Deflection basics:
What is deflection? Well, first, a little bit about the basics of CRT displays. The inside face of the picture tube is coated with phosphors in three colors, already discussed earlier when I talked about screen burn. Slightly behind the face of the tube is a metal grille, known as the shadow mask. It looks like a cheese grater - staggered tiny holes. You can't see it from the outside of the tube (you'd have to disassemble it to see it, and by disassemble, I mean "break with a hammer"), but trust me, it's there. Then, at the neck of the tube, is the electron gun assembly. This contains three separate electron guns, that are driven by the monitor electronics. The tube is filled with a vacuum (that is to say, not filled at all...).
So, the monitor electronics drive the electron guns to emit pulsed streams of electrons. There are three guns, each aligned with a particular set of colored phosphor dots. Electrons travel very nicely in a vacuum, and left to their own devices, leave the guns and travel directly straight ahead. This is great if you just want a tiny dot of light in exactly one place, but we'd much rather have them be able to light up the whole screen with a picture. That's where the deflection circuits come in. On the neck of the tube is a rather largeish coil of copper wires. This is the deflection coil, more commonly known as the yoke. The yoke actually contains two independent sets of coils - one for horizontal, and one for vertical. If you were to take the yoke apart, you could see the separation.
Now, electrons in a vacuum are affected by a magnetic field. They're deflected by it. So, by applying a signal to the separate coils of the yoke, the yoke creates magnetic fields that bend the stream of electrons and change the area of the face of the tube that they will hit. In a typical raster monitor, this beam of electrons deflected by the yoke and is scanned across the tube, from top to bottom, left to right, painting a picture as it goes. Imagine it like a REALLY fast printer. The electron beam is only touching one tiny point of the tube at a time, but since the phosphors stay lit for a while, and our eyes themselves have a certain persistence to them too, we see a stable, unmoving image. When in reality, only a tiny fraction of the screen is being written to at a time. This is why, when you record a video of a CRT display with a camera, you sometimes see this band of darkness waving through it. That's because the camera can capture things faster than the eye can, and it's catching the electron beam.
When the beam gets to the bottom of the screen, it's shut off, and the magnetic field is changed to bend it back up to the top. This takes a very tiny amount of time, but it's known as the "blanking interval". But, the beam isn't truly turned off all the way, it's still there, just with not enough energy to light anything up. But when you turn the SCREEN control on the flyback all the way up, it causes that "blanked" beam to be visible, resulting in the angled retrace lines on the screen.
What about vector monitors? Those are different, right?
Yes. They are. In a raster monitor, the deflection electronics are fixed at a particular pattern and frequency. They scan the tube exactly the same way, all the time, and synced up to the game board's signal via the sync line(s). Vector monitors, on the other hand, have no sync line. Instead, the deflection circuits are directly controlled by the game board. Vector games can manually bend that beam of electrons anywhere on the screen at any time, turn the beam on, then bend it someplace else, and shut it off again - allowing a perfectly straight line to be drawn from one point to another. It's the difference between coloring in squares on graph paper and drawing free-form. Vector monitors are thus different from raster monitors because they have these deflection amplifiers with direct inputs.
And how does convergence fit into all this?
Convergence is the alignment necessary to get the red, green and blue "beams" to converge at the same spot on the screen. Those three beams are separate, and need to be aligned so they work together. First adjustment is actually purity. Purity is the ability for a particular beam to ONLY hit it's proper color, and not "miss" and hit the wrong colored phosphors. Then, the convergence must be set up to align the beams with each other. Convergence is achieved by the way of tiny little magnets, mounted on movable rings, right between the electron gun and the deflection yoke. Basically, to "prebend" the beams of electrons, to get them lined up, before they're bent all together to scan the face of the tube. These tiny magnets can affect (mostly) only the one particular color since they're right next to the guns. But getting them aligned is a royal pain, and is outside the scope of this guide.
You said something about a shadow mask?
Yes. And I almost forgot. The shadow mask is basically a stencil. A piece of metal designed to ensure that the electron beams don't "spill over" onto the wrong dots on the screen, and only hit one at a time. This shadow mask sometimes gets slightly magnetized, however, and it bends the electron beams as they pass through, causing them to miss their mark! This causes a psychedelic colored picture. When this happens, the tube must be "degaussed", which is just a fancy way of saying "demagnetized". A picture tube is very sensitive to magnetic fields, including the Earth's magnetic field. Usually, when you move a monitor, it must be degaussed. Because this degaussing is so important, all color monitors and televisions have built-in degaussing coils. That's the fat (usually black, sometimes grey) cable that snakes around the back of the tube near the face. It looks like it was wrapped in lots of tape. It's another magnetic coil - and it gets fired for a second or so every time the monitor is turned on from being cold. That's the "thooom" sound that some monitors make on powerup. Sometimes, the shadow mask gets too magnetized for the weak little coil in the monitor to clear it up. Multiple cold powerup cycles can help it, but other times you need to degauss it manually with a handheld degaussing coil. Also, it IS possible for a strong magnetic field to permanently magnetize or even bend the shadow mask. Similarly, a tube that was dropped hard enough can bend or dislodge the mask. If this happens, there is no repair short of replacing the picture tube.
-Ian