Buried in power supply reviews is a spec most builders skip: hold-up time. It sounds obscure, but it directly affects whether a brief power flicker crashes your PC or passes unnoticed. Understanding it helps you tell a genuinely well-engineered unit from one that merely looks good on the box.
Defining hold-up time
Hold-up time is how long a power supply can keep delivering stable, in-spec output after the wall power is cut. It is measured in milliseconds. The ATX specification calls for at least 16 milliseconds at full load, roughly the length of one AC cycle, so the system can ride through a momentary dropout without losing power.
| Hold-up time | Interpretation |
|---|---|
| Below 16 ms | Below spec, risk of crashes on flicker |
| 16-17 ms | Meets ATX minimum |
| 17-25 ms | Comfortable margin, well designed |
| Above 25 ms | Excellent, robust bulk capacitance |
Why it matters in real use
Mains power is not perfectly steady. Brief sags happen when large appliances start up, during grid switching, or in storms. A unit with healthy hold-up time bridges these gaps so your PC never notices. A unit that barely scrapes the minimum, or falls below it, may reboot or freeze during a flicker that a better supply would shrug off.
What determines hold-up time
It comes down mostly to the bulk input capacitor: its capacitance and voltage rating store the energy that carries the unit through a dropout. Cost-cut units sometimes fit a smaller bulk cap to save money, which is exactly why hold-up time is a useful tell for build quality that the 80 Plus badge cannot reveal.
How it relates to a UPS
Hold-up time covers only milliseconds. For real outages lasting seconds or longer, you still want an uninterruptible power supply. Think of hold-up time as protection against flickers and a UPS as protection against actual blackouts; they solve different scales of the same problem.
Where hold-up time fits among PSU specs
Hold-up time is one entry in a small set of specs that separate a genuinely good unit from a merely adequate one. On its own it proves little, but read alongside ripple, voltage regulation, and efficiency it paints a clear picture of engineering quality. These are the numbers seasoned builders scan in an independent review before trusting a supply with a whole system.
| Spec | What good looks like |
|---|---|
| Hold-up time | Above 16 ms, ideally 17 ms plus |
| Voltage regulation | Within 3% across load |
| Ripple on +12V | Well under the 120 mV limit |
| Efficiency | Meets or beats its 80 Plus tier |
A unit that clears all of these comfortably is one you can forget about for a decade. A supply that meets efficiency but skimps on hold-up time and posts loose regulation is cutting corners where you cannot see them, and it is exactly the kind that reboots your PC during a storm flicker while a better unit rides straight through. Since you cannot measure these yourself without lab gear, the practical move is to lean on reputable review outlets that publish the figures.
FAQ
Should I choose a PSU based on hold-up time alone?
No, but it is a valuable data point in reviews. A unit comfortably above the 16 ms minimum usually signals thoughtful engineering, which correlates with other quality traits worth having.
Does higher wattage improve hold-up time?
Not automatically. Hold-up time depends on the bulk capacitor relative to the load. A larger unit lightly loaded can hold up longer, but a well-designed smaller unit can beat a poorly designed larger one.
Bottom line
Hold-up time is how long your PSU keeps the lights on after power is cut, and 16 milliseconds is the floor you want cleared with margin. It quietly separates well-engineered units from corner-cutters, so treat it as one more reason to read reviews rather than trusting the badge on the box.
