The deflection circuitry of a TV or raster-scan CRT monitor is not designed to be driven by random X-Y signals and optimized around the sawtooth patterns of the

The deflection circuitry of a TV or raster-scan CRT monitor is not designed to be driven by random X-Y signals and optimized around the sawtooth patterns of the raster scan. One thing is that driving the deflection coils of large-ish CRT at typical raster scan frequencies requires considerable power (the typical X-Y display cannot draw all the "pixels" each "frame") and another is that there is a bunch of analog tricks (funky magnetics in the "output filters" and such) where various non-linearities of the system partially cancel each other out, which works only for the typical saw-tooth-ish deflection waveforms. Another thing is that both CRT monitors and "modern" CRT TVs synchronize both their main SMPS and the EHT flyback to horizontal frequency of displayed image (one is tempted to say that typical CRT computer monitor is in fact one giant overcomplicated SMPS). The reasoning is more or less the same as to why the TV framerate is related to mains frequency: reduction of artifacts that move around in the image. Also, this is the reason why out-of-spec signal can damage (old, which essentially means not microprocessor controlled) CRT displays and TVs.