Mechanical Television

The Nipkow Disk

In 1884, German engineer Paul Nipkow patented (DRP 30105) a scanning disk that solved the fundamental problem of television: decomposing a two-dimensional scene into a one-dimensional signal. The disk is a flat circular plate with a series of small apertures (holes or lenses) arranged in a spiral pattern. Each successive hole is offset radially by exactly one scan line width from the previous one. As the disk rotates, each hole sweeps across the image in a circular arc, but because the holes are at different radii, each traces a different horizontal line.

Scanning Pattern Geometry

A typical Nipkow disk might have 24 holes arranged over a 180° arc of the spiral. As the disk completes one full revolution, the 24 holes collectively trace 24 scan lines—one complete frame. At 12 revolutions per second (720 rpm), the system produces 24 lines × 12 frames = 288 line-scans per second, yielding a visible frame rate of 12 fps. Some systems used 18 holes at lower speeds for better brightness. The radial spacing between holes determines vertical resolution: a 24-hole disk scanning a 4:3 aspect ratio image yields approximately 30 lines of vertical resolution (the extra lines account for blanking intervals and mechanical tolerances).

Rotation Speed and Resolution

The relationship between rotation speed, hole count, and resolution reveals the core limitation of mechanical TV. A Nipkow disk uses between 30 and 120 apertures depending on the desired resolution. To achieve 30 lines of resolution, the disk needs approximately 30 apertures spiraling from the outer edge toward the center. At 720 rpm (12 revolutions per second), this produces a 30-line image at 12 fps—a flickering but recognizable picture. To increase resolution to 240 lines, the disk must have 240 holes. At 15 frames per second, the disk must rotate at 15 rpm × (240/240) = 15 revolutions per second (900 rpm). Some high-resolution systems required up to 1,800 rpm.

The disk diameter must also increase to maintain adequate hole spacing—the outer holes must trace lines far enough apart to cover the full picture height. A 240-line system required a disk roughly 50–70 cm in diameter spinning at 900–1,800 rpm. At these speeds, windage losses, bearing wear, and vibration become severe. The practical resolution ceiling was approximately 240 lines, achieved by Baird in 1929 using a large aluminum disk with 240 apertures.

The Selenium Photocell

At the camera, a bright spotlight illuminates the subject. Light reflecting off the subject passes through the spinning Nipkow disk's apertures and strikes a selenium photocell behind the disk. Selenium's photoconductivity was discovered in 1873: its electrical resistance drops in proportion to incident light intensity. In series with a DC voltage source, the selenium cell generates a varying current that tracks the brightness of each point the aperture scans. This analog signal—varying between the dark current floor and the saturation current—directly represents one line of the image at a time. The signal is then amplified by vacuum tube amplifiers and transmitted over wire or radio.

Baird's Televisor

John Logie Baird built his first experimental television in 1924 at his flat in Hastings, using cardboard disks, darning needles for the apertures, and a biscuit tin for the housing. By October 1925, he achieved the first transmission of a recognizable human face—a toy doll's head, then his assistant William Taynton. The system used a 30-line format at 5 frames per second. Baird's first public demonstration on January 26, 1926, at the London Institution of Electrical Engineers, showed a live head-and- shoulders image of Taynton using a 30-line, 12.5 fps system. The image was roughly 2 × 3 inches—small but unmistakably a moving human face.

The camera used a single Nipkow disk spinning at 750 rpm (12.5 fps for 30 lines). A high-intensity arc lamp illuminated the subject with approximately 1,000 lux. The selenium photocell had a spectral sensitivity peak near 550 nm (green light), which is why early mechanical TV images were best at reproducing greenish tones. The photocell output was amplified by three stages of pentode vacuum tube amplifiers, producing a composite video signal with approximately 500 kHz bandwidth—enough for 30 lines of resolution at 12.5 fps.

The Neon Glow Lamp Display

On the receive side, Baird's Televisor used a neon glow lamp as the display element. A neon lamp consists of two electrodes sealed in a glass envelope filled with neon gas at low pressure. When voltage exceeds the striking voltage (approximately 90V), the gas ionizes and emits a characteristic orange-red glow. The lamp's brightness responds to the instantaneous voltage across it, making it a rudimentary amplitude- to-brightness converter. The received video signal modulates the voltage across the neon lamp, causing it to glow brighter or dimmer in sync with the transmitted image. A second Nipkow disk, synchronized with the camera disk, scans the glowing lamp to reconstruct the image.

The neon lamp could only display brightness modulation in a single hue (orange-red), so the image was monochrome even when Baird later demonstrated color transmission experiments. Approximately 20,000 Televisor sets were sold between 1929 and 1934, though reception was often poor, requiring the viewer to sit close to the tiny screen in a darkened room. The first model (1929) was a complete kit costing about £25—equivalent to roughly £1,500 today—consisting of a Nipkow disk driven by a hand-cranked motor, the neon glow lamp, and a simple amplifier circuit.

Baird's Color and Stereoscopic Experiments

Baird was remarkably forward-thinking. In 1928, he demonstrated a color television system using two Nipkow disks—one for the subject illuminated by alternating red and green filters, the other for reconstruction with matching color filters. The color sequential method transmitted red and green fields alternately, relying on persistence of vision to blend them. He also demonstrated stereoscopic (3D) television using two side-by-side camera disks and polarized viewing glasses. In 1941, he demonstrated a 600-line color system using three interlaced scans, but by then electronic television had already rendered mechanical systems obsolete.

Farnsworth's Image Dissector

Philo Farnsworth's Image Dissector (patented 1927, first demonstrated September 7, 1927) was the first fully electronic television camera—and represented a fundamentally different approach from mechanical scanning. Instead of a spinning disk, the Image Dissector used a photoemissive surface that emitted electrons in proportion to incident light intensity, forming an electron "image" of the scene inside the tube. A magnetic deflection coil then swept the entire electron image across a small anode aperture. At any instant, only electrons from one tiny region of the image passed through the aperture to the collector, producing a time-varying current corresponding to the scene's brightness at that scan point.

No moving parts. No spinning disk. The resolution was limited not by mechanical constraints but by electron optics and signal-to-noise ratio. The photoemissive surface was a thin film of cesium-antimony (CsSb) deposited on a glass window, with a quantum efficiency of approximately 10–15% in the visible spectrum. The electron optics used electrostatic focusing at approximately 2 kV, with magnetic deflection coils providing the raster scan. Farnsworth transmitted his first image—a simple line drawing of the letter "F"—at his lab in San Francisco at age 21.

Fundamental Limitations of Mechanical TV

Mechanical television suffered from four insurmountable constraints.Resolution vs. rotation speed: Doubling resolution requires either doubling the number of holes (larger disk) or doubling the rotation speed (mechanically impractical above ~1,800 rpm). A disk with 240 apertures at 900 rpm produces 240 lines at 15 fps—adequate but flickering. To reach 480 lines would require either 480 holes (a very large disk) or 1,800 rpm (causing dangerous vibration and windage losses).Brightness:The Nipkow disk blocks roughly 95–98% of incident light, because each aperture is tiny and only transmits light for a fraction of the rotation period. The resulting image is dim—viewable only in near-darkness.Response time: Selenium photocells have a response time of several milliseconds, causing motion blur and smearing on fast-moving subjects. Mechanical noise: Disk motors produce audible hum and vibration. Electronic television eliminated every one of these limitations: electron beams have effectively infinite resolution potential, phosphors emit light directly, photoconductive tubes respond in microseconds, and there are no moving parts.

The 1936 Berlin Olympics

The 1936 Berlin Olympics represented the high-water mark of mechanical television in public visibility. The BBC broadcast portions of the Games from Alexandra Palace using their improved 240-line Baird system, reaching approximately 400 home receivers and large-screen displays in London pubs and public halls. Public interest was enormous—tens of thousands of Londoners gathered to watch grainy, flickering images of athletes on screens no larger than a postcard. The broadcasts alternated between the Baird mechanical system and the new EMI-Marconi electronic system (405 lines). By January 1937, the BBC chose the electronic system exclusively, ending the mechanical era. Baird's system had proven the concept; electrons would carry television forward.