Build a Vector Graphics Display Clock with a Cathode-Ray Tube
Illustration: James Provost CRT Chronometer: The Oscilloscope Clock Kit uses PIC microcontrollers to turn old cathode-ray tubes into clock displays. Once upon a time, there was a type of particle accelerator so popular that it was mass-produced by the million. Engineers and scientists at their benches, and folks at home in their living rooms, would carefully arrange themselves to watch the dancing glow of a beam of subatomic particles smashing into a phosphorescent screen. This attention-hogging accelerator was, of course, the cathode-ray tube (CRT), which reigned supreme as the electronic display technology for decades, before being unceremoniously relegated to the figurative and literal trash heap of history by flat-screen technologies.
But there are still CRTs to be found, and you can put some of them to great use with Howard Constantine's US $100 Oscilloscope Clock Kit. The kit works with many CRTs that were designed to be used in oscilloscope-type displays and operated with relatively low voltages-in the range of hundreds, rather than thousands, of volts.
As CRTs are becoming as unfamiliar to modern engineers as amplifier tubes did to the transistor generation, a quick recap of a few salient points is likely in order here. Oscilloscope-type CRTs are different from those found in televisions and most computer monitors. TV-type CRTs use magnetic fields generated by coils located outside the vacuum tube to deflect an electron beam, which is scanned line by line across the screen to build up what's called a raster image. Oscilloscope-type CRTs use two pairs of horizontally and vertically oriented plates located inside the tube to electrostatically deflect the beam: This approach was handy for oscilloscopes because an analog input voltage can control the vertical position of the beam directly (albeit filtered through some signal-conditioning circuitry), while an internal timing circuit controls the horizontal deflection, letting engineers see time-varying signals.
Illustration: James Provost Clock Components (clockwise from left):
1. The kit does not come with a CRT, but you can purchase one from sites like eBay. The kit will work with many different CRTs: You will need a lower voltage one designed for use in oscilloscopes. I built the kit using a DG7-6.
2. A transformer and rectifier create the roughly 300 volts DC that the CRT requires. The clock can work on both 110- and 220-V power supplies selected via a jumper and can autodetect the frequency of utility AC power to use as a time base.
3. The clock is driven by PIC microcontrollers and all the components are through-hole.
4. The printed circuit board has space for optional extras, such as a battery backup, and Wi-Fi and GPS modules for setting the time.
The beam's horizontal deflection can also be controlled with a second input voltage (called X-Y, or vector, mode). This made oscilloscope CRTs appealing to early computer-graphics pioneers, who pressed them into service as displays for things like radar defense networks. Some seminal computer games were made using vector displays, including arguably the first-ever video game, the 1958 Tennis for Two, and the massive 1979 arcade hit Asteroids. But vector displays struggled when it came to, say, showing bitmaps or even simply a large area filled with a solid color, and so eventually lost out to raster displays.
But CRT vector displays have a distinct look that's hard to replicate-not least when the screen is round, which was the easiest shape to make back in the day. (I do find it a little ironic that after decades of engineers striving to create the most perfectly flat and rectangular displays possible, smartphone makers have begun rhapsodizing about offering partially curved screens with rounded corners.)
Illustration: James Provost Subatomic Beam: In a CRT, electrons are boiled off the cathode and accelerated and focused before being steered by electrodes. The Oscilloscope Clock Kit allows you to recapture that look. The kit itself comprises the printed circuit board and all the accompanying components, including two PIC microcontrollers-one controls the clock and generates the graphics and text, while the other is dedicated to periodically shifting the clock around the screen to avoid phosphor burn-in. Normally you have to supply your own enclosure and CRT (eBay usually has listings), but as Constantine has a limited stock he uses to make fully assembled clocks for sale, he kindly let me buy a 7-centimeter-wide DG7-6 from him for $75 and one of his acrylic enclosures for $40.
Getting a clear enclosure was important to me because I wanted to be able to show off the tube for maximum effect, while also keeping fingers safely away from the electronics. This is important because even though the CRT is considered "low voltage," that still means 300 volts in some parts of the circuitry. Perhaps I was particularly skittish on this topic because of childhood memories: When I demonstrated a burgeoning propensity for taking things apart with a screwdriver, my father, a broadcast engineer, headed off any designs I might have had on our TV set with lurid true tales of people being zapped by the charge stored in a television's high-voltage capacitors even after the set had been unplugged.
Fortunately for my nerves, the clock kit's smaller capacitors pose much less of a hazard. Nonetheless, now that even 5 V is increasingly shunned as a circuit-melting torrent of electricity, I recommend builders work with a little more caution than they are used to, especially as checking stages of the build do require probing it with a multimeter when the board is plugged in.
Soldering the through-hole components was straightforward, although because the tallest component-a transformer-is one of the first required, it's likely you'll need to use some kind of circuit-board holder rather than trying to lay the board flat when working. The biggest obstacle came when it was time to wire up my DG7-6 CRT. The kit provides 10 leads for operating a CRT-two that supply a few volts of alternating power to the heater filament to raise the cathode temperature enough for thermoelectric emission to come into play, one that has a constant negative voltage of about 295 V to provide the cathode's supply of electrons, three that connect to a train of accelerating and focusing electrodes in the neck of the CRT, and four that connect to the horizontal and vertical deflecting plate pairs. But my DG7-6 only had nine pins! A check of the DG7-6's data sheet [PDF] (which I found on Frank Philipse's wonderful archive) showed that the cathode and one side of the heater filament shared a pin. A quick email to Constantine revealed the solution was a quick fix: All I had to do was jumper the cathode connection to one of the filament leads. After that, the instructions stepped me through the calibration steps required to produce a sharp bright test dot in the center of the screen rather than a fuzzy dim elliptical blob off to the side.
When building the kit, you can incorporate one of two optional $40 add-on circuits that eliminate the need to set the clock manually-a Wi-Fi module and a GPS module. Without one of those, the clock automatically detects whether it is plugged into a 50- or 60-hertz wall socket and uses that frequency as a reference time base. Setting the clock manually is simply a matter of pushing a "fast set" and "slow set" button until the clock shows the correct time.
The end result is an eye-catching timepiece that restores a CRT to its rightful place: the center of attention.
This article appears in the January 2020 print issue as "Oscilloscope Clock."