[old_tv_nut]

I believe I've seen some attempts - have to go searching. I believe it's difficult to get the bandwidth and sensitivity you need at a reasonable power level with magnetic deflection [Edit: using the TV deflection yoke, that is].

The horizontal winding is designed to be mainly inductive at 16 kHz, so you can drive it with a switch (horizontal output tube or transistor) applying either full B+ supply voltage or open - the inductance then creates the current ramp to scan the beam from center to right, where the output device switches off. The inductance then resonates for a half cycle and reverses current through the damper diode to scan from left to center. Thus, the driver device and damper diode have relatively little dissipation during the fraction of a microsecond while they are switching from on to off. To produce a vector drive, you will need a class A driver with tremendous dissipation.

The vertical winding is normally operating with drive that is partially or mostly resistive and partially inductive with a class A drive, but develops some inductive kick. That's at 60 Hz. To drive it faster for vector use, the inductive effect will become much more prominent.

I'm not saying this is not a worthwhile project, but its major worth may be successfully doing the design calculations to see if it's paractical or not before you pick up a soldering iron.

Edit: above discussion applies to trying to use the existing deflection components. I'm not familiar with the yokes used in vector CRT games, but if you could obtain one or somehow get the specs, you woud have a better chance.