Box Fan vs. Shop Air Cleaner: Measured CADR, Real Filter Costs

Yo, let me tell you a story. I'm standing in my second bedroom — the one my landlord thinks is for guests — staring at two options on the floor. On the left, a box fan I hauled off somebody's stoop in Park Slope, twenty bucks and a subway ride where I got some looks. On the right, a printout of a Jet AFS-2000 going for $934.99 at Acme Tools. My shop is nine feet by eleven. My dust collector is loud. And I've been breathing MDF dust for a week because it's February and the windows are painted shut.

So I did what any reasonable person does: I fell down a rabbit hole of CADR ratings and pressure drop charts. And here's the thing I learned — the argument between a thousand-dollar ceiling unit and a forty-dollar box-fan build gets settled by which number you're actually measuring. Get that wrong and you'll spend money on the wrong machine.

Two Numbers That Don't Measure the Same Thing

CADR stands for clean air delivery rate. Per Wikipedia, citing AHAM and the EPA, it's "the cubic feet per minute (CFM) of air that has had all the particles of a given size distribution removed." It's measured under ANSI/AHAM AC-1 in a standard 1,008-cubic-foot room over a 20-minute test. You get three numbers: smoke, pollen, dust. Particle sizes are 0.09–1.0 μm for smoke, 0.5–3 μm for dust, 5–11 μm for pollen.

Here's the part that matters: WOOD magazine's air-filtration test didn't measure CADR. They measured CFM at the intake with a vane anemometer, in 12 spots, and converted to air changes per hour. So when you see the Jet AFS-2000 measured at 1,178 CFM, that's airflow. It is not CADR. Anybody comparing these two machines head-to-head has to say which number they're using, because the cheap build can win or lose depending on the choice.

What the Ceiling Unit Actually Delivers

WOOD tested nine ceiling units plus the box fan in an 850-cubic-foot chamber, generating dust with a drum sander on MDF and 120-grit paper, until the air hit 6 million particles per cubic foot. The Jet AFS-2000 cleaned it in 13 minutes at high speed. Their math: in a 24×24 shop with 9-foot ceilings — 5,184 cubic feet — that 1,178 CFM gets you 13 air changes per hour. Manufacturers suggest 6–8 ACH. You'd need two units to hit 6 ACH with the Grizzly G5955, the Rikon, the Shop Fox, or the box fan.

Two things about that Jet number worth flagging. Acme Tools lists the AFS-2000 catalog rating at 700 CFM. WOOD measured 1,178 at the intake. No source in my research explains the gap, so I'm not going to pretend they're the same figure.

And the consumables? That's where the "$935 machine" turns into something else. Jet replacement filters at Acme: $79.99 for the 5-micron electrostatic outer, $174.99 for the 1-micron inner. That's $254.98 for one filter change on a $934.99 machine. WOOD's own note: inner high-efficiency filters cost 4–5× the prefilters, so keep the prefilter clean. Clean it every 7–10 days of shop work. Vacuuming a filter with a shop vac that doesn't trap fine dust just returns it to the air.

The Box-Fan Build, Measured Properly

The most rigorous numbers on this build come from the EPA Office of Research and Development — Holder, Halliday and Virtaranta, published in Indoor Air in 2022. They ran a 20-inch box fan with one 1-inch MERV 13 filter in a 29.3 m³ chamber simulating the AHAM test, against simulated wildfire smoke. High speed: 111.2 ± 1.3 CFM CADR. Low speed: 79.7. Medium: 96.9.

Then they tested variations. Add a free cardboard shroud: 156.1 CFM, +40%. Swap to a 4-inch MERV 13: 247.9 CFM, +123%. Two filters in a wedge plus shroud: 263.1. Four-filter Corsi–Rosenthal box plus shroud: 400.9, +261%.

WOOD agreed. They named the $20 box fan with a $20 MERV 12 filter their Top Value, saying it "outperformed or equaled several of the dedicated air cleaners" at about $40.

But wait — the filter matters more than the fan. Same EPA setup with a MERV 11 instead of MERV 13: 63.6 CFM. Swapping down cost about 43% of CADR. And WOOD's advice on the fan ceiling: minimum MERV 10, and above MERV 12 "will likely lessen airflow."

Where the Cheap Build Genuinely Fails

This is the part where I stop being the box-fan cheerleader and look at my own shop honestly.

Loaded-filter collapse. The EPA study measured the same box-fan build after the filter loaded with smoke. CADR dropped to 4.3 ± 0.9 CFM. That's a 96% loss after collecting only about 10 milligrams of smoke. Pressure drop rose from 0.27 to 0.29 inches of water — barely moved. Airflow felt normal by hand. Filterbuy, a MERV media manufacturer writing from the production side, explains the mechanism: residential MERV 13 media is electrostatically charged, and charged particles neutralize the fibers. That's Filterbuy's explanation, not a finding of the EPA paper — and it's about smoke, not wood dust. Whether it transfers to sawdust isn't established. But the number is real.

Dust-loading airflow loss. WOOD measured the box fan dropping the most airflow of any tested device, stabilizing at a 30% loss. The Rikon 62-400 was 28%, the Powermatic PM1200 was 23%. Three machines fell off less than 10%. A furnace filter taped to a fan has no prefilter to clean, so you can't follow WOOD's every-7-to-10-days advice without pulling and replacing the media.

Sizing. WOOD's 24×24×9 example is decisive: the box fan cannot reach the 6 ACH minimum in that shop. You need two units. Filterbuy frames it similarly — the basic build gives about five ACH in a 12×14 bedroom but only two in a 20×20 room, and "most disappointment with this design is a sizing mistake, not a design failure."

The MERV ceiling. You can't buy your way to better filtration on a box fan without paying in CFM. The fan has no static pressure to spare. Filterbuy's pressure drop chart at 1,000 CFM shows it: a 20×20×1 filter runs 0.12" w.g. at MERV 8, 0.18" at MERV 11, 0.25" at MERV 13. A 20×25×4 MERV 13 is 0.11". Thicker media costs less in pressure than thinner media does.

The cheapest build isn't the most efficient. EPA found the free cardboard shroud was worth +40%. Filterbuy notes it's the one most people skip. That's free performance most builds leave on the floor.

Where the Filter Stack on a Dust Collector Goes Wrong

The instinct is obvious: I've already got a dust collector, why not stick a cartridge on the outlet and call it an air cleaner? Because dust collector blowers have almost no static pressure to spend. Dust collection blowers average about 7 inches of water. His phrasing: "almost any small pipe, bend, wye fitting, small port, restriction, or other obstruction will act just like a partially opened water valve and kill our airflow."

Every inch of resistance a filter adds comes straight out of CFM. Wynn Environmental's 35A274BLOL cartridge is $236, rated MERV 10, with 274 square feet of media, 17.5" by 23". Their older install write-up on a Delta 50-760 describes the added flow area as adding CFM — but that write-up calls the same filter "0.5 micron," and the product page now says MERV 10. I'm citing MERV 10. The discrepancy isn't mine to resolve.

And the source-capture principle Pentz states matters more than the filter spec: if fine dust escapes into the air, it takes hours for an exhaust fan or air cleaner to bring levels back down. Capture at the tool beats any filter downstream.

So Where Does the Money Go?

Here's my honest read. If your shop is small and closed — a spare bedroom, a basement, a one-car — the box-fan build is not a compromise. It's the correct first purchase. EPA's Corsi–Rosenthal build hit the lowest cost per CADR in the study at $0.18 per CADR. A five-filter 2-inch design measured by researchers at 600–850 CFM at roughly $0.08 per CADR — though note that's a different build and method than EPA's 400.9, and I won't average them for you.

But build it right. One filter, no shroud, no 4-inch media — that's the failure case. Two filters in a wedge with a shroud, or a four-filter box, and you're in the 260–400 CFM range from EPA's numbers. That's what I'd do.

The ceiling unit earns its keep in a bigger shop with more particle load. The Jet AFS-2000 cleaned the 850-cubic-foot test chamber in 13 minutes at high speed, and WOOD's ACH math works out to 13 air changes per hour for a 24×24×9 shop. The box fan doesn't get there with one unit. But the $935 purchase price is not the cost — filter it every 7–10 days, replace the inner filter on WOOD's 4–5× multiplier, and you're running a machine whose aftermarket media can swing performance more than the machine choice itself. WOOD watched the Grizzly G5955 go from 213 minutes to 33 minutes on American Fabric Filter media, while the Jet AFS-1000B got nearly twice as slow. That's filter-to-blower matching, and it's the thing that decides whether your machine works.

Two more things from the tests that I can't explain and won't try: both Powermatic models cleaned the air about 60% faster at their slowest speeds. Powermatic's Tobias Bridges told WOOD he didn't know why. And WOOD found no tested unit loud enough to require hearing protection.

Start with the box fan. Build the shroud. Buy 4-inch media if you can. Two units, not one. Then measure your shop again before you spend the thousand.