Michael,
It wouldn't seem to be too critical since you're switching DC and not trying to construct a sine wave, worry about harmonic content, prevent inverter switching overlap and desats, Etc. I would think you would find a compromise based on ripple amplitude and noise considerations. I would definitely use an RC snubber across the IGBT. Two 555s or a 556 could be used for an astable MV feeding a PWM stage by way of the control voltage pin. There are several one-shot chips that you could use also. Building the driver from scratch would be the biggest hassle because of the need to include the isolated dual power supply referenced to the IGBT emitter. Now if I were playing with this in our lab, I would just use one of our half-blown dual driver boards to feed the half-blown IGBT. I would just need to provide a +12 Volt source and a TTL level drive signal. Of course, my technical curiosity and interest in your challenge couldn't justify my actually rigging up these things in our lab.
Am I right in thinking that there are times that you would wish to gate this thing totally on or off? What about load side regulation?
Joe
