I am trying to wrap my head around why so many self-proclaimed experts keep insisting that driving a VFD filament with DC is perfectly fine. It is physically impossible to avoid a brightness gradient on longer displays when you use DC, yet every other blog post suggests some convoluted resistor bias network to 'fix' it. The math simply does not support this. You are trying to offset a linear voltage drop with static grid bias, which completely falls apart the moment you start multiplexing and changing the active duty cycle. Just use a proper AC driver or a simple H-bridge. Why is this basic concept so fiercely debated?