In a lifetime of working with electronics we see a lot of technologies arrive, become mighty, then disappear as though they had never been. The germanium transistor for instance, thermionic valves (“tubes”), helical-scan video tape, or the CRT display. Along the way we pick up a trove of general knowledge and special skills associated with working on the devices, which become redundant once the world has moved on, and are suitable only reminiscing about times gone by.
When I think about my now-redundant special skills, there is one that comes to the fore through both the complexity and skill required, and its complete irrelevance today. I’m talking about convergence of the delta-gun shadow mask colour CRTs that were the height of television technology until the 1970s, and which were still readily available for tinkering purposes by a teenager in the 1980s.
A shadow mask colour CRT has three electron guns, one for red, one for green, and one for blue. There is a grid of red, green, and blue phosphor dots on its screen, with a metal plate full of holes – the mask – aligned just behind it such that each colour phosphor can only be seen by its corresponding electron gun. As the three electron beams are traced in a raster across the screen, their relative intensities can synthesise any visible colour for your eye.
The problem facing the designer of a shadow-mask CRT is that of aligning electron guns, mask, and phosphor dots so that the alignment is perfect at all points on a screen which is not equidistant from the guns at all points. If you supply a poorly aligned CRT with a picture that is entirely white it would deliver a picture mostly white but with spots of colour, and when a TV picture is shown then it would feature magenta, cyan, and yellow shadows around its subjects.
Shadow mask CRTs developed from the mid-1960s onwards solved this problem mechanically, by clever shadow mask designs, different phosphor dot patterns, and careful alignment of all the components with the three electron guns in a line across the neck of the tube. Earlier tubes had their electron guns mounted in a triangular delta in the neck of the tube, and relied on complex arrangements of electronics, magnets, and electromagnets to actively correct the misalignment as the beams traversed the screen. It is the skill of making these adjustments that is the subject of this piece.
Broadcast colour TV has its roots in the years following the Second World War, with the first commercial sets both being American 525-line NTSC, the Westinghouse H840CK15 and RCA CT-100 from early 1954. (I should be using Noah Webster’s spelling color in those cases.) Earlier sets had round CRTs, but by the time we Brits got our 625-line PAL colour TV in the late 1960s the world had moved on to the curved-edge rectangular screens you will probably be familiar with.
To illustrate this piece I pulled out my last remaining delta-gun TV from storage. It’s an ITT CVC5 that came into my hands some time in the 1990s, it was made in 1972 and I hung on to it because I guessed it would be the last delta-gun that would come my way. It’s been in the wars – I should have held on to one of the many far nicer examples I ripped apart in the ’80s – but I got it working at the time with some repairs to a cracked PCB and replacement of a blown rectifier. Sadly it has now lost its field oscillator, so there’s a repair job for a future idle time.
Opening up the CVC5, you start to understand why these sets were so expensive when they were launched. It’s got a Mullard delta-gun CRT, a chassis that has a couple of square feet of resin-bonded-paper PCB densely covered in discrete components and tubes hinges down in front of you, a chunky PAL delay line and a large metal enclosure for the 25 KV EHT circuitry. It’s a live-chassis design, so extreme care is the order of the day if it’s connected. On the neck of the CRT is the deflection assembly, and behind it the convergence yoke attached by an umbilical cable to a large plastic convergence box containing a bank of pots and variable inductors. These and the adjustable magnets on the convergence yoke form the basis of the convergence adjustment.
The convergence yoke takes the form of an inverted Y (a ⅄) of coil and magnet assemblies, which put a magnetic field in the path of each of the electron beams when correctly placed. Each arm of the ⅄ has a rotatable magnet which moves the corresponding beam along the axis of the arm. On top is blue, bottom right is red, bottom left is green. There is also another adjustable magnet clamped to the top of the tube neck just behind the convergence yoke. This has the effect of moving the blue beam from side to side.
As can be seen from the picture, the blue electron beam can be adjusted in two axes while the other two only have one axis. To perform basic adjustment of convergence you turn off the blue gun and move the blue and green to the point at which they align, then turn the blue back on and move it as needed in both directions until it meets the other two. Ideally this should be done with a test pattern generator, but as a teenager such a device was far beyond my reach and I used Testcard F. Thank you BBC 2, for transmitting that!
Of course, with basic convergence done, you discover just how lousy the convergence of a delta-gun CRT is. You will still have coloured shadows on almost everything around the edges of the screen, each of the three rasters may be aligned in the centre but they still all have different shapes. It’s time to reach for the electronic convergence panel, and enter the black art of adjusting all those pots and inductors to drag the three images into line.
The convergence panel is always designed to be accessible from the front of the set, you need to have a good view of the screen to use it. On some sets it’s accessible behind a panel on the front, on others it’s a PCB that hinges up, but on the CVC5 it’s a plastic box that lives in a slot inside the case and can be brought out on an umbilical cable.
On the front are the array of convergence controls, each helpfully labeled by colour and area of the screen. There are also adjustments for the beam intensity to set up the grey-scale performance, and switches to turn off both blue and green guns for adjustments that don’t need them.
My reaction on looking at a convergence panel for the first time in years is that maybe this is a skill I *used* to have and have now lost. But in reality the controls are logically arranged, and rather than going crazy twiddling pots the technique is to make small adjustments in response to convergence problems on the screen. There is almost certainly a recommended progression, but if I had it in the first place I’ve lost it somewhere in the last twenty years. I do remember learning how to do this by trial and error on an endless succession of scrap sets back in the 1980s, first getting it spectacularly wrong and then progressively getting better at the job.
So there you have it, complex adjustments on a very old version of a now completely redundant technology — my most obsolete skill. What’s yours?
It’s worth taking a moment to delve into retrotechtacular territory and look at some other shots of the CVC5. Some close-ups of the colour decoder and timebase circuitry, and the high-voltage enclosure. The PL509 line-output tube, flyback transformer and EHT tripler. My personal favourite though is the focus pot, 5 KV across an open rod resistor, with the wiper adjusted by a long thin piece of fibreglass PCB material through a slot in a live chassis. Seems safe to me.