When I started making a cyclone filter to attach to my shop vac, I imagined it would turn out something like this:

Or at least this:

Instead, I got this:

Things started going wrong almost immediately, as noted in my first post on the topic, and I discovered several ways NOT to bend and cut acrylic (valuable lessons) and so this project degenerated into an experiment: how badly could a cyclone be built and still work?

Continuing where my last post left off, I had several things to do before my cyclone was complete:

1. add a top
2. reinforce the seams of the funnel
3. install intake and outflow tubes
4. attach the cyclone to a bucket lid

I made the top out of a piece of 1/2″ plywood that I marked according the to shape of the top of the funnel.

Since the funnel was not perfectly round, I found the center and drew a circle that was approximately the average diameter of the funnel and cut out the circle with my scroll saw (the bandsaw would have end the baby’s nap).

Next I used a torch to soften the plastic rim of the funnel enough to force the lid into the end and mold the plastic around the lid to hold it in place.

Everything was sealed up with Liquid Nails polyurethane glue, which fills gaps nicely and drys very hard overnight.

I then started reinforcing the seams by melting a steel mesh into the plastic across each joint.

This seemed to be working reasonably well, but just to be safe, I also smeared glue all over the joints to hold the mesh in place.

Using a holesaw, I drilled out enough of the top to insert a 1 1/2″ ABS pipe for the suction end of the cyclone, which I then glued in place.

For the intake, I cut another piece of pipe at a 45-degree angle (approximately) and glued it in.

On both pipes, I used masking tape to hold them in place while the glue dried.

Attaching the cyclone to the bucket lid required a small amount of creativity. If I just cut a hole in the bucket lid, I would have very little gluing surface, due to the bucket lid being very thin.

My solution was to cut a smaller than necessary hole in the bucket lid, and then to heat up the lid with a torch to soften the area around the hole, and then to force the cyclone assembly into the lid.

The result is a perfect sized hole in the lid that tapers along the cyclone and provides a large surface to join the two pieces.

Once everything had dried, I did a little more cleanup and reinforcement (neither of which helped the cosmetics of the cyclone), and hooked everything up for a test run.

I took the end of the hose to a pile of sawdust and watched in amusement as the dust flew into the cyclone, swirled madly, and went straight out the exhaust tube and into the shop vac.

I had successfully created a clear chamber in which to observe dust and trash along its path to the shop vac.

And then I noticed something important: I had hooked up the hoses backwards.

I switched the shop vac to the top, and the accessory hose to the side and tried again.

Success!

I watched in fascination as the sawdust, metal shavings, plastic shards, and welding slag flew into the cyclone  chamber and promptly spiraled downwards into the bucket below sending practically clean air out to the shop vac.

And thus the conclusion of my experiment was reached: even ugly cyclones can be effective. Don’t over-engineer it unless you have a lot of spare time and nothing else to work on.

Just make sure to cover the following criteria:

1. The inlet needs to come in at an angle that sends air and dust along the outer wall of the funnel
2. The outlet needs to be long enough to end somewhere below the inlet (this causes the incoming air to spiral downwards, instead of just getting dizzy in the middle)
3. The funnel should be shaped like a funnel (round and tapering downwards)
4. The seams and connections need to be air tight

If you build a cyclone that meets these requirements, it should work, even if it looks like a potato battery from a third-grade science fair… like mine.