A screening check for your controller. We test each part on its own, sticks, buttons, triggers, vibration, and connection, and show a pass or a fail for each part.
Each part passes or fails on its own. There is no single score. Test a failing part separately with its dedicated tool.
| Part | Pass | Fail |
|---|---|---|
| Sticks at rest | Rest value steady at 0.10 or below after a clean retest, with no creep | Steady above 0.10, creeping in one direction, or random jumps while untouched |
| Buttons and D-pad | Every button registers once per press, no ghost press, no sticking button | Any button misses, double registers, ghosts, or sticks |
| Triggers | Full range from 0.00 to 1.00, returns to rest | Stops short of full pull, does not return to rest, or misfires |
| Vibration | Each motor responds when tested | A motor does not respond, or responds weakly on one side only |
| Connection | Controller detected, steady report rate at the expected native rate for the controller and connection | Not detected, wild jumps, dropouts, or a rate far below the expected native rate |
Health here means a pass or a fail per part. There is no single health score, no percentage, and no grade.
It is the instability in your stick's raw data. Even without drift, high jitter means your aim will feel shaky. We show it as part of the stick result, not as a score.
No. Firmware can only adjust deadzones. Physical wear on the stick parts requires a part repair or replacement. Firmware alone cannot fix true mechanical drift.
Hall effect sticks use magnetic sensing with no friction wear in the sensing path, so they resist drift longer than potentiometer sticks. They are not drift-proof. A weak centering spring or a calibration problem can still cause drift.
Competitive gamers should run a full audit every 3 months or after 500 hours of gameplay.
No. Overclocking is a PC wired only option, and it can make readings unstable. It does not change the health result, because health here is a pass or a fail per part, not a score.
When a button press is registered without you touching it. A button that registers without a press fails the button check. Test every button again with the button test.
A steady offset above 0.10 is clear drift. Clean and calibrate first, then decide between repair and replace. A deadzone setting only hides the offset, it does not fix the hardware.
Battery is a separate module. We focus on Input Performance which directly affects your K/D ratio.
Internal boards are similar to standard PS5 units, but their modular nature makes them easier to repair.
Usually, the internal rubber membrane is torn or the spring has lost tension. Test the trigger range with the trigger test. If the trigger does not reach full pull or does not return to rest, it fails the trigger check.
Complete your controller audit using our specialized testing suite. Each tool is precision-engineered to target specific hardware components—from analog sensors to circuit response times.
A health check is a screening pass over your controller. It looks at the parts that a browser can actually read through the Gamepad API, so you can spot which part needs a closer look. It does not repair anything, it does not certify a controller, and it does not replace the dedicated tools below.
| This check screens | This check does not diagnose |
|---|---|
| Stick drift at rest | Lab-grade latency (browser estimates only) |
| Stick deadzone | Native haptics or impulse triggers |
| Buttons and D-pad inputs | HD Rumble detail |
| Trigger range | Module-level repair diagnosis |
| Vibration response | Whether a module needs replacement |
| Polling behavior | End-to-end system latency |
The rule: a health check is screening, not a repair diagnosis. If one part fails, test it separately with its dedicated tool and record the before number so you can compare after any fix.
If the health check flags one area, take that part to its own tool for a precise reading:
All readings here are browser estimates through the Gamepad API, not lab measurements. A pass does not certify a controller for tournament play. A single fail does not prove the controller is unusable. Retest the failing part separately with its dedicated tool first.
How we test is documented in our testing methodology and on the about page.
Last updated: 14 November 2026