Flying Spot Scanner

This is my Nipkow disc-based flying spot scanner, the construction of which was mostly detailed on this post in the NBTVA forum. At the moment it uses the club's 15 inch 32-line Nipkow disc, but a 30-line disc will also be made for it.

The whole thing is built around an Argus 300 slide projector which, as it's name suggests, uses a 300W lamp. The projector accepted 2 1/4" slides, so enlargement of the gate was not necessary for a disc larger than 12 inches. For the necessary disc clearance, the projector sits on 5 1/4" stilts. This also conveniently makes some extra room for the DC power supply.

Front of the machine with the front and top cover removed. The grey box holds the lamp rectifier.

From the beginning I knew I wanted to enclose the entire thing in a case, both to block stray light and also to hide the fact that it is made mostly of scavenged equipment. One of my concerns was that the projection lamp running on AC current would cause a noticeable ripple in the light output, so a rectifier was built. Stepping up the voltage from 120V RMS AC to a smoothed 150V DC also meant a brighter and hotter bulb, so I made room for extra cooling. The projector's rather puny (and noisy) built-in cooling fan was replaced with a 120mm mains-powered muffin fan on the projectors' underside and an extra 60mm AC exhaust fan was screwed to the wood case. A 90° plastic elbow originally for a gutter downspout made a perfect snorkel to route the exhaust out the back and block much of the light leaking from the vents.

The rear switch panel

In the end I found that the filament's thermal inertia was more than enough to overcome the periodic nature of the AC current supply, so the rectifier was abandoned. (On a side note, the lamp and rectifier were originally wired in series through the lamp switch to the main power rail. However, when switched on, the combined inrush current overwhelmed the little 5A-rated mini toggle switch and would partially weld the contacts together, creating a delay and strange hissing sound when the switch was opened and the contacts tried to separate. To get around that, I wired the rectifier in with the main power switch so it was always energized when the power was on, and the lamp stayed on its own dedicated switch.)

Due to a design error on my part, upon test I realized that the projected raster was reversed! I had failed to take into account the reversing action of the projector lens. Getting the lines to scan in the right direction was easily solved by mounting the disc on backwards, however the orientation of the arc-scan was still incorrect. That is, the raster appeared as if it was being viewed on the left side of the disc instead of the right. The solution was to add a mirror to reflect the beam 90°. This was accomplished by mounting an old chromed motorcycle mirror in front of the lens. The ball joint on the mirror would allow me to pan as well as tilt the light beam, or so I thought.

The old motorcycle mirror in position...

...and the new mirror door.

The added mirror worked perfectly as long as the beam was only reflected horizontally. Even the smallest vertical deflection caused the raster to tilt left or right, with the effect that reflecting the beam straight up rotated it a full 90 degrees! Not exactly what I had in mind. I soon swapped the motorcycle mirror for a metal door with a first-surface mirror glued to the inside that could also be closed to protect the lens when not it use. Bingo! 

One of the major drawbacks of enclosing this particular machine was that focusing became difficult. With this projector, there is no external focusing knob. Instead, the entire lens barrel must be rotated by hand or slid in and out manually. Putting this inside a box meant that I would have to reach inside from the front to get to the lens, blocking it during adjustment.

The Neon Lamp

The Electronics

When this scanner was conceived, I already had the framework plans for my Resurrecting 30-line TV project (article to come soon.) This machine was to serve as the caption scanner. Therefore, it needed to accept an external synchronizing pulse as well as generate it's own for standalone use. To be compatible with my project, the scanner had to use the club motor controller (instead of a synchronous motor or other speed control method) so the Peter Smith motor control circuit was built on perfboard. I also needed a timing generator to feed the motor controller. I had bought the club's Triple Waveform Generator circuit board and assembled it the year before, but never mounted it in a case and with no future plans for it, I used that instead of building something from scratch to save time. A sync separator circuit was also built next to the motor controller, to isolate the sync pulse from the club-standard video output of the generator. I could have picked the sync pulse directly off of the TWG before it was combined with the video signal, but using this arrangement meant that an external signal could not only be a TTL signal, but also be in the form of a "black burst". That is, a black picture with sync pulses. 

As a future project, I want to replace the incandescent bulb with an ultra-high brightness LED of the type used for outdoor or studio illumination. If it works, it would dramatically cut down on the power consumed and the heat produced. (I also wouldn't have to worry about finding a replacement bulb when it eventually burns out, although I do have a spare.) The other reason for this modification was that if the LED could be modulated, I would then have a Nipkow disc-based television projector, which to my knowledge hasn't been demonstrated before.

The TWG would also be able to provide a few simple test signals to set up and demonstrate this projector function. To take advantage of this feature I bought a six-way rotating wafer switch to select between the gray levels, checkerboard, pulse and bar, white field and black field patterns (all with internally generated sync pulse) as well as an external sync or video input via an RCA plug. However, after everything was installed and working properly, I ended up needing the TWG for another part of the project and replaced it with a scratch-built sync generator anyway, leaving the wiring and wafer switch disconnected.

I soon found out that when the unit was running, the sound of the fans drowned out the familiar audible feedback noise from the motor as it approaches sync speed, so some sort of timing light was called for to ensure the projector's disc was locked in. A 555 timer in monostable mode was made to drive an LED pointed at the disc inside the case. The sync holes appear stationary when the disc speed matches the pulse rate of the timing generator.

Timing LED with sync holes visible as shadows on the disc.

The framing slide as seen from the front.

Since the motor controller had to run from an externally generated timing signal, a method of getting the disc to be in phase with the signal was needed. A simple momentary “sync key” push button to disconnect the drive motor allowed for line slipping and rough image phasing. To ensure that the sync pulse was placed in the right part of the image during the line blanking interval, the optical fork that measured the disc's speed was mounted on a wooden slide. A threaded rod connects the slide to a knob on the outside of the case. 

The DC power supply is a switching type converter from an old laptop computer. It provides about 19 volts to a regulator on the circuit board and is also fed directly to the motor. The motor is a small transport motor from a VCR like the ones I have used in previous monitors. I was concerned about forcing it to drive such a large disc, but it seems to handle the task well, although it takes nearly thirty seconds to get up to full speed.

Last updated November 2020