Good enough solar cells, good enough membrane, good enough lightweight HHO, and a way to capture moisture from the air itself in the form on condensation or dew. It could be a small thing of water... whoa...
Build IT!!!
https://en.wikipedia.org/wiki/Super_hydrophilicity
https://en.wikipedia.org/wiki/Superhydrophobicity
It could just collect water into it, and using natural gravity it would accumulate @ the bottom where you could then, simply collect/filter it, and use it to run a HHO generator which who essentially fill the aircraft/dirigible, and power it, simultaneously...
Collecting water from the air itself...
https://en.wikipedia.org/wiki/Troposphere
Program the damn thing to track weather patterns, along with the usual air-pressure & humidity patterns.
You could essentially, fly through clouds to absorb the water, or just keep flying within specific patterns along with the humidity patterns of the EARTH.
If you wanted to do it the cheap hacker-way, and small scale...
https://www.youtube.com/watch?v=BY7uGJGvo9w&feature=rela...
http://hackaday.com/2007/12/27/24c3-build-your-own-uav/
https://en.wikipedia.org/wiki/Rain-X
A small jar HHO generator which would then feed a micro-engine(like in a R/C car) powering an electrical generator which would feed power to the HHO generator. This is already being done commercially.
Solar is a way better way then combustion in the end but, its the closest thing to a TYPE-2 civilization WE will see...
;-)
Wait, so you split water and burn it in a generator to generate the power you're using to split the water? I would love to see someone work out the math on that one... Everything I know about chemistry suggests that the best you'll ever do is 100% efficiency, which you of course cannot hit if you're not burning all of your H2 (which you wouldn't be, since the point is to get H2 to lift the whole thing).
"This is already being done commercially."
Do you have an example?