Vintage Television Sets and Colour Television Sets from the Dawn of Television until Now

Echard Etzold's Site

Those which live with everyday television today can't image how television began in the 1920s and 1930s. But believe it or not, television was earlier invented than the electric light bulb. Paul Nipkow was born in 1860. The idea to scan a picture ith the spiral scanning disk came in his mind on christmas eve 1883. The scanning disk divided the picture into lines which were scanned from the top to the bottom of the frame. The disk had to spin so fast that the human eye didn't notice the single line and frame writing, but only see the whole picture. But Nipkow was not the father of television like many Germans claimed. Before him, the Scottish clockmaker Alexander Bain invented the facsimile (fax). He used a kind of detector, installed on a pendulum, to scan a picture on a paper point by point, line by line. The pendulum was synchronized by a metronome. The first attemps for transmitting a moving picture used one transmitting wire for one selenic cell, connected with a lamp in the receiver. For a panel of 100 cells per line and 100 lines, 10,000 whires were necessary. The french inventor Constantin Senlecq had the idea three years before Nipkow invented the television scanning disk, to transmit the video signal seriell with one wire. Therefore the picture was scanned in lines. Behind a panel of photosensitive cells a rotating switch scanned the picture line for line. In the receiver the same rotating switch connected lamps line for line. While Senlecq described the theoretical basics of scanning and transmitting a television picture, Nipkow found a practical way to scan moving pictures with a high frame rate. But he had to wait until the photoelectric cell and electronic amplifying was invented until the first working televisors were realized. This kind of television is called mechanical television because the frame scanning was done with a mechanical disk instead of iconoscopes and kinescopes.

The first condition for mechanical television is a scanning disk. I have got it from Peter F. Yanczer, together with a D.C. motor and a hub. The scanning disk has a diameter of 12 inches and has a row of 32 holes in an inner circle for synchronisation. This format is similar to that what was introduced by John Logie Baird in 1928 in Great Britain. Those which want to built a scanning disk for themselfes can download here a Excel-file. (For creating a simple scanning disk copy the diagram into a vector graphic program and print it on a black piece of paper (with yellow ink). Then sting the holes with a needle. Be careful with the center hole, this had to be very exact, otherwise the lines will overwrite each other.

For the first attemps I took a piece of wood with two corner tins, which hold the Nipkow disk. Beneath the corner tins, the motor was fixed. Behind the disk on the right side six LEDs were placed, every LED with a 47 Ohm resistor in series. The motor was driven by a controlled D.C.-output transformer for model trains. The LEDs were connected over a 20 Ohm resistor directly with the audio output of a stereo amplifier. Parallel to the LEDs a 4.5 volts battery together with a 10 Ohm resistor was applied for generation the break-though voltage of the LEDs. With this very simple construction it was possible to generate the first television pictures on the Nipkow disk. The room was dimmed.But only som white slopes and lines were visible. I had to turn the polarity of the audio output. The motor synchronisation must be done manually with the D.C. control. It is very difficult to take a photo of a movie while synchronising the disk with the other hand and keep the vertical hold.

The empty housing of the monitor only with the loudspeaker and the power supply. The mechanics of the motor were very simple. The disk is driven by a belt. The ball bearing which hold the disk was fixed with two corner tins on a wooden plate. The motor was fixed too with a piece of tin and some rubber. The disk will be driven with a belt. For protection a glas door before the scanning disk was installed. For operating the motor in the first attempts I used an adjustable D.C. source for the model railway of my son. Later the motor was driven by an automatic speed control circuit. The schematics I got from the website of the Narrow-bandwidth Television Association.

The electronic board is pushed into the housing, but the scanning disk wasn't still installed. On the right and left side beneath the board some screws fixed the board into position to prevent removing of the board from the back side. TRhis could damage the scanning disk. The height of electronic board from the bottom of the housing is 11.7 cm.

The automatic speed control circuit is feed with the sync pulses from the video board and a pulse which is generated in an optical fork, which registrates the exact position of the Nipkow disk. The automatic speed control circuit compares both pulses and regulates the motor speed for best steady picture. Left: back view of the televisor, right: front view of the televisor.

See the set from the back and frontside with all units and mechanics. For the front side I will to install a glas. In the past scanning disk sets were closed with only a small window for the picture. But I think it is so interesting to watch a scanning disk set working so I decided not to put it into a closed housing.

In the 1920s and 1930s neon lamps usually were used for the light source. The light of a neon bulb is easily to modulate with the video signal. Today I used white LEDs instead. Six LEDs were installed parallel, each together with a 47 Ohm resistor. The LED cluster was connected with a video amplifier. But don't forget that this amplifier needs only the bandwidth of an audio amplifier. The LED video driver unit also provides the 400 Hz sync pulses for the automatic speed control circuit. Today computer software exist with which one can convert AVI videos and MPEG videos into NBTV 32 line standard video.

A clear sync pulse (see photo on the left side) is absolutely necessary for synchronising the scanning disk. The sync pulse will be compared in the 4046 with the reference pulse from the optical fork (see photo on the right side). In the first attempts the drive voltage for the motor was so high that no synchronisation was possible. But after increasing the d.c. current control the line synchronisation worked.

The power supply provides two D.C. voltages, separated for the LED driver and the motor control. A special problem is the light beam of the LEDs. This light is focussed. Instead of this a diffuse light is necessary which illuminates equally the frame window of the disk. The diffuser here was taken from an old slide viewer. With computer software an AVI file was converted into a 32 line WAV file which contain the video signal.

After testing the video unit the motor was tested. The picture is scanned vertical according to the video standard of the British Narrow-bandwidth Television Association (NBTV). The lines will be written from the right to the left. The disk's revolution is counterclockwise. After darken the room some fuzzy lines were visible on the disk, but no picture. Bright bold lines. I noticed that I have to change the polarity of the audio output. Because there is no automatic synchronisation in the first attemps, the disk has to be synchronised manually.

With the automatic speed control circuit is was possible to get an automatic synchronised picture, One should remember that tv watchers at the end of the 1920s have to synchronise the disk manually, usually with a wired remote control. And they were happy when they got a steady picture for some seconds.

When the first draft of the monitor was working, it was time to create a more professional housing and mechanic for it. Our master of the city here, the physicist Dr. Ulrich Barkow, is profund in woodworking. He disassembled the whole set and created a new wooden housing. The space inside was enlarged. A wooden board with the mechanics was inserted in such a way that one can pull it out to adjust the controls and mechanics.

The opto fork was placed on a wooden block which is fixed with screws. Motor and hub were also placed into wooden blocks for a more reliable operating of the disk. The electronics were placed on the rear side of the wooden board.

Now the monitor is easy to operate by the possibility to pull the wooden board without disassembling the glass door. The controls were placed at the right side. These are potentiometers for motor servo, synchron impulse level, contrast, white peak and black level adjustment and loudness. One switch is for interupting the synchron impulse to set the frame synchronisation into the correct phase. Examples for the video performance of the Nipkow scanning disk are online at Youtube: Mechanical Television I, Mechanical Television II, Mechanical Television III.

Update: In January, 2010, the set was upgraded to colour with the world converter of Darry Hock. The World Converter generates a 32-line NBTV-signal with an RGB output to drive a RGB LED-array.

A daughter-board for the world converter is available which has three optimised RGB drivers. For every colour channel, level and gain can be adjusted individually. The RGB LED-array has 32 powerful LEDs.

I have compared the colors of the RGB LED array picture on the Nipkow disk with the colors on a screen of a PAL color tv set, both feed with the same test patterns. Green of the LEDs seems to be much "deeper" and saturated than the green on the CRT screen, and also red of the LEDs seems to be much "deeper" and saturated than the red on the CRT screen. The LEDs seems to provide a wider gamut than the colors of the CRT. This obeservation was confirmed by the actual values for the wavelengths of the RGB LEDs of mechanical television:

632nm (red)
523nm (green)
465nm (blue)

This seems to be more similar to the wavelengths of the early NTSC color picture tubes than with PAL (except of blue):

636 to 658nm (red)
523 to 532nm (green)
442 to 452nm (blue)

For comparison PAL/SECAM dominant wavelength (these values are estimated from the color space triangle):

607nm (red)
552nm (green)
466nm (blue)