Tap one suspect key normally and let GearCheck Lab measure the time from its previous release to the next non-repeat keydown. Normal held-key auto-repeat is excluded so the test focuses on unusually fast same-key re-triggers that can appear as double typing.
Click ARM KEY SELECTION, then press the key you want to test once. After it locks, tap that same key normally at least 20 times.
For each complete press, the browser sees a keydown and then a keyup. If the same key registers another non-repeat keydown almost immediately after the previous keyup, that close re-trigger can match the symptom users describe as double typing.
Operating systems intentionally generate repeated keydown events when a key is held long enough. Modern browsers mark those events with KeyboardEvent.repeat. GearCheck Lab counts them separately and excludes them from suspected-chatter timing.
There is no universal browser-visible millisecond threshold that proves a bad switch. Start conservatively and interpret a pattern of repeated short gaps, not one isolated event.
Lower thresholds reduce the chance of flagging intentionally fast tapping. Higher thresholds are more sensitive but also more likely to classify a genuine second press as suspicious.
Browser/OS shortcuts, function keys, Tab, Meta/Windows keys and system combinations can be intercepted before this page sees them. For a clean test, use a normal letter, number or punctuation key where possible.
That is expected. It proves the browser is labeling held-key repeat separately; those events are not counted as suspected chatter.
No. Retest the same key and look for a repeatable pattern. Browser diagnostics cannot directly observe the switch contacts.
Yes. Test one key, reset, then select a known-good key and use the same threshold and tapping style.