A few questions - What size of particle are you looking for? Is your product cheap to produce, or do you have it in quantity and would you be willing to use some to test your construsted device? Are you willing to accept a hair-brained solution from someone who has never used one of these machines let alone constructed a cheap knock-off?
I looked around on the engineering and do it yourself forums I'm on and noone has ever made one, though there have been several questions regarding them.
The wikipedia has a diagram of how they operate, to me the schematics look like this thing is just a glorified paint sprayer with a hairdryer attached behind it. Depending, of course, on what exactly your product is and if it reacts with polyethylene, PVC, steel, or rubber, this scheme just might work.
Here's how I see it. You'd want a hairdryer or two, full size (better yet, a heat gun used for shring wrapping) a quantity of the foil lined tubing used for dryer exhaust (to direct airflow) and some foil duct insulation tape. Cut a hole a few feet back on the tube and mount the nozzle end of a paint sprayer pointing in the direction of airflow. Your cyclone chamber separator is just a 5 gallon bucket. The airflow comes in from the side towards the top, at a tangent to the side of the bucket. Put a length of PVC pipe straight down through the bucket lid so it ends a third of the way from the bottom. Slap a filter with a rubber band around the inside end of the pipe , and feed the hot exhaust back in to your intake. Perhaps, since it is a closed system, adding a dessicant somewhere along the way to dry the air even more might help. Product goes in the paint sprayer reservoir, and with a whole lot of fairy dust, it might work.
Explanation too poorly written for you? I'll draw diagrams in the morning. Maybe.