/diy/ - Do It Yourself

Going to sleep now, but someone tell me if they can see anything wrong with this method of driving a VFD.

I'm using UDN6118As for both grids and segments, which I'm pretty sure is fine.
More experimentally, I'm making out-of-phase ~40kHz 5V square waves, feeding them through an L293D, through some DC-blocking caps, and into the filament with only a pair of small resistors for regulation. Then I've built a clamp circuit with D1, D2, and Q2 that should keep the highest point of the square wave below a set-point. This should allow me to adjust the mid-point of the square wave (and hence the bias voltage of the filament) to anywhere between -4 and +8. I suspect a voltage of about +1V to +2V will be optimal, so I have plenty of room for it.

Actually looking closer, the typical total dropout of the L293D is 2.6V, so I'll be underpowering my filament. I'll work on duplicating the Q1/R4 circuit for the L293D.
>inb4 LM317
Too much dropout. I have some AMS1117-adj chips lying about, but since one of my 3.3V ones died I can't trust any of them.