Yep, circuit design 101 (though still quite easy to forget). 
Here is the schematic, though the circuit isn't in front of me at the moment so there may be an error here. It's nothing too elaborate, basically the first half creates a pulse of variable width in response to a falling edge of HSYNC, and the second half triggers on the end of the first pulse. Adjusting the R1 changes the delay; adjusting R2 changes the time spent low (which is set to match the original HSYNC).
It was all a bit of a crude breadboard hack just to see if it would work. It's powered from an external 5V supply, a multi-output unit which is also powering the yoke circuit. The grounds are all shared, including to the SE/30 analog board near the point where the HSYNC signal is taken. In fact a possible culprit is ground loops via earth, which comes separately off the SE/30, the power supply and the oscilloscope. Anyway, now that the concept basically works, I need to clean it up and it shouldn't be hard to squash that particular problem.
The other use for a circuit like this is that later on, when taking other resolution signals, this lets me adjust the retrace time independently of the HSYNC signal. At the very least, I need to make sure the retrace period isn't too short, which could melt the HOT if it switches on too early.
Here is the schematic, though the circuit isn't in front of me at the moment so there may be an error here. It's nothing too elaborate, basically the first half creates a pulse of variable width in response to a falling edge of HSYNC, and the second half triggers on the end of the first pulse. Adjusting the R1 changes the delay; adjusting R2 changes the time spent low (which is set to match the original HSYNC).It was all a bit of a crude breadboard hack just to see if it would work. It's powered from an external 5V supply, a multi-output unit which is also powering the yoke circuit. The grounds are all shared, including to the SE/30 analog board near the point where the HSYNC signal is taken. In fact a possible culprit is ground loops via earth, which comes separately off the SE/30, the power supply and the oscilloscope. Anyway, now that the concept basically works, I need to clean it up and it shouldn't be hard to squash that particular problem.
The other use for a circuit like this is that later on, when taking other resolution signals, this lets me adjust the retrace time independently of the HSYNC signal. At the very least, I need to make sure the retrace period isn't too short, which could melt the HOT if it switches on too early.