A Cyclone Chip Pre-Filter for My Dust Collector©
Hal Mahon

This describes construction of a cyclone pre-filter for the 2300 cfm dust collector I purchased for $100 from a furniture factory going out of business. This machine had 4 thirty-micron dust bags. (Thirty-micron filters common on dust collectors pull in everything but then blow the dangerous sized dust out of the bags and all over the shop.) I replaced 2 of its bag filters with one-micron needle felted polyester filters and the two bags on top with 1 micron pleated 3 ft high cartridge filters. Its 3 hp motor draws 8 to 10 amperes at 220 volts in typical operation. With a remote transmitter On/Off controller, this dust collector is handy, effective and results in a significant reduction in shop sawdust. In Fig. 1 the pre-filter I will describe in this article is shown connected to my dust collector.

There was however, still a significant problem that had been endured by the furniture factory: Large wood chips accelerated by the fan were causing deformation of the collector’s metal duct work, as shown in Fig. 2. (The stream of large chips from woodturning may have made this a more serious problem for me than for the furniture factory.)

My method for removing large chips from the air stream is to lead them downward against the barrel wall. Chips continue in a straight line from the input pipe until acted upon by an external force—in this case force from the barrel wall. Sliding in a cyclonic path around the barrel’s circumference they are slowed by friction with the wall. As they slowed gravity moves them downward and out of the main air stream. Lighter dust particles are entrapped by turbulence in the air stream and carried to the outlet of the pre-filter.

There are two types of ready made tops for 30 gallon pre-filters sized for 4 inch hose, but none for 6 inch diameter hose. Here I describe how I made my own 6 inch hose cyclone pre-filter.

First I cut the top out of a 55-gallon drum, as shown in Fig.3, using an ordinary wood cutting blade with my jigsaw. The plastic barrel is of tough plastic and the going was slow, but no blades were lost and the final cut required little trimming with a rasp.

I chose a ¾ inch thick glued up panel of ash from which to cut a 2 ft circle for the top surface (see Fig. 4). The 2 ft circle overlaps the sides of the barrel by about ¾ inch. I cut 2 more circles of flake board about 3 inches smaller in diameter than the ash top and glued them to the center of the ash circle. The 6-inch diameter intake and output pipes were selected from, 20 gauge long radius 90o bends (available from PSI as #N-90E06; 1-800-377-7297). PSI is a good source for filter media, and for complete cyclone and other dust collectors at prices less than the better-known manufacturer in Syracuse, NY.

As my local hardware store wanted $90 for a 6-inch carbon steel hole saw, I first thought of cutting the slanted holes for the elbows using my jigsaw. The result of my test on scrap flake board is shown in Fig. 5, and although not bad it was not a tight fit. Then I found a 6” saw with bimetal teeth at 20% of the hardware store price (Timberline #602-188; Toolstoday.com, 1-888-699-3939). Cutting a hole for the intake elbow at an angle of 30o is shown in Figure 6. This hole saw can cut down a depth of about 1 inch before the saw needed to be raised and the core cut away with a chisel. Although slow, this process worked and the result was a cleanly cut hole for the 6” diameter pipe protruding through the ash and flake board at 30o. True, the hole cut with the jig saw looks like a good fit, but the opening drilled with the hole saw fits the tubing better.

Both holes, as completed, are shown in Fig. 7. The outtake pipe admits air near the upper surface at the center of the drum. Foam weather stripping, 1¼ inch wide and ½ inch thick was added to provide a good seal between the top of the barrel and the ash top of the pre-filter. Initially I thought clamps would be necessary to hold the top tightly to the barrel. In practice the weight of the top provides an excellent seal. This was helped by the choice of heavy ash for the top surface.

A 9” length of 20-gauge pipe was added to the intake-90o tubing at the point where it aims into the barrel. Three ¼ inch short sheet metal screws secure the two tubes together with minimal protrusion into the air stream. As shown in Fig. 8 the extension of the added tube below the lid has been nibbled away along its half diameter facing the side of the drum. This helps direct incoming chips against the wall while eliminating a direct path for chips to be carried out by turbulent up draft at the exit pipe. Centripetal force at the wall engages chips in cyclonic motion around the drum as the force of gravity pulls them to the bottom. All chips for which these two forces are greater than those from turbulent air deposit at the bottom of the barrel.

In practice air flow into the 6 inch diameter hose has picked up wood as large as 4 inches and the pre-filter appears to keep all but particles smaller than about 1/16th inch, i.e. a mm or so. The larger of these particles are sufficiently small that they cause no damage as they pass through the impeller to the second cyclones and drop to the bags below. 

Four ¼ inch thick angle brackets secure the tubes to the top. In addition epoxy fills most of the space between tubes and wood. Tight joints at all parts of the system aid effective dust collection. Caulking has been added to seal joints. Other fittings have been taped at their joints.

My nearly final system is shown in Fig. 9. Notice the orange wire connects metal components of the pre-filter with metal of the main dust collector. The building ground system of course is connected through building wiring to the metal of the dust collector. This serves to eliminate static build up of charge that could cause explosion of fine dust. There is spiral metal wire for ground continuity along the flex hoses used in my system.

A popular image of a cyclone dust filter has a conical shape extending below the fan with a container below the cone to collect dust. This style has an advantage of taking dust and chips out of the turbulence of the incoming air stream. In my design these larger chips used to rattle around at the bottom of the barrel creating nuisance noise. This noise was quieted by placing a 2 ft long two-by-twelve inside leaning against the side of the barrel.

Two methods I follow to enhance system air flow are—

1. Using the largest diameter tubing possible for distant runs, as for example to the drum sander and jointer. (Both 5 ft lengths of flex tubing shown in Fig.9 are actually 6½ inches in interior diameter.)
2. Minimizing curvature and bends in the tubing. Notice the impromptu support that reduces curvature in the tubing between the dust collector and pre-filter exhaust. I try to keep tubing runs as direct and compact as possible.

The manometer for monitoring the health of my filter is shown in Fig. 10. (This was added after previous Figures had been taken.) Its water column reading shows the level of vacuum inside my pre-filter. With 20 feet of 6 inch diameter hose connected to the table saw a preliminary measurement showed 18 cm water column (i.e. 7 inches) inside the pre-filter. The drop across the pre-filter with 5 ft of hose open at my lathe measured 6 cm, and measured 2.8 cm with 20 ft of 6” hose between pre-filter and table saw.

The level of vacuum inside the pre-filter may grow higher with improved configuration and tightening of the entire system. As the filters become clogged with dust the manometer reading will drop. Experience will tell at what level drop the filters need to be cleaned. Not yet installed is a pulley system to the ceiling above the pre-filter. This will be used to lift the top of the pre-filter clear of the barrel when it needs to be emptied.

As shown in Fig. 9 this dust collecting system is a joy to use. It is excellent at collecting nearly all sanding dust from work on my lathe into the opening of the 5 ft length of 6½” diameter hose (where there is 4 cm water column of vacuum an inch or so inside its opening), and into a 4” diameter hose from my drum sander. My set up does less well at collecting dust from my table saw, but I am working on that. It is poor at collecting dust from my JET band saw and from work on the drill press. I am also working on improving large chip collection while rough turning. At this point at least 50% of large chips go flying off into my shop. Because I use flexible hoses rather than fixed metal ducts (metal ducts do have up to 25% less resistance compared to the same length of flex hose), my system is more work to change from one machine to another. A system with fixed metal ducts around my shop and blast gates at each machine would definitely be more convenient. However, my airflows are greater because my hose lengths are significantly shorter, are more direct, and have fewer elbows with loss inflicting fittings.

A $1473 system with nominal 1000 cfm airflow from a well-known manufacturer in Syracuse, NY draws chunks of wood directly into the fan before dust enters its 2 cyclones. Its filters do trap 1-micron dust but they have only 10% of the area of my system, and it is expensive for what one gets. In comparison I’m pleased with the performance and the cost to assemble my two-stage duct collection system with its pre-filter for large chips.