Switch health

Double Click Test

If your mouse has started registering two clicks when you clicked once, this finds it. Not a speed test — it times the gap between presses, because that gap is what separates a worn switch from a deliberate double-click.

Gap between each click and the last

under 35 ms — a fault 35–80 ms — borderline over 80 ms — a hand can do this

Each bar is one gap, newest on the right, on a logarithmic scale so the 5–500 ms range that matters is readable. The page does not try to spot deliberate double-clicks: a double-click is a pair set apart from its neighbours, which a continuous run cannot show.

Double-fires: 0
Borderline: 0
Normal gaps: 0
Shortest gap: —

How to

How to test for double-clicking

Click the pad at an ordinary pace, twenty or thirty times. Do not click fast: a fast run puts real gaps into the same range as a fault and hides the problem.
Watch the plot. Each bar is the gap since the previous click, newest on the right.
Read the verdict. Any gap under 35 milliseconds is a double-fire — no hand produces that.
Run it again if the result is borderline. One stray reading is not a diagnosis; a pattern across two runs is.

The measurement

What double-firing actually is

A worn contact, not a software setting

A mouse switch closes a metal contact, and as that contact settles it bounces. The firmware ignores state changes for a few milliseconds afterwards — the debounce window — so the bounce never reaches your computer. As the switch wears, the bounce lasts longer than the window, and one physical press arrives as two clicks.

That is why the fault is progressive, why it gets worse under a hard click, and why no driver setting fixes it. Windows' double-click speed slider changes how the operating system interprets two clicks; it does nothing about a switch that is sending two.

Why 35 milliseconds is the line

The fastest sustained human clicking on one button is around 16 presses a second, which is a gap of roughly 60 milliseconds. A trained double-tap can be quicker for a single pair, but a hand cannot get under about 35 milliseconds — the finger cannot leave and return that fast.

So gaps below 35 ms are a mechanical fault by elimination. Between 35 and 80 the tool says borderline rather than guessing, because that band contains both a very fast deliberate pair and an early-stage bounce, and only a repeated pattern tells them apart.

What to do about it

A double-firing switch is a hardware repair. Replacing it is a soldering job and the parts cost very little; some mice have hot-swappable switches that need no soldering at all. Some people report success cleaning the contact with contact cleaner, which sometimes buys months.

Until it is fixed, the mouse will keep getting worse, and it will quietly corrupt any click-rate measurement you take with it — a chattering switch inflates a click test, which is the reverse of the usual complaint about a dying mouse.

The bench

Every tool on the bench

One instrument per input, all on the same clock. A rate is only comparable against the same tool and the same interval.

Background

Read the method

Longer entries on what this tool measures and where the number comes from.

Debounce, chatter and phantom clicks

Every mouse ignores its own switch for a few milliseconds after each press. When that guard fails your click rate rises — and your mouse is dying.

Read the entry

Auto-clickers and meaningless numbers

A script firing 50 clicks a second reads as 3,000 CPM. The chronometer counts it, says so, and explains why filtering would be worse.

Read the entry

Questions

Asked of the chronometer

How many clicks should I test before deciding?

Thirty to fifty ordinary clicks, at a normal pace. Early-stage chatter is intermittent — a worn contact might misfire once in twenty presses — so a short run can easily miss it and a single clean run proves very little.

Could it be a setting rather than the mouse?

The gap figure settles it. Windows' double-click speed slider decides how far apart two clicks can be and still count as a double; it cannot create a second click. A pair under 35 milliseconds is a switch sending two presses, and no setting on the computer produces that.

Does a chattering switch affect my click speed results?

Yes, and upwards. Every spurious press is counted like a real one, so a dying mouse quietly inflates a click test rather than lowering it. If a CPM figure jumped without your technique changing, this is the first thing to rule out.