As you may know, in the very earliest days of television in the early 1900s, inventors were experimenting with mechanical means of generating a raster scan, but apparently, “all” of them failed to produce sufficient resolution to gain popularity with the public. Later, when electronic raster scanning came along through cathode ray tubes in the 1920s, the resolution was greatly improved, thus started the adoption of television by the public at large. Then, when LCD monitors came to market in the late 1990s, businesses at large switched to the space-saving, and energy-saving technology. Never again would anyone want to use the old CRT monitors because they contained so many hazardous materials, unlike the modern flat-panel display technology, and pretty soon, in the late 2000s, it was a race among companies to get rid of CRT monitors as quickly as possible. CRT manufacturing was pulled to a halt, and old stock CRTs were sold at considerable discounts to deplete the inventory as quickly as possible.
Well, this isn’t the full story, as it must turn out. Although it is an accurate reflection of the public consciousness, it is wrought with many misconceptions. Worst of all, if you take this viewpoint to heart, it stifles your ability to build useful Raspberry Pi electronics projects in the modern era.
So, let’s retrace our steps to the beginning, review how each such earlier technology is still being used in the present, and ultimately, present a viable “modernized” design for building a raster-scan display.
First, there was the mirror galvanometer, generally commercialized sometime in the late 1800s. Yes, this is a pretty old technology. The main design intent and improvement over the traditional needle display galvanometer was to allow viewing the voltage properties on a finer scale. The mechanism was to attach a galvanometer to a mirror and shine a beam of light at the mirror; changes in voltage would deflect the light at a different angle. Ah, now you’re talking, just about all the properties you would need to build a “high resolution” television display, using relatively primitive opto-mechanics, free of hazardous chemicals. In fact, high-speed mirror galvanometers are used in modern laser light show graphics displays to drive deflection of the laser beams to draw graphics.
Indeed, from this point followed a few parallel paths of development. Up and coming was the galvanometer’s more sophisticated cousin, the oscilloscope. Early implementations built off of the mirror galvanometer. For example, one particular design used photographic film to record the path that the light beam took. The other parallel path of development was cathode ray tube oscilloscopes.
From these two branches, cathode ray tubes were primarily used for oscilioscopes for quite a while, but mechanical television research, design, and development picked up quickly from rotating apertures and mirror galvanometers. It wasn’t until quite a while later that cathode ray tubes were improved enough to be practically usable outside the lab, so during the early 1900s, mechanical television had an edge. But then, as times must have turned out, World War II came around in the 1930s, and this radically changed engineering priorities across the world. The end result was that at the time, since cathode ray tubes had a clear use as radar screens, they were viewed as much more valuable for the war effort, and although there were some quite impressive mechanical television projects out and about, they basically got pulled to a standstill. Then, after the war was over, it was pretty much unanimous that only CRT televisions could be considered ready for prime-time viewing.
One of the most interesting mechanical television project that made it pretty far along in development until getting canceled because of World War II was the Scophony. This was a mechanical television projector device that used rotating mirrors to scan out a television frame. The highest speed rotating mirror motor operated around 30,000 PRM. That’s totally doable with modern DC micro-motors, although the vintage design used an AC motor, probably because brushless DC motors were not practical at the time. The high-speed motor-mirror was used for horizontal scan, and a lower speed motor-mirro was used for vertical scan.
Interestingly, beam focusing was done using two cylindrical lenses rather than one spherical lens. The light beam was generated from a halogen lamp. As for modulating the light signal, this system used an interesting technique. Rather than modulating the current to the lamp, which could risk making it burn out faster, a chamber of water is acoustically vibrated at varying frequencies to change its index of refraction. These changes can range from passing the light beam clearly to scattering it off in a different direction. This was called a “Jeffree cell.”
As you can see, these are all relatively primitive technologies involved in this projection television display that can be built by hand or by primitive machine tooling without the assistance of an already-invented computer.
20190726/https://en.wikipedia.org/wiki/Mechanical_television
20190726/https://en.wikipedia.org/wiki/Scophony
20190726/https://en.wikipedia.org/wiki/Jeffree_cell
20190726/https://web.archive.org/web/20071015191810/http://nightlase.com.au/lasermame/
20190726/https://en.wikipedia.org/wiki/Radar_display
20190726/https://en.wikipedia.org/wiki/Oscilloscope
20190726/https://en.wikipedia.org/wiki/Oscilloscope_history
20190726/https://en.wikipedia.org/wiki/Mirror_galvanometer
So, that is the history behind television. One of the big reasons why CRT monitors caught on in the mainstream was because of a fortunate turn of events, but one might still argue that they still would have gone mainstream simply due to their technical advantages over the mechanical systems even if it weren’t for the war. But, one thing is clear, that if you want to reinvent a raster scan display that is worthy of use with modern electronics hobby projects, building your own CRT monitor is a no-go in the modern era. Yeah, people of times past did it, but that is just too culturally obsolete to be repeated, not to mention the need for controlled hazardous chemicals that are harder to come by in the modern era.
Also, with the lack of vector CRTs in the modern era, some retro gaming users have turned to laser light displays as a viable replacement for a true vector CRT. It certainly does give a more authentic look than primitive emulated rasterization may give.
So, let me detail and propose my design that works well with Raspberry Pi.
You have two rotating triangular prism mirrors. These are designed to be like “long bars” rather than short prisms, and this is necessary so that beam deflection in two dimensions can be done with ease. These are controlled by (brushless) DC motors. The fastest motor, for horizontal scan, rotates at 30,000 rpm. Additionally, the motors have some sort of “resolver” or rotation sensor that can effectively determine the timing of the “start” of a cycle. For your light source, you have an LED that can have its power modulated by the Raspberry Pi. There is a focusing lens and maybe also an aperture to concentrate the light to be more beam-like. The other alternative would be to use laser light directly, and although that is cheap to do, that makes the project feel a lot less hand-made, plus it is somewhat more risky with more safety implications.
The main caveats of this system:
-
Horizontal and vertical sync signals are not controlled by the video card, but rather entirely by the display device. The video card must adjust its display properties to the synchronization pulses generated by the raster scan motors.
-
The projection display itself is necessarily going to be somewhat small and dim. A narrow projection angle, which results in very little magnification with increasing distance, is a necessary artifact of the deflection mirror design, and the use of triangular prism mirrors rather than octagonal prism mirrors is a means of compensating for this, albeit with a great susceptibility to distortion. The brightness of the display, of course, is limited by the brightness and focusing power of the LED light system, and it decreases with increasing projection distance.
-
Because the system is mechanically controlled by motors, and a rather fast one for the horizontal sync, it is going to be rather noisy compared to pure electronic displays.
-
The resolution is going to be somewhat low, of course. No more than 320 X 240 is my recommended expectation, but you might be able to get up to 640 X 480 resolution. The horizontal resolution is also limited by the rate that the Raspberry Pi can toggle the GPIO pins, though some trickery with the UART or PWM can be used to effect a higher resolution.