Air cleaner with homemade pleated filter

I built this air cleaner back in 2012, and it made a big difference in my shop. I wanted another air cleaner to use in my big garage workshop, but by now, these big filter cartridges cost $35 (+tax) each, and I had an idea for making my own pleated filter.


Making a pleated filer with chicken wire

My motivation came from finding this 30"x62" "hammock filter" for just $12. it seemed to be made of a similar material as the filter cartridge I last used. It would make a much cheaper filter, if I could figure out a way to support it in a pleated fashion.

On pleated furnace filters with the soft felt lining, the filter fabric is always held by some thin wire mesh.


I figured I could try to create a pleated wire mesh using some chicken wire. As this was just an experiment, and I didn't want to spend a lot of money (that would defeat the purpose), so I used some rusty old chicken wire that I found.


From previous experiments, I knew a wire mesh wasn't terribly easy to fold, so I made a jig to help.


The "jig" consists of several boards, supported between two workbenches. The boards all have a hole in either end so I can put a nail through them to align the layers.


Then I can just fold the wire mesh over, add another board, fold it back, add another layer, and fold in the opposite direction, and so on.

I only made four boards, so I have to pull them out from below as I go.


Half the folds made.

I started the folds in what would be the middle of the filter, working towards one end. After that, I cut as much as I needed for the other half off the roll and folded to the other end. That way, I didn't have such a long end to keep folding back and forth.


The folding was more difficult than I hoped. But it looks like it's not too difficult to get the filter material to follow the wire.


I needed to make a box for the unit, and for that I used some silver maple that I milled on my own sawmill some years ago. First I glued pieces side by side to get sufficient width, then surface jointed, edge jointed, planed and ripped them to width.


Then cutting box joints in the corners using my screw advance box joint jig. I cut 1/4" fingers.

It's always satisfying to see joints with that many fingers go together.


Folded chicken wire stuffed into the box. When I folded it, I had it slightly skewed, which made it want to go into the box crooked.

I worked out that I had enough filter material for 11 1/2 folds. I then worked out how big the folds could be, and laid it out. It was tempting to do it in SketchUp and print it with my BigPrint program, but I ended up just using a calculator, tape measure, and bevel gauge. Once I worked out the filter spacing that I needed, I used "constant calculation" mode on the calculator to calculate the intervals. Constant calculation is just repeating the previous operation if the "=" key is pushed again. Nearly all five function calculators do this, but most scientific calculators don't.


Then cutting the supporting combs out on the bandsaw. After cutting the first one, I drilled holes in the corners where I needed to turn, which made the cutting easier.


Screwing a comb on either side of the box.


Then fiddling the folded chicken wire onto them.

The result was not as straight as I had hoped.


So I made two more combs to support the mesh closer to the middle.


Here's what it looks like close-up. The rusty old chicken wire was brittle and quite a few pieces broke out. But there was still enough to support the filter material.

So far, this was a lot more work than I expected.


I had been thinking of other ways of supporting the filter material before I started. For example, just a series of thin rods to hold the folds (top), but I figured the filter material would probably end up billowing a bit and pinch off part of the surface on the back (as shown at right on the image at left), so I would need a whole lot more rods to hold the filter material apart on the back (bottom left part of image). So this could get quite complicated, and, if I did this, I'd probably figure chicken wire would be so much easier!


The filter material was 30x62" (76x157 cm), but the chicken wire was only 24" (60 cm) wide, so I had to cut some off the side. To get more area, I then sewed what I cut off on the short edge to get more length. Glue would have done the job too, but sewing is more elegant.


Fitting the filter material into the pleats.

The last scrap I sewed on didn't cover the full width, so I put some blue shop towel in the corner.


Adapting an oscillating fan

I wanted to buy a 20" box fan, but it appears these are only in stores in the summer so I couldn't buy one. Instead I adapted an old 16" oscillating fan, here shown already partially disassembled.

The challenge with these is how to mount the motor. I realized, if I took the gearbox off the back, I could use those screw holes to mount it.


The wires from the motor to the buttons to select the speed went through the plastic base. I didn't want to cut the wires, so I cut the housing in half to get them out.


The motor has an E-ring on the shaft to keep the fan blade from sliding too far back. But I didn't have much depth to work with in the box, so I needed to move the fan blade closer to the motor. Here, prying off the E-ring with a screwdriver.


Then cutting off the shaft with an angle grinder. I also cut off the worm gear part of the shaft sticking out the other side.


I then used a Dremel cut-off disk to grind a new groove around the shaft near the motor so I could put the E-ring on there.


The fan blade keys onto the shaft with a flat spot, but that flat spot was part of the shaft that I cut off, so I had to grind a new flat spot.


With the fan blade further back, I could now use the groove that was behind the blade for an E-ring in front of the blade. Except I was maybe half a millimeter short of space, so I widened the slot using the same technique as making the previous slot.


Quick test, spinning it up in the vise.


Update: I later realized the shaft on these motors is only press fit into the rotor, so on my next air cleaner, I just pounded the shaft back through the rotor to move the blade closer. Much easier!


Now I need to make a mounting bracket to mount the motor in the middle of the box. I figure I'll make it out of some 6 mm plywood.


To transfer the screw hole pattern, I put the screws in the holes, then the plywood against the screws, and tapped it with a hammer.


I could barely see where the screw heads dented the plywood. I center punched those locations with an awl and drilled the holes. It fit perfectly.


Having mounted the motor, I outlined the frame on the plywood with a pencil, then drilled some holes around the motor frame for better air flow to cool the motor.


I wanted to cut away more wood around the motor, but to not weaken the bracket I first glued some strips of hardwood on it at the edges.


I then used two spring clamps to position the bracket in the housing and drilled through the holes in the bracket to get the pilot hole locations in the fan box.


Then with the motor screwed onto the bracket again, I screwed the bracket in place.


I mounted the motor's fan buttons to the side of the case, with a slot for the buttons. The buttons sit just flush with the box, and just about match the colour too, I like how that looks.
I also added a clamp for strain relief of the cord


To make sure all the air that the fan blows actually comes through the filter, I needed some sort of baffle. I cut that out of some hardboard (or Masonite) that used to be the back of some discarded furniture.

I always question my sanity when I save material like this, but end up glad when I find a good use for it.


Drawing circle for the fan using my beam compass


And cutting it out on the scrollsaw.


I figured a screen in front of the fan would be good, and the old fan guard was just perfect for that. I drilled some new screw holes, large enough for some #4 screws.


I glued some bits of 6 mm plywood to the back of the masonite where the screws go, then screwed the grille on.


I added some rails around the edge to screw the baffle on. To fit the baffle, I drilled through it and through the rails to get the holes perfectly aligned.


But the hardboard had some mold stains on it. I tried sanding those off, but no success. Then figured green paint would look good, especially because I still had the can of paint out from painting the wooden lathe


Finally screwing the panel on.


Testing it

Setting it up for some testing. My first test was to run it without having made any dust to see how it would knock down the ambient airborne particle count. I also set up an oscillating fan to make sure the shop air was well mixed, and my Dylos air quality monitor to measure airborne particle counts.

I hadn't used any machines yet that day, but it still dropped the ambient particle counts substantially. The dark blue line is particles from 0.5µm to 2.5µm, while the magenta line is particles 2.5µm and larger. Counts are particles per 0.01 cubic foot.


My next test was making some extra dust to test with. I'm using the top wheel of my strip sander to grind some MDF, with none of the dust sucked into the dust collector. I had an oscillating fan running to spread the dust evenly throughout the shop. I only ground away at the MDF for about 35 seconds to raise the fine dust 4x and coarse dust 10x over ambient levels.


With the air cleaner running it would cut the dust level in half every three to four minutes. This is almost twice as fast as my other air cleaner. The higher rate is probably because I have a higher air flow rate with this air cleaner (this one has a bigger fan and straighter path for the air)

What I found odd is that the dust levels dropped faster and further than with my previous ambient air test. My first thought was that maybe the MDF dust particles clumped with the ambient particles and helped to collect them, but then I realized, I'd left the air cleaner running while I took the Dylos to my computer to enter the data in a spreadsheet. My guess is, whatever the ambient particulate was, it was harder for my filter to catch it than the MDF dust. Having left the air cleaner running for half an hour between experiments had probably cleared out most of the ambient particulate. With the added MDF dust easier to clear, the dust levels dropped faster than in my first test.


I then repeated the same test with my old air cleaner. This produced a similar exponential drop-off as the new air cleaner, though it's not able to get the particle count quite as low as the new one. It used to be able to get the particle count much lower when the filter was still new.


Going back to the first test, that the ambient particles were much harder to filter out suggests that there is different dust of a different nature. The ambient particulate had about 5x higher readings on the fine particles than the coarse particles, so it suggests that it was altogether a finer dust.

I also read recently that 3D printers can cause a lot of "nano particles". I don't have a 3d printer, but 3d printing is a bit like a glue gun, and using the glue gun in my shop did raise the particle count very substantially (higher counts than grinding up MDF).

A lot of people very paranoid about wood dust, but not many people worry about glue guns. But I think I'd rather breathe wood dust than plastic particles!


My old air cleaner still uses the same filter that I bought for it in 2012, though I have knocked the bulk of the dust out of it twice in that time.

I'm not so worried that it can't get the dust levels as low as it used to — it's the really high dust levels that I'm more concerned about, and this filter is still quite effective at reducing those.


The filter looks quite filthy, but as long as it works, I'm not complaining.


See also:

Mini dust collector
for single, small tools

More about dust collection on my woodworking website.