The inverted downdraft gasifier, invented by Reed and Larson of Golden, Colorado, and made backpack-friendly by Ray Garlington. He offers a nice discussion of the design features here: http://www.garlington.biz/Ray/WoodGasStove/

I've been experimenting with this stove because i want to get away from fossil fuels entirely. I'm using a bigger, 4" diameter can that fits inside my 800ml Optimus Terra pot.

I'm getting close to a final "product", but am not there yet. However, I made a discovery. It is possible to increase the window for gasification of the wood, in other words the time that it happens can be lengthened indefinitely.

Read Garlington's discussion, it's concise. I will assume you've read it, so you will understand what I am about to say.

After about one minute, with the batch-loaded stove being lit afire from the top, gasification is operating. Blue and yellow flames appear to be coming out of the secondary air holes and there is no smoke. This little stove burns wood FAST, and we are using twigs, basically, so gasification is completed anywhere from 4-10 minutes after starting the batch. What I discovered is, gasification can be continued by chucking a piece of twig into the can every few seconds, before the flames stop. Instead of the "fire" going out, the flames will continue in gasifier mode as long as wood is added. Finding a rythym is easy, by watching the results 60 seconds later. It takes about 60 seconds for twigs put in the stove to affect the flames. They don't burn, they emit gas which rises, mixes with hot secondary air, and erupts into flames.

The stove's batch of fuel is added per Garlington's instructions, and it stops below the level of the secondary air inlets. (When this hot secondary air hits the wood gas, it instantly burns, thus the appearance of flames coming out of the secondary air slots/holes.) Tossing twigs through the feed slot of the stove keeps new wood in the gasification stage. As soon as the wood hits the hot center, gas starts coming off it. I was watching my twigs closely, and as they landed in the gassy zone, they looked like wooly caterpillars: little fuzzy-haired jets of gas immediately started coming out of the twigs. It's cool looking. It's not often I see a physical phenomenon displayed so graphically. Toss in the twig, gases come out, but not flames. The flames are above the pile of wood. The wood may be at the lower half of the stove's loading area, because it settles as it burns away, and the flames are at the top, above the secondary inlets. So it's a smoke-free fire.

I need to run more experiments with batch size, and I may want to go to the 3" diameter can, just to see if it's better. Also, my windscreen is a little large. I may make a "clip-on" screen, which would be like an extension of my added-on secondary wall. Maybe 3 slots for 3 legs of the windscreen. I'll do some pictures here. Directing the flames to the pot with the windscreen helps heating efficiency dramatically, but it makes for smoky puffs.

I can only add 5 photos per post, so I'll do more posts until all photos are up.