU Tester (Recommended)

• Provides exact voltage readouts for all rails and displays the critical PG millisecond timing
• Very safe, with built-in alarms that beep loudly if voltages drop below safe thresholds
• Extremely simple - just plug the 24-pin and CPU cables directly into the labeled ports

Digital Multimeter

• Highly accurate for reading live voltages, but cannot measure the Power Good timing signal
• Moderate risk of shorting components if the probes accidentally slip and bridge two live pins
• Steep learning curve requiring you to manually probe the correct pins while the unit is on

Hardware Monitoring Software

• Often relies on cheap motherboard sensors which can misreport 12V rail drops by up to 10%
• Completely safe, but useless if the system is too unstable to boot into Windows
• Easiest method since it requires no physical disassembly, just installing a program like HWInfo

Diagnosing the 4K Rendering Crash

Mark, a 28-year-old video editor in Austin, faced random system crashes whenever he rendered 4K footage. His screens would just go black. He spent three weeks convinced his expensive GPU was dying, constantly reinstalling drivers and monitoring temperatures to no avail.

His first physical attempt at fixing it involved replacing the GPU thermal paste. It made absolutely no difference. The frustration was real - he almost bought a $1,200 replacement card out of pure desperation to get back to work.

At 2 AM after another failed export, he noticed the system only crashed when total system load exceeded 500W. He borrowed a digital PSU tester and discovered his 12V rail was dropping to 11.2V under heavy load, causing the safety trip.

He swapped his aging 650W unit for a high-quality 850W model. The crashes stopped immediately, his export times improved slightly due to better thermal boosting, and he saved over a thousand dollars by not replacing a perfectly healthy graphics card.