How does this keyboard test read physical switches via event.code?
Every keycap on this visual tester listens to keydown and keyup events directly inside your browser tab, tracking event.code, event.key, Peak rollover, and mechanical switch bounce locally. No keystrokes, timings, or hardware identifiers leave your device, making this NKRO switch bounce test safe even on shared work laptops.
Comparing physical event.code vs localized event.key output
Browsers expose two distinct identifiers per keypress. Under the W3C UI Events KeyboardEvent code Values standard, event.code names the physical switch position (KeyQ, ShiftLeft, ShiftRight, or IntlBackslash), whereas event.key reports the typed character produced by your operating system language map.
Binding each visual tile to event.code ensures the physical switch highlights accurately even on French AZERTY, German QWERTZ, or Dvorak layouts. Meanwhile, the sidebar displays Last event.code beside Last event.key so you can verify both hardware wiring and software mapping in one glance.
- Untested vs tested states: Untested caps stay white, held caps invert to solid ink, and released keys remain tinted green until you click Reset test.
- Built-in system limits: Firmware-only keys like
Fnand OS-reserved hotkeys such asPrintScreenor hardware volume controls are intercepted before any browser receives aKeyboardEvent.
Preventing browser hotkeys and Alt-Tab stuck modifiers
Testing Space, Tab, Arrow keys, F1, F3, or F5 normally scrolls the page, moves focus, or opens browser find/help dialogs. When Capture active is enabled, the listener calls preventDefault() on those keys outside editable controls so key presses stay centered on the diagram.
Switching tabs with Alt+Tab or Cmd+Tab fires window.blur, which clears Active keys immediately instead of leaving phantom held modifiers on screen. On hybrid laptops or tablet docks, pair keyboard verification with our Touchscreen and multi-touch test to test the full chassis.
Measuring NKRO vs 6KRO rollover and matrix ghosting
Keyboards wire switches inside an electrical grid of intersecting rows and columns rather than running an individual wire per key. Testing how many keys appear simultaneously under Held keys and Peak rollover quickly reveals whether your board supports 2KRO, 6KRO, or full N-key rollover.
Why wired USB supports NKRO while Bluetooth drops to 6KRO
Standard USB HID boot protocol packets carry at most 6 simultaneous non-modifier key codes plus modifier bits (Shift, Ctrl, Alt, Meta). True NKRO keyboards switch to extended HID report descriptors over wired USB or a 2.4 GHz dongle so every switch on the board registers together.
When switched to Bluetooth Low Energy, many wireless mechanical keyboards revert to standard 6KRO HID profiles to conserve battery and stay compatible with phones and BIOS screens. If Peak rollover stops at 6 plus modifiers over Bluetooth, plug in the USB-C cable and re-test all finger combinations.
Spotting blocked diagonals vs phantom matrix ghosting
As documented in the MDN KeyboardEvent code Documentation, left and right shifts (ShiftLeft, ShiftRight) report distinct physical codes, making it straightforward to test symmetric modifier combos. Inexpensive membrane boards without per-switch diodes sometimes exhibit ghosting or anti-ghosting blocking when three keys share matrix corners.
- Matrix blocking: Holding
W,A,ShiftLeft, andSpacecauses a fourth key likeDorRto stay silent in Held keys because the controller blocks ambiguous paths. - Phantom ghosting: Holding three corner keys lights a fourth untouched tile in the grid because electrical current back-feeds across shared traces.
If you switch between keyboard WASD movement and analog controller play on PC, inspect both devices before competitive matches. Keep a browser gamepad stick circularity checker handy to test thumbstick drift alongside your keyboard.
Running a key chatter test with the millisecond bounce slider
Mechanical key chatter happens when worn metal contact leaves vibrate open and closed on a single physical keystroke, producing doubled characters such as tt or ee. Adjusting Chatter bounce limit lets you screen suspect switches quantitatively down to 0.1 ms resolution.
Tuning the 15 ms to 80 ms bounce limit without false alarms
The Chatter bounce limit slider spans 15 ms to 80 ms in 5 ms steps, defaulting to 35 ms. Every time a key releases (keyup) and triggers a fresh keydown within that active millisecond threshold, the tester appends a fault entry and increments Chatter.
The diagnostic retains up to 50 bounce events per session and lists the 4 most recent faults inside the Switch chatter detected card with exact millisecond gaps (for example KeyE re-triggered in 18.4 ms). Remember that optical and Hall Effect switches lack physical metal contacts and rarely fail below 15 ms.
Separating degraded metal leaf springs from fast tapping
Because holding down a key causes your operating system to fire rapid keydown repeats, the diagnostic explicitly ignores every event where e.repeat is true. Only genuine release-and-retrigger cycles count toward Chatter, ensuring held movement keys never pollute your fault log.
Tip: Tap each suspect key 20 times at natural typing speed with Chatter bounce limit at
35 ms, because even fast deliberate double-taps rarely fall below85 msbetween key release and the next bottom-out.
When a worn keyboard switch shares the same desk as an aging gaming mouse, test both sets of contacts together. You can measure switch bounce gaps in your mouse using the same millisecond threshold approach.
Comparing ANSI, ISO, TKL, 75%, and Mac keyboard layouts
Choosing the right entry under Keyboard layout aligns the on-screen diagram (1u = 38 px wide, 36 px tall) with your exact physical chassis. All five layouts reuse the same underlying W3C physical key codes while toggling the navigation cluster, numpad, ISO split keys, or Apple command row.
| Layout option | Key blocks shown | Physical layout distinctions | Best hardware match |
|---|---|---|---|
| Full ANSI (104-key) | Main, navigation, numpad | Wide 2.25u Enter, long 2.25u ShiftLeft, 6.25u Space |
Standard US full-size desktop keyboards |
| Tenkeyless TKL (87-key) | Main and navigation block | Omits right numeric keypad (NumLock to NumpadEnter) |
87-key gaming and developer mechanical boards |
| Compact 75% / Laptop | Condensed main + inline arrows | Tight F1–F12 row ending with Del, tucked arrow cluster |
75% custom keyboards and Windows laptops |
| Full ISO / UK / EU (105-key) | Main, navigation, numpad | Vertical Enter, # key (Backslash), short 1.25u ShiftLeft + IntlBackslash |
UK, German, French, and Nordic keyboards |
| Mac laptop / Magic Keyboard | Compact main + Mac bottom row | Control, Option, Cmd, 5.5u Space, right-hand arrows |
MacBook keyboards and compact Magic Keyboards |
Physical lock buttons such as CapsLock, NumLock, and ScrollLock light up normally whenever you press and release them on the key map. Checking whether an operating-system lock state was already latched before opening the browser page relies on the MDN KeyboardEvent getModifierState API (CapsLock, NumLock, or AltGraph modifiers).
If you test new laptop hardware straight out of the box, verify every switch here before peeling off screen film. Then switch over to spot stuck subpixels on matte or IPS monitors to complete unboxing checks.