Homemade Cyclone Dust Collection System

by Steve Silca

Part 1: Introduction & Project History

My dust collection adventure began with a single stage 1 HP Reliant brand from Trendlines in December of 1995. As the hobby, the tool collection, and the shop area grew (now 15’x27’ – about 400 square feet), this did not suffice in terms of suction or in terms of safety. I was spending more and more time exposed to the dust that flows right through the cheap filter bags (see Bill’s site for more information on this).

Initially, I had planned to use the old blower and motor in conjunction with a drop box in the shop. Then, I pondered the possibility of exhausting the dust collector outside and not worrying about filtering. This, however, is known to create a very dangerous situation, as fumes from the furnace and water heater would flow directly into the shop, particularly in the winter when I can’t comfortably open a window to help replace the air in the shop.

So, it was clear to me that I needed to set a system up that was completely contained within the shop. It was also clear that, because of their small size, the 1 HP motor and 10” diameter fan were not going to do me any good for the kind of system I envisioned (see Bill's site for more information on these size requirements as well).

Bill Pentz convinced me to construct a cyclone, as intimidating as that sounded. This website thoroughly explains how a cyclone separator, in conjunction with a home-made blower and filter set-up, effectively collects woodshop dust. However, for those unfamiliar with this topic, the diagram at the side should help you understand exactly what happens in this system. The blower, mounted on top of the cyclone, consists of a motor, fan, inlet and outlet. As the motor turns the fan, suction is created at the inlet, which is connected to the cyclone outlet pipe (4). Dust and chips from the network of ducts throughout the shop enter through the cyclone inlet (1) and are forced to circulate around and around in the upper cylinder of the cyclone. As the air approaches the funnel shaped cone of the cyclone (2), dust and chips continue to be spun against the sides and eventually settle down in to the collection bin (3). The remaining air and very fine dust are sucked out of the cyclone through the cyclone outlet pipe (4), into the blower, and blown out the blower outlet toward a filter. This filter captures the very fine dust (which are now known carcinogens), and allows perfectly clean air to be exhausted back into the shop. Many websites go into greater detail on this topic.

A month and a half later my cyclone was complete, the shop was re-organized, and the ducting was being finalized. This is the story of that process. Hopefully, it can help others decide what they want to do with their dust collection system. I know that I would not have endeavored upon this project if I hadn’t known that there were many others who had successfully completed the project and were available to help me along the way. And help me they did! Through dozens and dozens of emails with fellow woodworkers, I was able to put together a powerful dust collection system that makes my shop cleaner and my household a safer place to be. Hopefully, this information can help the next woodworker who has these same goals in mind.

Part 2: Blower

The blower is powered by a Harbor Freight 5 HP compressor duty motor propelling a 14” impeller from Jet Tools (now called WMH Tools). Please look at the Cost Analysis for the prices of these items. Although it’s a 5 HP motor, it draws 15 amps, so it is really much closer to a 3 HP motor. I had my friend who’s a machinist make a custom bushing to make the hole in the impeller (which is some metric figure around 7/8”) adapt to the 5/8” arbor coming out of the motor. He machined the bushing, and drilled and tapped four holes in it which accept set screws that affix the bushing to the motor arbor. Then, he took the flange that came with the impeller and drilled and tapped two holes 90 degrees apart in it. These holes accept set screws that affix the impeller to the bushing. The bushing was milled with two flats 90 degrees apart against which these two set screws are tightened. These flats make the assembly extra secure. All the set screws were installed with lock-tite for security as well.

WARNING: After about a month of use, I was turning on the dust collector one night to show it to a friend. Right when I turned it on, there was a screeching sound and a clatter in the blower housing. I knew what had happened, but just to be sure, I disconnected the blower outlet from the hose that goes to the filter housing. Sure enough, the impeller had slipped off the bushing, and was resting on the bottom of the blower housing. Luckily, no damage had been done to any of the components, since the impeller didn’t fall off when it was spinning at 3450 RPM, rather when the motor was just being turned on. I was actually somewhat glad this happened, because I had been looking for a reason to take the whole cyclone down and re-mount it to the cement wall. The first installation in the rafters of the floor above was causing major vibration trouble (see the Installation page for further discussion on this). Taking the cyclone down was a little depressing, but I was thrilled that I had taken Bill's advice and designed the blower so that the housing could be easily removed from the motor/impeller assembly (see discussion in the last paragraph of this page). Otherwise, I would have had to completely rip apart the entire blower housing to re-affix the impeller to the motor.

Upon closer inspection of the impeller, my machinist friend who had made the bushing discovered what he thought was the problem. He said that the two set screws which hold the impeller against the bushing had worn away because they were stainless, which apparently makes them softer. So, we replaced those with two hardened steel set screws. This was fine, but I did not want to go through the process of taking this dust collector apart again, so we elected to take one more step. We put the impeller on the bushing exactly where it should be when it’s installed, and drilled a hole through the flange on the impeller and through the bushing – all the way through both. The bushing is longer than the motor arbor, so we did this at a point beyond the arbor so we wouldn’t be drilling through that. This hole accepted a roll pin that goes through the impeller flange and the bushing. Now, with the better set screws set in with lock-tite and the roll pin, I am 100% confident that the impeller will not fall off again.

The picture at the side shows the blower components: the motor with the bushing attached to the arbor, the impeller with the two threaded holes in the flange to accept set screws to tighten against the bushing, the motor mounting board with the rabbeted hole that accepts the motor arbor (this rabbet received a bead of caulk when the motor was attached to seal that connection), and the piece of ½” angle iron and bolts that I used to attach the motor to the board.

The bolts that hold the 1/2" angle iron to the motor mounting board had to be recessed because the blower housing gets pulled up tight against this board (the picture at the side is of the bottom of the motor mounting board and shows how these four bolts are recessed as well as the four bolts and fender washers that hold the motor mounting board to its hangers).

Blower outlet sealed with epoxy. Note 14 1/2" access hole on top fits over impeller, and is caulked where the sheet metal spiral meets the MDF

The blower was constructed a la Bill Pentz airfoil blower plan. The materials are ¾” MDF and 20 gauge galvanized sheet metal (I was fortunate to find a 7” wide scrap at the sheet metal supplier which I cut down to about 5 ¾” to make this blower). I cut an 8” diameter hole on the bottom of the housing for the inlet to the blower, because the center pipe in the middle of my cyclone was going to be a piece of 8” snap-lock. On the other side of the blower housing is a hole with approximately a 14 ½” diameter. This is the case because the blower housing is designed to first be attached to the top of the cyclone, and then slip over the impeller, which is already affixed to the motor. The outlet of the blower is 5”x7” and designed to have an area slightly larger than that of a 6” circle, since that is the size hose that would convey the air and fine dust to the filter. I used a basic 6” HVAC take-off to adapt the rectangular output of the blower to the 6” flex hose. The take-off was sealed with epoxy, and the groove where the MDF receives the sheet metal spiral was sealed along the outside with a caulk gun.

Part 3: Cyclone

I have no experience with soldering, welding, or metalwork. So, reading Bill's site and the various links, I determined that I was not going to solder this beast. They spoke of enormous torches and irons that I did not want to get involved with. A friend of mine who works with metal and was familiar with the cyclone style dust collector that I was trying to create suggested that I use pop rivets and epoxy. Luckily, I had a plethora of pop rivets from a suspended ceiling installation, and am fortunate to have a stock pile of West System epoxy and thickening additives.

I used a 4' x 10' sheet of 26 gauge galvanized sheet metal, which I got for $25. Unfortunately, I am dealing with a 7' ceiling height, and even with the motor sticking into the rafters, Bill's ideal cyclone size would not fit. Bill wants a cyclone to have an 18" diameter and a 1.64 cone/length ratio. My cyclone is 17" in diameter with a cone/length ratio of 1.3. The result of this, I think, is that some of the chips that are sucked into the system rotate around the lower part of the cyclone for a while, sometimes until it is turned off, rather than falling right down into the bin. This is not a problem: I have seen no particles in the filter housing, so it hasn't affected the cyclone's ability to separate. But (I think), because of the physics of the air circulating and then going up and out the top, my cyclone seems to suspend some of the chips as they go around the bottom of the cyclone.

The cyclone went together very well with the aid of a 17" diameter plywood circle. This circle eventually became the top of the cyclone that was fastened to the blower housing. I was also careful to make sure that the overlap of the sheet metal in both the upper cylinder and the cone was oriented in such a way that the air/dust didn't hit a ridge as it rotated around, but rather saw a "ski jump", as Dale Critchlow called it.

Tabs were formed on the cone, and it was pushed over the outside of the bottom end of the upper cylinder. I put the 17" diameter piece of wood at the very bottom of the cylinder during this process which gave the proper shape to the cone and cylinder. Pop rivets were installed from the inside out whenever possible, and I took special care to make sure they were inserted flat and as tight to the surface of the sheet metal as possible. Those rivets that had to be installed from the outside during assembly were eventually drilled out and re-installed from the inside out. I took great pains to make sure that the inside was as perfectly circular and as smooth as possible. All seams, every pop rivet, and every tab was sealed with epoxy from the outside.

The air ramp proved a little more complicated. The picture at the side is my first one. Everything is fine with it, except it spirals the wrong way, and these are not reversible. The way I wanted my ducting to enter the cyclone meant I had to have the directing of rotation within the unit be the opposite of what most cyclones have, and so I had to re-do my air ramp. This one is identical to the one that's actually in my cyclone in every other way: the rivets go from the inside of the 8" outlet pipe to the outside, the tabs on the ramp all are oriented up so they are out of the air flow in the cyclone, and the seam where the ramp meets the outlet pipe was sealed and smoothed with epoxy.

The picture at the side shows what the air ramp and outlet pipe look like from the very bottom of the cyclone when they're installed. I was able to reach all the way down with a caulk gun and run a bead of duct sealant around the joint where the air ramp meets the outside edge of the cylinder, which you can see in the picture. I was originally concerned about using 26ga rather than 24ga sheet metal for this cyclone project, which some recommended. Until I had installed the air ramp, I felt that the upper cylinder might be a little fragile and flexible. The air ramp made the entire cyclone much stronger and much less likely to crush (I have run this thing for a long time with all gates closed and neither the ductwork nor the cyclone seem to have been crushed as some said might happen.) In this sense, the air ramp has additional benefit. Not only does it get the air flowing in the right downward direction within the cyclone and reduce turbulence, it also adds strength to the entire cyclone assembly.

The air inlet is rectangular, so I understand, because it reduces turbulence in the cyclone by feeding in the air and chips in a laminar fashion. The air inlet transition transfers the round network of ducts to the rectangle of this air inlet. Rather than forming my own transition piece, I was able to find a good HVAC fitting that was almost the perfect size. A little hammering with a ball peen hammer to smooth all the inside ridges, and this piece was riveted and sealed in place. At the bottom of the cyclone I began with a 7" to 6" HVAC reducer, cut off the 7" portion of it, cut tabs, and fit it over the outside of the bottom of the cone. I was able to install the rivets from the inside to the outside, and took special care before I put this piece on to hammer out any ridges on the inside of it that would catch dust or affect the downward circular rotation out of the bottom of the cyclone. This piece was epoxied from the outside as well. I spent a good deal of time making sure the entire cyclone was air tight from the moment of entry to the bottom of the bin.

Finally, the entire cyclone was spray-painted with two coats of machinery grey paint.

Part 4: Installation

I bought the bin from Oneida since it was the ideal size (a low height yet it still has a pretty decent capacity (35 gallon)), its top was self-sealing, it was cheap, and there was free shipping from Oneida. I cut a hole in the top myself. I used a 6" starter collar, cut tabs in it, riveted it into place, and sealed it up tight with epoxy.

The motor was mounted to the motor-mounting board first (see blower page), and the hole where the arbor pokes through was sealed with silicone.

Original Installation

As the pictures show, originally the motor-mounting board was attached to two 2x4's in the rafters. The bolts that mounted this board to the 2x4's had to be recessed since the top of the blower housing would sandwich right up against this board (see picture above). The motor mounting board was connected to the 2x4's in the rafters with four 3/8" bolts and 8 nuts for safety sake (don't want these things vibrating loose). I bought 8 anti-vibration pads from McMaster-Carr, drilled 3/8" holes in them, sandwiched one between the motor mounting board and the 2x4's, and put one on top of the 2x4's between them and the 3/8" nut/washer that pulls the motor mounting board up tight against the 2x4's. The idea was to make a break in the wood-wood and metal-wood connections, which would help isolate vibration and keep it out of the rafters.

Then, the impeller was slipped over the bushing and tightened in place (with lock-tite on the set screws) about 1" below the surface of the motor board. This would put it right where it needed to be in the blower housing.

Since I had a 14œ" diameter access hole on top of the blower, I first attached the circular piece of plywood within the top of the cyclone with screws from the outside. Then, I set the blower on top of the cyclone making sure the outlet pipe in the cyclone lined up with the blower inlet, and that the bottom of the blower sat flat on the top of the cyclone. I screwed from the top down - through the bottom of the blower to the top of the cyclone - since this was easy to do through the 14 œ" access hole. I put a bead of silicone around the outside of the cyclone cylinder where it meets the blower housing.

Finally, I bought 50' of œ" foam strip that is about 3/8" thick before compression from McMaster-Carr. I put a circle of this around the 14 œ" diameter hole that would compress up against the motor-mounting board and seal that joint, yet still allow it to be taken apart. You can see this piece of foam in the picture of the blower on top of the cyclone. At this point, the blower/cyclone assembly was bolted to the motor mounting board with four 3/8" bolts and fender washers. These bolts were tightened well enough to significantly compress the foam on top of the blower.

Unfortunately, when the dust collector was turned on for the first time, I discovered that the anti-vibration pads were utterly useless. They reduced vibration slightly to be sure, but the entire area around the dust collector vibrated like crazy when this thing was on. The master bedroom is located right above the shop, so this would not do.

I pondered making a big stand and mounting the whole thing right to the cement floor. This would interfere with emptying the bin, however. So, I decided to fashion wall hangers out of angle iron. This would allow me to bolt the unit to the poured basements walls, which would eliminate vibration in the rafters above and diminish the overall cacophonic sound this unit sends through the house. See the discussion below for details on the Second Installation.

Once the cyclone was in place it was just a matter of connecting it to the ductwork, and using about an 8” piece of 6” flex to connect it to the drum. I have yet to experiment with lining the bin with a bag so it is even easier to empty the dust, but I plan to soon. I used about two feet of 6” flex to connect to the right end of the filter housing. However, I found in my attic several lengths of 6” flexible insulated ductwork, and thought about using this to connect the blower to the filter. This is supposed to reduce the noise by about 5-8 decibels. It wasn’t terribly loud in the shop, only in the room above where everything was vibrating since the motor was mounted in the rafters. Once this problem was solved with the wall hangers, sound throughout the house was greatly reduced. As well, when I put my ear to the blower, it seems to me that more noise emanates right from the actual blower (right through the sheet metal sides of the housing) rather than coming out the blower outlet and through the clear 6” flex hose. So, it didn’t seem to me that changing the hose which connects the blower to the filter was not going to significantly reduce noise. I have elected not to experiment with the insulated duct muffler.

Second Installation

The impeller falling off the motor gave me additional motivation to redo the installation to eliminate the vibration. As mentioned, I used angle iron to create basic 90 degree angle brackets capable of supporting the entire weight of the cyclone and blower (this probably amounts to around 100 lbs.). I used Tap-con screws to attach 2x4’s to the cement wall – I had to do this because my mounting procedure had to span the window opening you can see in the pictures. I made the brackets in my shop, bolted them together, ensured that they were the right size, and took them to my machinist friend to weld all the joints. This may or may not have been necessary, but it will make this set-up even more secure. The 2x4’s and the brackets were painted the same machinery grey as the cyclone.

It’s difficult to get into the details of this second installation, but hopefully the pictures and my explanation will help a little. The angle brackets had to be far enough away from each other so that the blower housing can fit between them. For me, this meant about 27”. Now, the motor mounting board is not nearly that long, so I had to effectively extend the motor mounting board so it would reach the brackets. For this I used 2 1/4” solid oak, since these would bear the weight of the entire dust collector, and put the unit at about the same height it was at before.

I first mounted these two oak hangers to the motor mounting board using Grade 8 bolts, the anti-vibration pads, and two nuts for security (see top picture at the side). Then, I drilled 3/8” holes at the two ends of each oak hanger, lifted that assembly, held it in its place between the brackets, marked a hole on each bracket where the new 3/8” holes were located on the hangers, drilled the four holes on the brackets, and bolted the oak hangers to the wall brackets.

Note that some of the long bolts that squeeze the blower housing together had to be cut off on the top so they didn’t protrude above the motor mounting board and run into the oak hangers (see bottom picture).

At this point, the procedure was the same as it was during the original installation. I attached the impeller (this time with hardened steel set screws and a roll pin – see the Blower page), affixed the blower housing/cyclone assembly to the motor mounting board, and connected the cyclone to the collection bin, filter housing, and the network of ducts throughout the shop.

Part 5: Filter & Housing

Steve Cater (http://www3.sympatico.ca/caters/Wood/Cyclone.php) alerted me to the fact that these cartridge filters we use are designed to be used from the outside to the inside. Most manufacturers go from the inside out since it’s easier and less space-consuming. The reason that it’s better to go from the outside-in, as I understand it, is that you maximize the surface area of the pleats of the filter, which will result in better air flow, and less resistance from the filter. From Bill’s page I see that another way to go about this is to use two filters going from the inside out, which creates enough surface area.

I didn’t want to have two filters, and at $56 from buy-filters.com, one was enough for me. I built the filter housing based off Steve Cater’s, with one minor difference. His was designed so the output from the blower was directed right at the pleats of the filter, because the pipe entered from the side. I was worried about this causing damage and causing a build up of dust on that side of the filter. As well, I didn’t want to take up as much floor space as this set-up would necessitate.

This mix of design requirements worked out well, and as the pictures show, I mounted the filter horizontally in a box that has holes on either end. So, the incoming air and fine dust from the blower hit the metal plate that’s at the bottom of the filter (now the right END of the filter) – it acts as a buffer.

The fine dust is captured in the pleats, and the air is expelled back into the shop out the hole in the other end. This works out well since I am building an air cleaner from an old squirrel cage fan I have from a replaced furnace. I will mount this on the opposite side of the shop, and, when both the dust collector and the air cleaner are on, it will create a nice circulation of air throughout the shop space resulting in better filtering and a safer environment.

A few notes on the design. All the joints where MDF meets MDF are sealed with silicone. Remember, I was taking extra pains to make sure that the cyclone, blower, and filter housing were hermetically sealed. The left end of the housing, as well as the plastic viewer in front, are detachable. I used the same ½” x 3/8” foam strip material to seal these joints. The reason the end is detachable is quite simple: it is the end that the filter is connected to via a ½” threaded rod and wing nut, and if the filter ever needs to be replaced or otherwise adjusted, that end will have to be removed. The plastic viewer is designed to see the condition of the filter. It also acts as a clean-out to clean the filter and any stray shavings or chippings that aren’t separated and find their way into the filter housing. Thus far, not a single one has, so I don’t anticipate using this too much. When the time comes to clean the filter, I plan to blast the thing with compressed air from the inside out by sticking a pipe with several holes along its length into the left end of the filter. This should put a nice caking of fine dust throughout the housing. In that event, I might remove the plastic to vac all that out. However, it occurs to me that it might be much easier to merely detach the flex hose from the right end of the housing, and stick my vacuum in that way. We’ll see.

Finally, the housing is mounted with hinges to the floor joists above. This is the case because the left end butts right up to my scrap rack. If I ever wanted to pull the filter out from the left end, I would need to bring the housing away from the wall. The hinges allow the housing to move away from the wall without detaching it.

Part 6: Ductwork

Money Savers

For the number of times Bill proclaimed that we must run 6” ductwork to all our machines, I decided to take his advice. It was not feasible for me to buy everything from Oneida – I would probably spend $2000. Here are a few things I did to save money:

1. I made 9 of my own 6” blast gates. The design is pretty straightforward. One thing that I did is make the hole in the collar that receives the ducting about 3/16” larger than the hole in the actual gate. This allows the 6” duct to insert into the collar and bottom out in the blast gate, and then it is easily sealed at the collar. To make these blast gates I drilled the holes in the gates (which are made of ¼” luan) with a large circle cutter on the drill press, set to about 5 15/16”. This was just right for the 6” duct to not go through the gate. These drilled fairly easily.

   I wanted about a 6 1/8” hole in the collars, so the duct would slip through the collar and bottom out against the gate. As well, 6 1/8” was large enough for some of the ductwork fittings to fit into the gate as well. For instance, several times a wye inserted right into a blast gate and these fittings need about a 6 1/8” hole. The circle cutter was not cutting well through the ½” plywood that I used for the collars, so I made just one collar with a precise 6 1/8” hole. I then rough cut with a jig saw the holes for the rest of the collars to just under 6 1/8”, tacked each one over the perfect 6 1/8” template, and used a flush-trim bit in the router to make 17 more 6 1/8” collar pieces. These collars with their 6 1/8” holes were then glued to either side of the gate pieces perfectly centered over their 5 15/16” holes, leaving a 3/32” reveal all around. This is a little difficult to explain, so hopefully the above picture helps.

   

   

   

   Planer, horizontal belt sander, and four-inch floor sweep hook-ups
2. I was lucky to have four metal blast gates and several plastic gates all in the 4” size from my previous system. As well, I had a few lengths of 4” spiral pipe and a great deal of 4” flex from the previous system. So, these costs are not included in the cost section of this site, since I didn’t have to purchase them. I was able to use these gates and lengths of pipe and hose only at the very end of duct runs, as the pictures show. For instance, the planer dust hood has a 4” outlet. So, the 6” duct was reduced to 5” flex, which was reduced to 4” right at the blast gate at the hood. Or, for the horizontal belt sander and band saw, I used a reducing wye and then installed the 4” blast gates and hose, since I think that these are large enough for the tools they serve. The only other place that 4” gates and pipe were used was for one of the floor sweeps, which has a 4” outlet.
3. I bought as much as possible from home centers (Menard’s and Home Depot around here) and as little as possible from dust collection supply companies. Home centers have the HVAC pipe, reducers, flexible elbows (I used a tremendous amount of these), hose clamps, and various take-offs, boots, and hoods that can be retrofit to work in the system. The only items I purchased from dust collector supply companies were the 6”x6”x6” wyes (at $20 a pop, I had 12), reducing wyes (I bought about six of these), a couple 5” and a few 3” blast gates (since I grew tired of making my own and they were only $10 and $6 respectively), and 3”, 5”, and 6” flex hose (since you can’t really find good flex hose for dust collection applications anywhere else, and Oneida is famous for their efficient flex hose). So, avoiding purchases from dust collection companies was difficult, as there are certain things you have to get from them, but I think I saved some money by going to home centers for the parts they had.

   

   HVAC boot used as floor sweep

   These money-saving ideas do not make the ductwork cheap by any stretch of the imagination, but they did save hundreds of dollars for me. For instance, for a few hours of work with scrap materials I had nine 6” blast gates that would have cost me $109 plus shipping. I saved about $200 by purchasing three HVAC boots, and making two cuts on them that turned them into phenomenal floor sweeps. As well, the adjustable elbows (of which I probably used 20-25), reducers, and ductwork are all just cheaper at the home centers.

   A cost analysis of the entire system is located on page 8 of this article.

Special Considerations

Anyway, I took great pains to seal the ductwork UP TO every blast gate, because every leak that exists prior to a blast gate will be a leak that affects performance at every machine at all times. I was not quite as concerned with sealing every single seam after the blast gates, as these would not have a detrimental impact on overall system suction. The various sources of ductwork integrated well together – that is, crimped ends always fit into uncrimped ends.

I discovered a pretty easy way to handle two uncrimped ends meeting one another: rather than trying to create crimps or cut slits to make these things go together, I would just butt them against each other, and surround them with a short piece of snap lock duct ONE SIZE LARGER than the ducts that were being coupled.This piece merely wrapped around the joint (the snap lock was not utilized), overlapping itself, and was tightened down with two hose clamps. I used this technique about four times. The picture shows how the connection of an uncrimped end of an adjustable elbow to the uncrimped end of the miter saw hood is made by wrapping the two with overlapping 8” snap-lock (extra from the 8” diameter cyclone outlet pipe) and tightening with two hose clamps. It’s very cheap and it felt good to use the small pieces of ductwork scrap I had left over from cutting other ducts to length. Hopefully the picture and explanation help you understand this idea. I used 6” duct scrap to join two 5” uncrimped ends, 5” duct scrap to join two 4” uncrimped ends, and 4” duct to join two 3” uncrimped ends.

A few other innovative ideas I utilized were at the sanding center and the table saw. The sanding center, which is the bench that has the 1” belt sander as well as the 6”disc/4”belt sander, would necessitate two reducing wyes in addition to the wye off the main duct. This would be costly and take up more space than was there. So, I built a box with a 5” hole to the dust collector, a 4” hole for the 4” belt sander, and two 3” holes for the 1” belt sander and disc sander. I built custom dust ports for the belt/disc sander to capture the dust at the source, and the hood for the belt sander is adjustable so it can be put below the plane of the belt if necessary for longer parts sanding. The blast gates fit right into the box so they’re all centralized, and the box is secured to the wall.

The table saw vertical duct drop is removable. It is set back from the table saw about 15”, and allows ripping up to 43 ½” in its current location. I did NOT want it to run on the floor, and I didn’t want it to be further to the right of the table saw because it would interfere with walking through the shop. As well, since the surface planer is to the right of the table saw, it might get smacked a lot as boards are being run through the planer, brought back around to the front, and fed back in. I don’t anticipate disconnecting it often, since it allows me to rip 43 ½”. The only problem I foresee is if I’m using the fence as a stop block to cross-cut a long board or glue-up that is wider than about 15-18”. This is the only circumstance in which I see a board making contact with the duct, which will only happen rarely. So, I’m pretty happy with the layout.

The other thing to note is that I extended the blast gate with a 5/8” dowel I had laying around, so I can open it from the front of the table saw. I connected the 6” duct to the table saw with an HVAC hood that takes it from 6” round to 10”x4” rectangular, which works out well for the base entry of the Unisaw.

General Set-up

As the photos show, the set-up is essentially a 6” main trunk pipe running diagonally across the shop with several 6” branches coming off it. The longest duct run is to the miter saw, which is roughly 35 ft. of 6” ducting through a few wyes, and several adjustable elbows. I used no tees, only wyes. This cost more, but there is unanimity in the dust collection world that wyes are the right fittings to use. I located the blast gates as close to the collector as possible without putting them in an inconvenient location to use at each tool. I did this because the more duct there is before a blast gate, the greater the chance is that there will be a leak that affects the entire system.

Because of the design of my blast gates (they are the “self-cleaning” design), they can’t be installed against a wall to simply be pulled out when I want to use that machine. This is because when they’re closed, the moving part of the gate extends about 7-8” behind the gate. So, these gates were often installed along the ceiling above the machine, since there was just enough room in the floor joists above to shut the gates.

As much as possible, I used two adjustable elbows to make a 90 degree bend, as this would make a less abrupt, longer sweeping turn in the system. Oftentimes the 90 degree turn was coming off of a 45 degree wye, so only one adjustable elbow (set to 45 degrees) was required to complete the turn. These adjustable elbows were lifesavers for me – they allowed me to maneuver around all the crap that is found along our basement ceiling – heating ducts, plumbing and electrical pipes and fixtures, phone lines, etc. I couldn’t get a consensus on whether or not every seam on those adjustable elbows needed to be sealed or not, so, prior to the blast gates in the system I sealed every seam, and after the blast gates I only sealed the seams where the crimped end inserted into an uncrimped pipe. This sealing was all accomplished with duct tape, and the sealing at the blast gates (where the duct inserts into the collar of the blast gate) was done with duct sealant, and, in some cases, epoxy. Oneida’s site confirmed that the we don’t have to seal the long seam along the snap-lock pipe. All in all, I think that this recipe works well.

Tool Hook-up Specifics

1. I ran flex hose to every machine that I foresee moving a little from time to time. This meant 4” flex to the horizontal belt sander (which is on wheels), and the band saw (which has yet to be hooked up but has that 4” black plastic gate designated for this tool), 5” flex to the drill press and planer, and 6” flex to the router table which may need to be adjusted from time to time depending on how it interacts with the miter saw, which is directly to its left. Also, the sanding center utilizes 3” and 4” flex even though those machines are stationary, merely because it was much easier to hook them up that way.
2. All other machines were hooked up with either a direct run of 6” rigid duct (radial arm saw, miter saw, two floor sweeps, jointer, and table saws), 6” reduced to 5” (one floor sweep), or 6” reduced to 5” reduced to 4” rigid duct (one floor sweep).
   

   

   6" rigid duct right to cutterhead
3. 3. At the jointer, the 6” duct is run right up to just below the cutterhead. Nothing escapes on this tool.
4. As indicated before, I used the 4” port on the planer, but ran a short length of 5” flex from a 6” wye, so that works just fine.

   

   5" floor sweep using HVAC boot
5. Also as indicated before, I used HVAC boots and register adaptors in several places. My two 6” and one 5” floor sweep are adapted HVAC boots, I used a boot to attach to the table saw (which went from 6” round to 10”x4” rectangular). As well, HVAC fittings were used at the miter saw and radial arm saw, discussed next.

   

   Radial-arm saw dust port
6. The radial arm saw collection is accomplished with a 6”x5”x3” wye. This wye has a 3” flex hose going to the front dust port of the blade then through a 3”x2” reducer. Behind the blade I used a common HVAC fitting that transfers the duct into a 10”x4” rectangle. I made a couple cuts in it to open it up and receive as much of the dust as possible, and on the left side I added an extender to do the same. I don’t ever really plan on using the various angle settings on the radial arm saw – it’s really only a rough cut saw that I recently inherited from a good friend of mine – so I went ahead and screwed the dust port to the table right up next to the blade at its 90 degree setting. Up to this point, I haven’t found it necessary to connect the 3” flex to a spring or bungee cord to hold it up.

   

   6" miter saw dust port
7. The miter saw uses a similar HVAC fitting as the radial arm saw, but it is taller with a narrower opening, 14”x2”. I thought this was a better height for the way I wanted to collect dust from this tool. I pop riveted some wings and a small top to the HVAC fitting to capture as much of the dust as possible. It seems to do a good job. I considered using a wye like at the radial arm saw and running a hose to the port where the dust bag is generally attached, but seeing as all that dust gets shot directly backwards anyway, this would prove useless. All that comes out of the port gets sucked up instantly by the main hood. Cutting miters, this hood is not quite as good at collecting everything, but gets the job done. It certainly collects the fine dust from this tool which is what I’m most concerned with here.

   I am considering getting a saddle-tee to tap into this duct run and have a hood to collect at the mortiser, which is right next to the miter saw.

   

   Dust collection bag used to collect at the open-based contractor's saw
8. The 6”x6”x6” wye on the table saw drop splits to the Unisaw and to the contractor’s saw. These tools do not move much, so rigid duct was run right up to each of these machines. The 10” contractor’s saw presented a dust collection challenge, but I am pretty happy with my solution. It was accomplished by getting a dust collection bag from Harbor Freight (about $3.00) that fits open-based table saws. I cut a hole in the back of this bag, and fit a 6” starter collar through the hole. As well, I dropped a piece of ¾” particleboard in the bottom of the collection bag so it would keep its shape amidst the suction from the dust collector.

   

   5" flex for drill press
9. The drill press currently has a 5” flex duct hanging down towards it. I bought 10 ft of 5” flex, and this is what was left after the planer run. This will be used in conjunction with various adaptors to collect from the drill press when it’s being used as a drum sander (I have a box with a collection hole), and also will just sort of rest on the table when doing large drilling operations hopefully collecting some of the fine dust that might result from those. This and the floor sweep next to it are where I used the two 5” blast gates I bought.

   

   6x6x3 wye with adaptor for shop-vac hose
10. There is a 6”x6”x3” wye over the center workbench in the shop, along the run that goes to the miter saw. This has my third 3” blast gate which is hooked up to a 3”x2 ½” adaptor. I happened to have an extra length of 2 ½” hose for my shop vac, and this is hooked up to this gate. So, I can use this hose in conjunction with another adaptor (a 2 ½” x 1 ¼”) to collect dust from routers with dust ports and sanders. This set-up is great because the hose comes from above, and stays out of the way. This hose can also be used for general clean-up after hand planing or drilling on the bench.
11. There is a 5” branch duct that extends beyond the center bench over to the workbench with the pegboard. This is on the opposite side of the shop from the dust collector. I have no idea what it might be used for, but it’s there right above the bench if I ever need to tap into it.

Part 7: Electrical Work

This area deserves a brief write-up. The motor is 220V. I ran 10/3 flexible conduit from the circuit box, through a double pole switch, and then to the motor. I was fortunate to have a spool of this stuff from a previous electrical project. Otherwise, it is probably $50 in wiring. In the circuit box, which I am fortunate to have right in the shop, I used a double pole 20A breaker. I think I could get away with a 15A double pole breaker, but I elected to use a 20A. This proved a good idea, because I ended up having a couple other items operating off these breakers. I added a low voltage switching system and a fan to cool the motor, both of which were powered with the dust collection circuit.

I was not satisfied with having just one switch to turn the collector on, so I replaced it. I purchased a DPST (Double Pole Single Throw) relay from McMaster-Carr for $12, which operates with 24 VAC input. Essentially, this is a double pole switch actuated by the completion of a 24 VAC circuit. So, I wired right into one of the hot wires of the dust collector circuit a 120VAC to 24VAC transformer. These are available for about $12 from a hardware store (your doorbell probably operates with one of these). The reason you transform this circuit to low voltage is because then you can just run wires without having to worry about conduit or electrical boxes and connectors. I then located six total switches in convenient locations around the shop – two 3-way switches, and four 4-way switches. I used about 100 ft. of 18/3 thermostat wire to wire all these switches together so that the dust collector can be turned on or off from anywhere in the shop.

I chose this system for two reasons. The first was its cost effectiveness. I was able to procure the switches and switch plates for free – so this set-up only cost me $34 for the relay, the transformer, and the thermostat wire. I didn’t want to spend money on a fancy remote system, and didn’t think it would fit my needs. Secondly, I didn’t want to put a switch on every blast gate as some have done. This would mean that every blast gate has to be closed to shut the system down, and sometimes I don’t want to have to do that. As well, if I’m moving from tool to tool, I would have to go open the second tool’s blast gate, then close the first tool’s blast gate, just so the motor doesn’t turn off and back on. Not only is this inefficient because of the extra time it takes, it is also a waste of electricity (it takes a lot of juice to start a motor over and over again) and is extremely bad for the motor. I am no expert with electricity or motors, but I have read that more than 6-10 starts in an hour is horrible for a motor. It has to do with the capacitors getting too hot, or so I understand. As well, you can just tell intuitively when you turn this baby on that the starting phase of the motor is the most painful thing for it.

Anyway, my system of 3 and 4-way switches meets my needs best, reduces the number of times that the motor gets turned on and off, and I think gives me more flexibility. The switches are located over the miter saw, over the radial arm saw, amidst the band saw/belt sander/drill press cluster, in between the sanding center and the surface planer, above the jointer, and above the table saw. Basically, they’re located around the perimeter of the room and one in the middle of the room.

The second item that was added to the dust collector circuit, and made me glad I used 20A circuit breakers, was a fan to cool the motor. The motor I used is a compressor duty motor, and is not designed to be mounted up against the motor mounting board as it was. They are designed to have air flowing through them to cool them as they operate. My set-up made it very difficult for air to flow through this motor. Not only was it mounted up against a board (it’s actually about 3/16” above the surface of the motor mounting board because it rests on four small bolt heads that hold the motor case together, so there’s space for a tiny bit of air flow), but it was mounted up high in the corner of the shop where little air flow can get to it. To complicate things, since heat rises, any heat that the motor creates stays right there in that corner. I effectively created a self-heating oven for my motor.

Now, this oven takes a little while to develop - about 25-30 minutes. I noticed it for the first time when I had the collector on for about an hour, and then it wouldn’t turn back on because its own internal thermal protection system wouldn’t allow it. I’m actually surprised it didn’t just shut itself down, which is what I think these thermal protection switches are supposed to do, but I’m not sure. In any event, I wired a fan to turn on when the dust collector does. This fan is aimed right at the motor and hopefully accomplishes two things. One, it gets more air flowing through the motor to cool it. Two, it brings cool air to that upper corner and pushes some of the hot air that may develop out away from the motor. It does not keep the motor cool, but it does keep it from getting so hot that it burns up. In the future, I might put up a fan that can move a larger volume of air to see if it cools the motor more effectively. As well, in warm weather, I can open the window behind the dust collector and really create a good ventilation system through and around the motor.

Part 8: Cost Analysis

Cyclone/Blower/Filter Assembly

I was fortunate not to have to pay for pop rivets, epoxy, or the foam that makes the seal at the top of the blower, the left end of the filter housing, and between the plastic viewer and the front of the filter housing.

Ductwork

Home Depot and Menards:

I spent a total of $221 at these two home centers. This got me all my 5” and 6” snap-lock pipe, adjustable elbows in the 5” and 6” size, hose clamps, several boots and take-offs that functioned as hoods, most of my reducers, and my tube of duct sealant.

Kencraft:

This was the cheapest supplier I found for dust collection accessories. Remember, I only purchased items that I had to from these types of suppliers since the prices are quite high. I bought 12 6”x6”x6” wyes, a few reducing wyes, a couple 3”x2” reducers which aren’t available at home centers, three 3” blast gates, and two 5” blast gates. Unfortunately, this supplier set me back $425, including shipping.

Oneida:

This is the pinnacle of cyclone dust collection and accessories, as well as prices. I stayed as far away as possible from them! They do offer free shipping for orders over $50, though. So, from them I bought from them only items I couldn’t get from Kencraft. This included two reducing wyes which I couldn’t find at Kencraft (6”x6”x3”, and 6”x5”x3”), my 35 gallon drum (its cost is included in the cyclone cost), their 3”x2 ½” reducer to connect to the shop vac hose over the center bench in the shop, 10’ of their 3” flex for the radial arm saw and sanding center hook-ups, 10’ of their 5” flex for the drill press and surface planer, and 5’ of their 6” flex, half of which was used for the cyclone (and its cost is reflected there), and half of which was used to run to the router table (its cost is reflected here). As it applies to the ducting cost, I spent $170 at Oneida.

Electrical Components

As I indicated, I was able to get the 3- and 4-way switches for free, as well as the 10/3 flexible conduit. This probably saved me nearly $130.

Grand Total

Part 9: Conclusion

Well, I saved a tremendous amount of money building the dust collector. The dust collector itself only cost $350. Although the low end cyclones can be as cheap as $600, the cyclone that I built is probably at the same level if not stronger than the Woodsucker II unit at @ $750, or Oneida units upwards of $1000. I am thrilled about this aspect of the project. I don’t have access to gauges and meters to give actual CFM readings or any other readings, but I can say that this unit’s suction is phenomenal, that it captures the fine dust at all my machines, and that it will also suck up some surprisingly large chunks of wood as well. However, a fellow woodworker has gotten in touch with me via email and has graciously offered to send me the necessary equipment to measure the power of this dust collector. I will post those figures if and when I have them.

Building a collector this size locked me in to running 6” to every machine, or else it wouldn’t be worth all the effort. The ducting cost over twice as much as the dust collector itself ($816)! And it would have cost much more had I not made my own blast gates and floor sweeps, not purchased as much as possible at the large home centers, and not had all the 4” blast gates, spiral pipe, elbows, and flex hose from my previous system. Without all these money savers, I expect I would have spent closer to $1300 on the ductwork alone. I never saw this ductwork figure coming – it crept up on me as I spent more and more for the network of ducts to reach every machine.

All in all, this was worth it. For one, I fulfilled a New Year’s resolution a month and a half after the new year – I really wanted and needed to upgrade this aspect of my shop. I think that it has increased the overall caliber of my workspace. Secondly, this will keep the surfaces of the shop as well as the entire household much cleaner…forever. Most importantly, I have taken an enormous step toward a safer environment for myself and my family. You can’t put a price on that. To be honest, I feel that I found the cheapest possible way to create an extremely effective system that meets all my needs and my family’s needs. And, it’s a conversation piece that will last a lifetime…

Thanks for visiting this page. I hope you have found my ramblings useful, and that this page might help you decide where you want to go in your dust collection adventure. I say go for it – spend some time and money and invest in shop safety and cleanliness – you’ll be more than happy with the results. Good luck.