Keyboard Tester: Detect Chatter, Double Typing & Rollover Limits

Detect duplicate key presses, rapid switch chatter, stuck keys and keyboard rollover locally.

Keyboard Chatter & Switch Tester records every keydown and keyup your browser delivers while you type across an on-screen layout, then analyzes that transition log for duplicate presses, unusually fast re-presses, and how many keys the browser ever saw held down at once. It reports what the browser observed key by key; genuine electrical chatter and an intentionally fast double-tap can look identical in the data, and the tool says so rather than pretending otherwise.

What counts as chatter, and what the test deliberately excludes

A keydown event carries a repeat flag when the operating system's own key-repeat feature fires it, and this tool counts those separately from real presses; auto-repeat while holding a key down is never treated as suspicious chatter, since it is expected behavior rather than a switch double-firing.

What does get flagged is a key going down again within an adjustable window, thirty-five milliseconds by default and tunable from ten to one hundred fifty, after its own previous release, called a rapid re-press, and a keydown arriving while the tester never saw a matching release from the previous press at all, called a duplicate down. Both get counted and shown per key, along with the shortest re-press interval recorded for that key.

How to test your keyboard for chatter and rollover

  1. Click Start keyboard test; the dashed capture area takes focus and recording begins.
  2. Press every key once and confirm the on-screen map marks it as tested. The map draws the main block, the function row and the arrow keys; the numpad, Insert/Home/Page keys and media keys are not drawn but still get their own rows in the results table.
  3. For a key you suspect of chatter, tap it 20 to 30 times at a normal pace, lifting fully between presses, then check its row for rapid re-presses or duplicate downs.
  4. To check rollover, hold several letter keys at once and watch the Held now line; skip Ctrl, Alt and Windows or Command combinations, which the page leaves to the browser and system.
  5. Press Escape to stop, review the per-key table, and export the event CSV before starting a new test, because a new test clears the session.

Rollover reflects what the browser saw, not a hardware ceiling

Maximum rollover tracks the largest number of keys the tester ever saw simultaneously held during a session, updated every time a new key goes down while others are still down. That is a genuinely useful number for spotting ghosting, combinations of keys that silently fail to register together, but it is a browser-observed figure rather than a hardware specification; a keyboard's real n-key rollover capability depends on its own USB report protocol and can differ from whatever a given test session happened to exercise.

The escape key is reserved to stop the test rather than being captured as test input. While recording, the page blocks the default action of ordinary and navigation keys so they stay inside the capture area, but it deliberately leaves Tab and any combination with Ctrl, Alt or the Windows or Command key to the browser, so shortcuts keep working; pressing Tab moves focus away, and losing focus stops the recording.

The on-screen map, the results table, and the exportable log

A physical keyboard-shaped map highlights every key by its physical position, spanning the function row, the number row, the three letter rows, and the bottom modifier row with the four arrow keys, using the key's physical code rather than the character it produces, so the map stays correct regardless of your operating system's keyboard layout or language setting. The numpad and the Insert, Home, End and Page keys are not drawn, but presses on them are still recorded in the table.

Every key that registered a press gets its own row in a results table listing total presses, releases, auto-repeat count, possible-chatter count, shortest re-press interval, and average and longest hold duration. The full transition log, up to the four thousand most recent events, can also be downloaded as a spreadsheet-ready file with a timestamp, event type, physical code, and character for every row.

Chatter, debounce and ghosting explained

How to fix keyboard chatter

Where browser-side detection stops being conclusive

Questions people ask about Keyboard Chatter & Switch Tester

What is keyboard chatter, and why would I want to test for it?
Chatter is when a single physical key press registers as multiple presses in rapid succession, usually from a worn or dirty switch contact briefly reconnecting after release. It shows up as characters double-firing in games or text editors inserting duplicate letters, and this tool flags it by watching for keydowns arriving unusually soon after that same key's own previous release.
Does this test measure true n-key rollover?
It reports the maximum number of keys the browser ever saw held simultaneously during your session, a useful real-world signal but not the same as a certified NKRO rating. The real ceiling depends on the keyboard's own USB reporting protocol and can be higher than anything a given test session happened to exercise.
Why does not holding a key down and triggering auto-repeat count as chatter?
Every keydown event carries a flag indicating whether the operating system's own repeat-while-held feature generated it, and this tool checks that flag specifically to exclude those from the chatter count. Counting expected auto-repeat as suspicious activity would flag nearly every normal keypress, so only unexpected rapid re-presses and duplicate downs are treated as anomalies.
Can I save my test results to look at later?
Yes — the full event log, up to the four thousand most recent keydown and keyup events, can be exported as a downloadable file with a timestamp, event type, physical key code, and character for each row, so you can review or share the raw data outside the browser session.
How do I fix keyboard chatter or double typing?
Confirm the flag by retesting the key, then clean the switch with compressed air, raise the debounce time if your firmware allows it, and rule out the port or cable. A switch that keeps double-firing after that needs replacing, either by hot-swapping it or through a repair or warranty claim.
What is keyboard ghosting?
Ghosting is when holding several keys together makes a key you never pressed appear, or makes one you did press go missing. It comes from how the keyboard's scan matrix is wired, so holding combinations while watching Held now shows which groups of keys fail together.
What chatter threshold in milliseconds should I use?
Start with the default 35 ms, which catches most switch bounce without flagging ordinary typing. If deliberate quick double-taps keep getting flagged, lower it toward 20 ms; to catch slower bounce, raise it toward 50 ms. Changing the value recalculates existing results once recording has stopped.
Are my keystrokes stored or sent anywhere?
No. Key events stay in this browser tab and disappear when you start a new test or close the page. The exported CSV includes the character each key produced, so avoid typing passwords or private text during a test if you plan to share that file.
Why doesn't a key light up?
The map omits the numpad, the Insert, Home, End and Page keys and media keys, so check the results table for those. Escape stops the test instead of lighting up, Fn is normally handled inside the keyboard, and system shortcuts may never reach the page. A key missing everywhere may be faulty.

Related tools on BrakeTools

Content reviewed 2026-09-27.