[kfogel]

That part about "...you wouldn’t want to wing it with the configuration, because allegedly you could break your monitor with a bad Monitor setting" -- strike the "allegedly"! Or at least, let me allege it from personal experience: I did that to one monitor, in the early 1990s. You could smell the fried electronics from across the room.

[immibis]

For the interested: CRT monitors have a high-voltage power supply which uses an oscillator. Cheap(er) monitors allegedly reused the horizontal sync frequency for the power supply oscillation, to save an oscillator, so if the horizontal sync frequency was very different from expected, or worse, completely stopped, it could burn out the HV power supply.

Has anyone tested this hypothesis? It could also be that the horizontal sync itself burns out, although that seems less likely.

(In even more detail: Like any other switching power supply, the HV supply in a CRT runs on a two-phase cycle: first, a coil, which creates electrical inertia, is connected to the power source, allowing current to build up. Then the current is suddenly shut off, and the force of the coil attempting to keep it flowing creates a very high voltage, which is harvested. If the circuit gets stuck in phase one, the current never stops increasing, until it's limited by the circuit's resistance, much higher than it's supposed to be. The excessively high current overheats and burns out the switching component. Anyone working on switching power supplies will have encountered this failure mode many times.)

[dfox]

It is not really about saving one oscillator, but about two things:

- saving the drive circuitry for the flyback, which is usually combined with horizontal deflection amplifier. Also such a design probably simplifies the output filter for horizontal deflection as the flyback primary is part of that filter.

- synchronizing the main PSU of the display to the horizontal sync as to make various interference artifacts in the image stay in place instead of slowly vandering around, which will make them more distracting and noticable.

It is not that hard to see the whole CRT monitor as essentially being one giant SMPS that produces bunch of really weird voltages with weird waveforms. And in fact is you take apart mid-90's CRT display (without OSD), the actual video circuitry is one IC, few passives and lot of overvoltage protection, rest of the thing is powersupply and the deflection drivers (which are kind of also an power supply, as the required currents are significant).

[disqard]

Your (parenthesized) explanation of switching power supplies made a lot of "secondhand knowledge" click in my head -- like, for instance, why there's lots of high-frequency noise in the DC output. Thank you!

[meatmanek]

I was briefly pleased with the ability to run an 8" monitor that looked like the kind on 90s cash registers at the impressively high resolution of 1024x768. Then after about 10 seconds it blinked out, smelled like burning electronics, and never worked again.

[satori99]

Neal Stephenson's Cryptonomicon made reference to a hacker dubbed The Digi-Bomber, as he could make his victims CRT monitors implode in front of them by remotely forcing a dangerously bad configuration.

[hawski]

That reminds me of using a CRT monitor to broadcast audio through radio waves: http://www.erikyyy.de/tempest/