What This Online Keyboard Tester Does

This keyboard diagnostic tool tests physical switches, rollover matrices, contact bounce, and keystroke timing across four interactive modes:

  • 1. Key Tester (Matrix Visualizer): Displays an interactive 100% Full, 80% TKL, or 60% Compact virtual keyboard that highlights active and previously tested keys, logs raw W3C KeyboardEvent data (event.code, event.key, hexadecimal virtual keycode, and location index), and optionally plays synthesized Cherry MX Blue, Brown, or Red switch clicks.
  • 2. NKRO & Ghosting Test (?mode=nkro): Counts simultaneous held keys in real time and records your peak N-Key Rollover (2KRO, 6KRO, or full NKRO) to detect matrix blocking or ghosting on gaming and office keyboards.
  • 3. Switch Chatter Detector (?mode=chatter): Monitors the millisecond interval between consecutive actuations of the same key and flags any repeat under 50 ms as hardware contact bounce (chatter).
  • 4. Typing Latency & Hold Duration (?mode=latency): Measures the exact press-to-release hold duration (keydown to keyup) in milliseconds, calculating last, minimum, maximum, and average actuation times.

Whether you are diagnosing a sticky key on a laptop, verifying N-Key Rollover (NKRO) on a custom mechanical keyboard, checking for switch chatter (unintended double-typing under 50 ms), or inspecting raw W3C KeyboardEvent scan codes, our Online Keyboard Tester provides zero-latency hardware telemetry directly in your browser. Simply press any key—or hold multiple keys at once—to light up the interactive 100% Full, 80% TKL, or 60% Compact matrix.

1. From 1868 Typebars to X/Y Matrix Multiplexing: How Keyboards Scan Keys

In 1868, Christopher Latham Sholes patented a mechanical typewriter whose characters sat on pivoted metal typebars. Pressing adjacent keys in rapid succession caused the swinging arms to collide and jam inside the basket. The QWERTY layout separated common English digrams across opposite sides of the basket so typists could not outrun gravity. Although modern keyboards replaced mechanical levers with solid-state electronics, wiring 104 individual switches directly to a microcontroller would require 104 separate GPIO pins.

To solve this, keyboard PCBs arrange switches into an orthogonal X × Y matrix (typically 6 rows × 18 columns). The microcontroller's matrix scanner pulls one column high (or low) at a time and reads the voltage state of all 6 rows before stepping to the next column. Completing a full sweep of all 18 columns at a 1 000 Hz scan rate takes just 1 millisecond. When you press a key, its physical switch bridges the intersection of Row Ri and Column Cj, allowing the MCU to translate the (Ri, Cj) coordinate into a standardized USB HID usage code.

2. Ghosting, Jamming, 6KRO, and True NKRO Diode Architecture

When you switch to the NKRO & Ghosting tab and hold down multiple keys simultaneously, you directly test the electrical isolation of your keyboard's PCB matrix:

  • Matrix Ghosting & Phantom Keys: Imagine pressing three keys located at coordinates (R1, C1), (R1, C2), and (R2, C1) on a budget membrane keyboard that lacks per-key diodes. Because all three switches are closed, electrical current flows backward from Column C2 through (R1, C2), across Row R1 to (R1, C1), and down Column C1 into Row R2—falsely convincing the microcontroller that the fourth corner (R2, C2) is also pressed!
  • Matrix Blocking (Jamming / 2KRO): To prevent phantom characters from appearing on screen, membrane keyboard controllers implement software blocking: whenever a 3-corner rectangle is detected, the controller simply ignores the third key press. This is why holding W + A in a game often prevents Space or Shift from registering.
  • 6KRO vs. True NKRO with Schottky Diodes: Standard USB HID Boot Protocol descriptors allocate a fixed 6-byte array for active keycodes, limiting even diode-equipped boards to 6 simultaneous keys (6KRO). True NKRO keyboards solder a small-signal diode (1N4148 or SMD SOD-123) in series with every single switch so current can only flow in one direction, and transmit a 16-byte+ USB HID bitmask where each key owns a dedicated bit ($120+$ keys simultaneously).

3. Mechanical Switch Chatter: Contact Bounce Physics & Debounce Filtering

Inside a traditional mechanical switch (such as Cherry MX, Gateron, or Kailh), pressing the stem pushes a movable beryllium-copper leaf spring against a stationary gold-plated crosspoint contact. When two elastic metal surfaces collide at high velocity, they do not establish an instant, continuous circuit; instead, the leaf microscopically rebounds 3 to 10 times over 0.5 ms to 4.0 ms, generating a train of rapid high/low voltage spikes known as contact bounce.

Keyboard firmware suppresses normal bounce using an algorithmic debounce filter (such as eager or defer debouncing over a 5 ms window). However, as a switch ages, suffers leaf fatigue, or accumulates microscopic carbon oxidation and dust, the physical bounce duration stretches beyond the firmware's 5 ms threshold. A single finger press then emits two distinct USB HID keydown packets separated by 8 ms to 45 ms. Our Switch Chatter Test monitors every key's inter-release-to-press interval (Δ t < 50 ms) and flags defective switches in red.

4. Matrix Scan Rate vs. USB Polling Rate (125 Hz to 8,000 Hz)

Modern gaming keyboards heavily market 1 000 Hz (1.0 ms), 4 000 Hz (0.25 ms), and 8 000 Hz (0.125 ms) USB polling rates. However, total end-to-end keystroke latency is the sum of four sequential hardware stages:

Ltotal = Tmatrix_scan + Tdebounce + Tusb_poll + Tos_compositor

If a keyboard polls the USB bus at 8 000 Hz (0.125 ms) but scans its internal switch matrix at 500 Hz (2.0 ms) and applies a conservative 8 ms defer-debounce filter, its actual latency will be worse than a 1 000 Hz board running an eager 0 ms debounce algorithm or contactless Hall Effect magnetic sensors (which measure continuous magnetic flux density in milliteslas with zero physical contact bounce).

5. Key Hold Duration, Actuation Force (cN), and Biomechanical Ergonomics

In the Typing Latency mode, our telemetry engine measures your exact Hold Duration (dwell time in milliseconds)—the interval between keydown and keyup. Experienced touch typists using light linear or tactile switches (45 cN–55 cN, actuating at 1.5 mm–2.0 mm) typically exhibit crisp dwell times between 65 ms and 95 ms. Consistently long dwell times (> 140 ms) indicate heavy "bottoming out" against the 4.0 mm switch housing or steel plate, which increases repetitive strain on finger flexor tendons.