DIY Welding Fume Extractor: Build a Safe Shop System

You finish a Saturday afternoon of MIG welding a trailer hitch in the garage, hang up your helmet, and realize the whole shop still smells like a foundry — a grey haze hangs at head height, your throat is scratchy, and you have a headache that wasn’t there when you started. That haze is welding fume, and every home-shop welder has stood in it. The good news: for a fraction of the cost of a commercial fume extractor, you can build one that actually captures fume at the arc — if you get the airflow, placement, and filtration right.

Why Your Garage Shop Needs a Fume Extractor (and What the Fume Actually Is)

Walk into most home shops five minutes after the arc has been struck and you can see it: a faint grey veil hanging at head height, a sharp metallic taste at the back of your throat, and a smell that lives in the curtains for a week. For a lot of DIY welders that haze has quietly become proof that work got done. It isn’t. It’s the visible edge of a fume cloud you’ve been breathing all afternoon, and it is entirely controllable.

The real question isn’t whether welding fume is bad for you. It’s whether a homemade extractor is worth the sheet metal, or a feel-good gadget that just moves the problem from your lungs to your rafters. The answer, before any physics:

Welding Fume Is Not Smoke

This distinction matters more than anything else in this guide. Welding fume is an aerosol of ultrafine metal and metal oxide particles, formed when the arc vaporizes the consumable electrode, the flux coating, and a thin layer of the base metal itself. That vapor cools within millimeters of the arc and condenses into particles that are largely respirable and sub-micron — small enough to slip past the upper airway, settle deep in lung tissue, and carry into the bloodstream.

What rises off the joint is three things mixed: the fume, a column of hot shielding gas, and burning flux or surface contaminants like oil, paint, and plating. Strip away the smoke from dirty metal and you’re still left with a metal oxide aerosol you can’t see, can’t reliably smell, and can’t cough back out once it’s deposited. The recipe changes with the job too — running flux core wire on aluminum produces a different exposure profile than stick welding mild steel, and stainless work adds hexavalent chromium. That variation is exactly why “it’s only mild steel” is not a safety argument.

The Open Garage Door Is Dilution, Not Capture

The most common home-shop fume control is also the weakest: open the door, crack a window, and let the breeze sort it out. That’s dilution ventilation. It lowers the average concentration by mixing fume with more air, but it does nothing about the concentrated plume rising off your workpiece into your face. The fume has to travel through your breathing zone before it ever reaches outdoors.

Worse, hot plume rises, cools, stalls against the ceiling, then drifts with every door slam, box fan, and footstep. On a still day, or with a cross breeze pushing inward, an open door can hold fume inside longer than if you’d done nothing at all. Ventilation guidance from university EHS programs makes the same point industrial hygiene textbooks do: local exhaust ventilation — a hood positioned at the arc — captures fume before it dilutes into the air you’re standing in. Capture beats dilution, every single time.

The OSHA Rule That Describes Your Garage

OSHA doesn’t inspect hobby shops, but its welding standard is the best free design brief you’ll ever get. 1910.252(c)(2) requires mechanical ventilation whenever welding, cutting, or brazing is done in any of these conditions:

• In a space of less than 10,000 cubic feet per welder.
• In a room with ceilings lower than 16 feet.
• In an area where partitions, walls, or barriers restrict natural cross-ventilation.

A typical two-car garage runs 600 to 800 square feet. With an 8- or 9-foot ceiling, that’s roughly 5,000 to 7,000 cubic feet — under the volume threshold, well under 16 feet, and stuffed with barriers like benches, shelving, and a water heater. Your garage trips all three triggers at once. The standard then offers two routes: general mechanical ventilation (the familiar 2,000 CFM per welder figure) or local exhaust ventilation positioned to pull fume from the breathing zone. We’ll build on that second route.

Why Health Authorities Stopped Being Casual About This

For decades, welders measured exposure against OSHA’s total particulate PEL of 5 mg/m³ — a number with thin toxicological roots. NIOSH now describes welding fume exposure as a preventable occupational hazard and recommends limits several times stricter, reflecting follow-up research into respiratory and neurological effects. The International Agency for Research on Cancer classifies welding fume as carcinogenic to humans, and OSHA regulates individual constituents like hexavalent chromium separately. In plain language: this is not a nuisance odor. It’s a documented health risk, and every major health body’s position is the same — control it at the source.

What This Guide Covers

The sections that follow get specific: fume chemistry by process and base metal with an exposure-limit table; OSHA’s local-exhaust table translated into usable numbers; the capture-velocity and CFM math that decides whether your build works; fan, filter, and hood selection; a step-by-step wheeled source-capture unit; the mistakes that make homemade extractors useless or dangerous; and the smoke tests and maintenance schedule that prove it still works six months later.

Keep one principle in mind throughout: fume control is a capture problem before it is a filtration problem. Get the hood close and the airflow right, and everything downstream gets easier.

What Is in Welding Fume and Why It Harms You

Before you size a single fan, it helps to know exactly what you are chasing. Welding fume is not smoke in the campfire sense. When the arc melts the electrode and the base metal, a tiny fraction of that metal vaporizes, rises, and instantly condenses in the cooler air into an aerosol of ultrafine particles — mostly metal oxides, often under 0.1 micrometres across. That size matters: particles that small stay suspended for hours, drift with the slightest air current, and bypass the body’s natural filters to settle deep in the lungs. OSHA’s welding and cutting guidance and NIOSH’s welding safety overview both treat fume as a ventilation problem precisely because of that behaviour.

The Fume Recipe Changes With the Metal

The plume coming off your arc is a chemical signature of whatever is being melted, plus the shielding gas, the electrode coating, and any coatings or platings on the workpiece:

• Mild steel: mostly iron oxide with a side portion of manganese from the steel alloy and wire. Relatively low toxicity compared with what follows, but still the source of manganese exposure.
• Stainless steel: iron, chromium, and nickel. The chromium oxidizes in the arc to hexavalent chromium [Cr(VI)], the most concerning constituent a home welder routinely breathes. Any stainless work — a kitchen splashback, a bike rack, a boat transom repair — produces it.
• Galvanized steel: the zinc coating boils off first, generating zinc oxide fume. This is the classic trigger for metal fume fever.
• Aluminum: aluminum oxide fume plus, in MIG work, additives from the wire and possible coating residues. If you work aluminum, our Aluminum MIG Welding: The Complete Practical Guide covers prep and process details that also reduce how much fume you create in the first place.
• Brass and bronze: copper and zinc oxide — a reliable way to give yourself the “zinc shakes.”
• Coated, plated, or painted stock: whatever is on the surface ends up in the plume. Primers, galvanizing, anodizing, and old paint can add lead, cadmium, or phosphorus compounds to the mix.

Metal Fume Fever: The Acute Hit

Zinc oxide fume carries an OSHA permissible exposure limit of 5 mg/m³ as a ceiling — never to be exceeded, even briefly. Breathing it produces metal fume fever, historically called “brass chills” or “zinc shakes.” Symptoms typically come on 4 to 12 hours after exposure: a dry throat, chills, muscle aches, nausea, and sometimes a headache that peaks overnight and fades by morning. It is miserable rather than usually permanent, and it is almost entirely preventable with source capture. The tricky part is that the fume is invisible by the time it reaches you, so a shop can feel fine right up until it isn’t.

Manganese: The Slow, Quiet Risk

Manganese deserves special attention because the official limits disagree with each other by orders of magnitude. OSHA lists the manganese fume PEL as a 5 mg/m³ ceiling, while NIOSH sets its REL at 1 mg/m³ as an 8-hour TWA with a 3 mg/m³ short-term limit, and ACGIH puts its TLV at just 0.02 mg/m³ respirable and 0.1 mg/m³ inhalable. Long-term overexposure is linked to manganism, a parkinsonian condition with tremor, stiffness, and coordination problems — the kind of neurological damage that shows up after years, not after one afternoon.

Hexavalent Chromium and the Cancer Classification

OSHA sets the Cr(VI) PEL at 5 µg/m³ as an 8-hour TWA, with a 2.5 µg/m³ action level under 29 CFR 1910.1026. Stainless steel welding in a poorly ventilated shop frequently blows past those numbers within minutes of striking an arc. Cr(VI) is a confirmed lung carcinogen, and in 2017 the International Agency for Research on Cancer classified welding fumes themselves as carcinogenic to humans (Group 1), based on sufficient evidence for lung cancer, with stainless-steel welding fume specifically noted (see the IARC Monographs). That reclassification is why “I only weld occasionally” is a weaker argument than it used to be.

Exposure Limits at a Glance

Constituent	OSHA PEL	NIOSH REL	ACGIH TLV
Manganese fume	5 mg/m³ ceiling	1 mg/m³ TWA; 3 mg/m³ STEL	0.02 mg/m³ respirable; 0.1 mg/m³ inhalable
Hexavalent chromium	5 µg/m³ 8-hr TWA (2.5 µg/m³ action level)	1 µg/m³ TWA	0.2 µg/m³ inhalable
Zinc oxide fume	5 mg/m³ ceiling	5 mg/m³ ceiling	—

The practical takeaway is simple: ventilation is not a luxury upgrade. Capture velocity at the arc, adequate CFM, and real filtration — the subjects of the next section — are what keep these numbers on the right side of the table in a garage where no inspector will ever walk in.

The OSHA Ventilation Rules That Also Make DIY Extractors Work

You are not running a shipyard, so OSHA will not inspect the garage. That does not make its welding rules irrelevant. The airflow numbers in 29 CFR 1910.252(c) are the most useful engineering data you will find for a homemade fume extractor, because they answer the two questions that decide whether your build works: how much air must I pull, and how close must the hood sit to the arc?

OSHA gives a shop two ways to handle welding fume. Either move at least 2,000 cubic feet of air per minute past each welder to dilute it, or capture it at the source with local exhaust ventilation (LEV). A DIY extractor is the second option, and OSHA explicitly lets LEV stand in for the general-ventilation requirement. That substitution is why a well-built shop unit counts as a real engineering control rather than a fan with a filter taped to it.

Two Paths: Dilute the Fume or Capture It

Under OSHA 1910.252(c)(2)(ii), general mechanical ventilation must deliver a minimum of 2,000 CFM per welder unless the shop provides local exhaust hoods or booths, or puts welders on airline respirators. Read that “unless” carefully. It is the hinge the whole DIY concept swings on: the moment your hood pulls fume off the arc, the obligation to flush the entire room with 2,000 CFM per welder falls away.

It falls away for the right reason. Dilution ventilation only spreads fume thinner until it leaves the building, and it fails whenever a draft blows the plume across your face before it reaches a grille. Source capture works in the opposite direction: it grabs the fume in the first 12 inches, before it becomes the shop’s problem. For the same reason, general ventilation only counts if the make-up air itself is clean and comes from outside, as Virginia Tech’s EHS guidance on welding ventilation spells out. A shop fan blowing air around the room is neither.

OSHA’s Local Exhaust Table: Your Build’s Spec Sheet

Attached to the LEV allowance is a small table tying hood distance to required airflow and duct size. It is the most quoted piece of welding ventilation data in the trade, and it belongs on your workbench during the build:

Hood opening to arc	Minimum exhaust airflow	Duct diameter
4 to 6 inches	150 CFM	3 inch
6 to 8 inches	275 CFM	4 inch
8 to 10 inches	425 CFM	5 inch
10 to 12 inches	600 CFM	6 inch

Airflow values are OSHA’s; the nominal duct sizes follow from the standard’s 4,000 fpm transport basis. Confirm against the current standard text before cutting sheet metal.

Notice the pattern. Push the hood from 5 inches to 11 inches away and the requirement roughly quadruples. Fume does not drift in a straight line to your inlet; it expands into a widening cone, so the hood must swallow an ever-larger volume as standoff grows. This is why “I will just mount it across the bench and turn it up” fails. Distance is the most expensive variable in the design, and the one you can give away for free by simply moving the hood closer.

4,000 FPM: The Speed That Keeps the Duct Clean

The same table bases duct sizing on an air velocity of 4,000 feet per minute inside the pipe. Below roughly that speed, fume particles stop following the airstream and settle onto the duct wall. The consequences are annoying at best: your effective duct shrinks, the fan loses the flow it was rated for, and accumulated conductive dust in a duct running near a hot process is a fire hazard you built yourself.

Airflow equals velocity times cross-sectional area, so use that relationship when you size pipe. A 3-inch duct passes about 195 CFM at 4,000 fpm, a 4-inch about 350 CFM, and a 6-inch about 780 CFM. Go one size too large and velocity drops under transport speed; go too small and the fan fights rising pressure to deliver the CFM the table demands. Short, straight runs with minimal elbows are not a nicety, they are how you protect the number on the fan’s rating plate.

100 FPM: The Capture Velocity Floor

Distance and CFM are only half the standard. OSHA 1910.252(c)(3)(i) requires local exhaust hoods to hold a capture velocity of at least 100 feet per minute at the point of welding. That is the air speed measured right at the arc, not at the hood face and not at the fan outlet. A hundred feet per minute is about half a meter per second: a gentle breeze you can feel on the back of your hand held over the weld pool. It is the pass mark, not the goal, and heavier fume, an overhead joint, or a shop draft all push the practical requirement higher.

Five Design Rules for a Home Builder

1. Design backward from the arc. Pick your standoff, read the required CFM off the table, choose the duct that holds 4,000 fpm, then buy the fan. Buying the fan first is how shops end up with equipment that cannot do the job.
2. Keep the inlet within 6 inches when the joint allows. Every extra two inches of standoff costs roughly a hundred CFM of extra suction, a bigger motor, and more noise.
3. Keep ducts small, short, and straight. Transport velocity survives friction, bends, and filter boxes; long flexible runs with three elbows do not.
4. Never count the open garage door as ventilation. General ventilation under the standard is mechanical, sized per welder, and fed by clean make-up air. A breeze through the door captures nothing at the arc.
5. Match suction to the process. Cellulose-coated stick rods, flux-cored wire, and gouging sit at the heavy end, so an extractor used for Air Arc Welding and Carbon Arc Gouging needs to be built to the high side of the table. Aluminum is its own case, covered in Aluminum MIG Welding: The Complete Practical Guide. The capture physics stays identical; the fume composition does not.

None of this needs an industrial budget. It needs three numbers: 150 to 600 CFM depending on distance, 4,000 fpm in the duct, and 100 fpm at the arc. Everything you buy, cut, and bolt together in the next sections exists to hit those targets and prove you hit them.

Capture Velocity, CFM, and Hood Design: The Physics Behind Every Working Build

Every homemade fume extractor that actually works runs on three variables: capture velocity, airflow in CFM, and the distance between the inlet and the arc. Change one and the others have to follow. Get all three right and a modest inline fan will hold a MIG plume at the source. Miss one and you have built a noisemaker that shoves fume around the garage. This section turns the airflow math into shop-floor decisions, with worked numbers you can copy into your own build.

Distance Is the Enemy: Why CFM Climbs With the Square

Arc fume leaves the weld as a hot, buoyant plume. As it climbs it spreads and slows, so the volume of contaminated air you have to capture grows far faster than the gap itself. The working rule: required airflow rises roughly with the square of the distance from the arc. Double the gap and you need about four times the CFM.

OSHA’s local exhaust table for welding and cutting puts hard numbers on that rule. An inlet held 4 to 6 inches from the arc needs roughly 150 CFM. Slide the same inlet back to 10 to 12 inches and the requirement jumps to about 600 CFM — four times the air to do the identical job (OSHA 1910.252(c)).

The useful consequence is that the cheapest performance upgrade is never a bigger fan. It is a longer arm or flex hose that gets you three inches closer. A 150 CFM unit with its inlet 5 inches off the arc will outperform a 600 CFM unit parked a foot away, on a fraction of the power.

From Face Velocity to CFM: The One Equation You Need

Capture velocity is the air speed required at the arc to overwhelm the rising plume. Face velocity is the air speed across the mouth of your hood. The two are tied together by a single conversion:

CFM = face velocity (fpm) × hood opening area (square feet)

Fifty to one hundred feet per minute across the hood face is a common design starting point for capture hoods in industrial hygiene guidance (Virginia Tech EHS ventilation guidance). Hoods sitting inches from a welding plume typically run well above that band, because the convective updraft leaving the arc is strong.

Worked example. A tapered hood with a 10 × 10 inch mouth has an area of 100 square inches, or 0.69 square feet. At 100 fpm of face velocity it flows only about 69 CFM. To push the 150 CFM that OSHA’s table calls for at 4 to 6 inches through that same opening, you would need roughly 217 fpm at the face. Now run the equation backwards: 200 CFM at a gentle 100 fpm would require a 2-square-foot opening, around 17 × 17 inches — a hood too large to sit close to the work. The design lesson is clear: small hood, close inlet, higher face velocity, and a fan sized to match.

For light MIG and TIG work on mild steel, with the inlet 4 to 6 inches from the arc, a DIY unit delivering 100 to 300 CFM at the hood is genuinely adequate. Save the big-air setups for dirtier jobs. Aluminum MIG throws noticeably more fume per pass than mild steel (see our Aluminum MIG Welding: The Complete Practical Guide), and flux-cored wires are smokier still because the flux itself participates in the reaction.

Ducting and Static Pressure: Where Rated CFM Goes to Die

A fan’s catalog number is free-air CFM — output measured at zero resistance. The instant you add duct, elbows, a spark screen, and a filter, delivered airflow drops. Every one of those components adds static pressure, and every fan’s performance curve slopes downward as pressure climbs. That is why real delivered airflow is always lower than the rating on the box, and why experienced builders size the fan a step above their target so the finished system still makes its number.

Diameter matters twice over. Too small and friction losses eat the fan’s output. Too large and the air slows until sparks and heavy particles drop out of the stream and settle in the duct, where they are a fire hazard. Welding exhaust designs generally aim for duct velocities in the 2,000 to 4,000 fpm range:

CFM at the hood	Duct diameter	Approx. duct velocity
150	3 in	~3,050 fpm
250	4 in	~2,860 fpm
400	5 in	~2,930 fpm

Check the math yourself with the same equation: a 4-inch duct has a cross-section of about 0.087 square feet, so 250 CFM ÷ 0.087 gives roughly 2,860 fpm.

A few pressure-drop habits keep the run honest. Each 90-degree elbow costs about as much static pressure as 20 to 30 feet of straight duct, so use two 45-degree bends in its place. Smooth sheet metal outperforms ribbed flex hose, which belongs only on the short moving section of an arm. Keep runs short and straight, and don’t bury a fine screen directly in the airflow at the inlet — it catches sparks, but it blinds with spatter and adds pressure drop you will feel as lost CFM.

Work the three variables in this order: get the inlet as close as the process allows, read off the CFM your chosen distance demands, then select the fan, duct size, and filter package that can still deliver that number once real-world pressure drop is stacked on. A build that follows that order works. One that starts with the fan and works backward usually doesn’t.

Choosing the Components: Fans, Motors, Filters, Hoods, and Ducting

A DIY welding fume extractor is only as good as its weakest part. Most failed builds fail at component selection, not fabrication. Somebody buys a fan that looks impressive on the box, pairs it with a filter that only traps odor, and runs it through twenty feet of dryer duct. It moves air, sort of, and captures nothing. This section is about choosing parts that work together, so the finished unit performs the way the calculations in the previous section promised.

The Fan: Where Most Budgets Go Wrong

Three fan families are realistic for a home-shop extractor, and they are not interchangeable.

Fan type	Typical airflow	Best for	Where it fails
Inline mixed-flow duct fan	200–450 CFM	Portable single-arm extractors, most DIY builds	CFM collapses against filter pressure drop
Centrifugal blower (scroll)	300–1,000+ CFM	Fixed installs with long ducting and real filters	Heavy, bulky, expensive, louder
Bathroom exhaust fan	50–110 CFM	Almost nothing in welding	Undersized by 4–10x, motor not rated for particulate

Inline mixed-flow duct fans, the acrylic-body units sold for grow rooms and ventilation, are the most popular DIY extractor fan for good reason: they are cheap, compact, mount inline, and commonly deliver 200 to 450 CFM. The catch is the fine print. That number is a free-air rating measured with nothing attached. Bolt on a filter box, two elbows, a flex hose, and a hood, and you can easily lose half of it.

Motors: Shaded-Pole vs. EC and Brushless

Cheap inline fans and nearly every bathroom fan use shaded-pole motors: simple, rugged, inefficient, and with no real ability to hold speed under varying load. Step up to an EC (electronically commutated) or brushless DC motor and you gain three things that matter here. First, far lower running cost, since EC motors can be 60–70% more efficient. Second, true variable speed control via a 0–10V signal or a potentiometer, which lets you dial capture velocity at the hood instead of guessing. Third, longer bearing life under continuous duty, which matters if you weld for hours at a stretch.

If budget forces a shaded-pole fan, plan to control speed with a simple inline voltage controller and accept the noise penalty at full speed.

Filters: HEPA, Pleated Media, and the Carbon Trap

This is where DIY extractors earn or lose their claim to being a real engineering control.

HEPA media captures 99.97% of particles at 0.3 microns, the most penetrating particle size. Welding fume is not a single size; it is a cloud of ultrafine metal oxide particles, with abundant material both smaller and larger than 0.3 microns. Because HEPA performs well across the whole curve, it is the standard to aim for, either as a true HEPA filter panel or as a filter box packed with HEPA media. The trade-off is pressure drop: real HEPA is dense, and your fan must be chosen to push through it.

Pleated MERV-rated panels are the sensible middle ground. A MERV 13–16 pleated filter captures the bulk of welding fume at a fraction of the pressure drop of HEPA, costs less, and is easy to source in standard sizes. For mild-steel work in a ventilated shop, this is a defensible choice.

Activated carbon adsorbs gases and odors. It does not capture metal fume particles. Carbon alone, and especially the thin “activated carbon mesh” sheets sold as odor pads, is not a welding fume control. Use carbon only as a secondary stage after a particulate filter, and only if you want to knock down smell, not as your protection.

Hoods and Ducting Materials

For hoods, sheet metal is the right answer: it is spark-resistant, rigid, and easy to taper into a proper capture shape. Welded or riveted mild steel with high-temperature paint handles stray spatter that would melt plastic.

Ducting is where physics punishes convenience. OSHA 1910.252(c) ties hood distance to duct diameter in its local exhaust table, with ducts ranging from roughly 3 inches up to 5.5 inches, sized to keep transport velocity near 4,000 feet per minute so particles stay airborne in the stream instead of settling out and piling up inside your duct.

• Galvanized sheet-metal pipe: the best all-round choice. Rigid, spark-safe, smooth-bored, seals well.
• Rigid PVC or ABS: smooth and cheap, fine if you keep runs short and grounded, but it builds static and will ignite under the right conditions. Never use PVC for hot spark-laden exhaust without grounding and a spark trap.
• Flexible ducting: use only for the last few feet on an arm if needed. Ribbed flex destroys airflow, traps spatter, and is a fire hazard if it is the cheap foil type.

Controls, Gates, and Hardware

Blast gates at the hood let you shut off airflow to a capped inlet, keeping your CFM where you are actually welding. A speed controller on an EC or brushless fan lets you tune capture velocity. Put the fan on rubber vibration isolators so it does not turn the cart into a resonating buzzer. Add locking casters rated well above the finished weight, because a filter box, steel hood, and fan add up fast.

Note: If you plan to weld aluminum, remember that processes like Aluminum Flux Core Welding: What Actually Works in 2026 and Aluminum MIG Welding: The Complete Practical Guide generate fume profiles that deserve the same disciplined capture velocity you would give steel work. Component choices that survive steel spatter are the ones worth building around.

Step-by-Step: Building a Portable DIY Welding Fume Extractor

Theory ends here. What follows is a wheeled, floor-standing source-capture unit that delivers roughly 150 CFM at the hood — the airflow OSHA’s local exhaust table calls for when the inlet sits about six inches from the arc (OSHA 1910.252(c)). That covers light MIG, TIG, and stick work on mild steel in a home shop, and it fits in a weekend with a drill, tin snips, and roughly $200 in duct-aisle parts.

Subsystem	Target specification
Hood inlet	0.5 to 1.5 sq ft opening, flared, smooth inside
Face velocity at inlet	100–300 fpm (area × face velocity = CFM)
Delivered airflow	About 150 CFM with the inlet 4–6 in from the arc
Ducting	6-inch rigid metal, two elbows maximum
Fan	Inline blower, 250+ CFM free-air, must hold ~150 CFM at 1 in w.c.
Filter stack	20×25 in panel: MERV 8 pre-filter, true H13 HEPA, optional carbon layer
Controls	Speed controller plus a damper, GFCI power, bonded frame

Step 1: Fabricate the Hood

Cut 26-gauge galvanized sheet into a tapered transition that runs from your 6-inch duct up to a rectangular or shallow-canopy opening. Keep the taper gentle — an included angle under about 15 degrees — so the air stays attached instead of tumbling and losing pressure. File the inlet edges smooth.

Here is where most builders get the sizing backwards. A 0.5-square-foot inlet running at 100 fpm face velocity moves only 50 CFM, no matter how big the fan is. The opening is what converts fan airflow into real capture. For OSHA’s 150 CFM at the arc, you either enlarge the inlet toward 1.5 square feet and keep a lazy 100 fpm, or hold the 0.5 square foot opening and push it to 300 fpm. The smaller, faster inlet is handier around tight work like alloy wheel repair; a bigger canopy suits bench work.

Bevel the front of the hood at roughly 45 degrees so you can still see the arc through it, and bolt on a replaceable steel spark shield across the inlet. Spatter will chew through the hood itself within a few months, and a shield costs pennies to swap.

Step 2: Build a Lockable Arm

The arm is a safety component, not a convenience. A flex hose that sags into the arc is a fire and a ruined workpiece. Two workable options: a wire-reinforced flexible hose threaded over a support wire, or rigid pipe joints with hand-tight locking collars. Whichever you pick, the joint must hold position when you let go.

Mount the arm so the inlet can sit 4 to 6 inches from the weld zone, slightly above and upwind. Fume rises fast and cools into a buoyant plume; catching it within a few inches is what keeps the required airflow near 150 CFM instead of the 500-plus you would need at two feet.

Step 3: Mount the Fan on Isolation

An inline duct blower rated 250+ CFM free-air is the sweet spot. Free-air CFM is the number printed on the box when nothing is attached; once you bolt on a hood, three feet of duct, and a filter, expect something closer to half of it. Check that the published fan curve still shows your target 150 CFM at roughly 1 inch of water column static pressure.

Bolt the fan to the cart through rubber grommets or a foam isolation pad so motor vibration does not turn the sheet-metal filter box into a speaker. If you plan to recirculate, put the fan downstream of the filter: the blades stay clean, and any duct joint leak pulls air in instead of pushing fume out.

Step 4: Build an Oversized Filter Box

Build the box from plywood or 24-gauge steel, gasketed with foam tape and held with toggle latches for fast filter changes. Oversize it deliberately. A large panel-style filter has far more media area than a small round cartridge, which means lower face velocity through the media, dramatically lower pressure drop, and a fan that does not have to work twice as hard. Keep air speed through the panel under about 250 fpm.

Stack three stages: a cheap MERV 8 furnace filter as a pre-filter to catch spatter dust, a true H13 HEPA panel for the ultrafine metal oxide fraction, and an optional loose activated-carbon layer for odor. The NIOSH welding guidance treats HEPA-grade particulate capture as the baseline for a reason — the particles that hurt you are the ones you cannot see.

Step 5: Add Airflow Control and Power

Put a blast gate or butterfly damper in the duct, then add a plug-in speed controller rated for the fan’s amperage. The damper alone throttles by wasting energy as heat at the motor; slowing the motor cuts airflow, noise, and power draw together. Bond the cart to the plug ground with a short wire and run everything through a GFCI outlet.

Step 6: Decide — Outdoors or Recirculated

Exhausting outdoors through a short wall or window run is the simplest and most forgiving option: rigid metal duct, as short as possible, one elbow, terminating well away from any window, soffit, or fresh-air intake. If your work includes aluminum MIG welding, vent out and wear a respirator — aluminum fume is an irritant that filtration alone does not fully characterize.

Recirculation is fine in winter, but it is only as safe as the filter. It requires true HEPA-grade particulate media. A carbon-only pad, an “activated carbon mesh” screen, or a shop-vac cartridge does not capture ultrafine metal oxide, and per OSHA 1910.134 a half-mask with P100 cartridges remains your required backup regardless of what the machine does.

Assembly order matters: hood and arm first, then filter box, then fan, then controls, and only then ducting cut to length. Seal, gasket, and test the fan on the bench before you commit the cart to full assembly. The next section covers proving that it actually works.

The Mistakes That Make DIY Extractors Dangerous or Useless

A homemade fume extractor can be a genuine engineering control, or an expensive fan that moves hazard from one corner of the shop to another. The difference is almost never the sheet metal; it is a decision made early and never revisited. Here are the failure modes that show up in home-shop builds, and how to fix them.

Venting Into the Attic, Soffit, or Living Space

The most dangerous shortcut is treating the roof or the garage ceiling as a dump site. Welding fume is not like bathroom humidity. The particles are hygroscopic metal oxides; they grab moisture from the air and form a slightly corrosive film on whatever they settle on. Dump that into an unvented attic and you seed rust on nail heads, feed mold on sheathing, and create a dust reservoir that gets pulled back into the shop.

The fix is straightforward: exhaust outdoors through a wall or roof penetration, discharged away from doors, windows, and any fresh-air intake for your home’s HVAC system. If outdoor venting is genuinely impossible, you are left with filtered recirculation, which demands real HEPA media and honest airflow, not a compromise to make on stainless or coated steel. Virginia Tech’s EHS guidance is explicit: fume must be carried away from the welder’s breathing zone, not relocated indoors.

Trusting “Activated Carbon Mesh” as Your Only Filter

Those thin black pads sold for odor control are a trap. They hold a few grams of loose carbon dust spread across a mesh, with no dense particulate layer at all. Welding fume is overwhelmingly particulate, so a carbon-only pad captures neither the particles nor, at that loading, much of anything gaseous. The fume passes straight through: your shop smells different while your lungs get the same dose.

Use carbon only as a final stage behind a real particulate filter, and size it generously if you add it at all. When you work aluminum, the fume chemistry changes; our Aluminum MIG Welding: The Complete Practical Guide covers why that matters for prep and process choice.

Undersized Bathroom Fans and Optimistic CFM

A shaded-pole bathroom exhaust fan rated at 110 CFM free-air delivers a fraction of that once you bolt ducting and a filter box to it. Free-air rating is measured with nothing attached. Add 10 feet of flex duct, two elbows, and a loaded filter, and static pressure climbs while delivered CFM collapses. Bathroom fans are also not rated for continuous hot-air service or spark exposure.

Buy an inline duct fan or blower with a published fan curve, and size it for the system CFM measured at the hood, not the number on the box.

Sparks, Fire Risk, and Static Discharge

Welding spatter and hot slag travel. In a length of ducting they lodge in dust deposits or land on filter media, exactly the fuel-air mixture a fire needs. Install a spark arrestor screen upstream of the filter, use a metal (not plastic) housing, keep duct runs short and straight, and clean dust on a schedule. An airflow interlock that shuts the welder off when extraction stops is the professional-grade answer if you weld often.

Air moving through plastic duct and dry filters also generates static charge; bond and ground the ducting with conductor wire and a clamp to a known ground. OSHA’s 1910.252 ventilation and fire-prevention provisions apply in workplaces, and they translate directly to your garage.

Mounting the Hood Too Far From the Arc

This is the most common reason a build works on paper and fails in the shop. Capture velocity falls off sharply with distance; the OSHA distance table earlier shows required CFM roughly doubling for a modest increase in hood-to-arc distance. A hood idling two feet away captures almost nothing, however strong the fan. Get the inlet within 12 to 18 inches of the arc whenever the work allows.

No Speed Control, No Maintenance, No Backup Respirator

A unit with no damper or speed controller runs at one setting forever, usually too weak on thin material and wastefully loud otherwise. Worse, filters load gradually: airflow drops, fume escapes the capture zone, and nobody notices because the fan still sounds like it is running. Build in a variable controller, check airflow monthly, and change media on visible loading rather than a calendar guess. Weld stainless, coated, or plated material and the extractor becomes the primary control with a P100 half-mask as required backup. That ordering is not opinion: OSHA’s respiratory protection standard 29 CFR 1910.134 places engineering controls ahead of respirators. For magnesium and coatings, see our Alloy Wheel Welding guide for why they change the fume you breathe.

None of these mistakes need expensive parts, only one honest afternoon: where the exhaust goes, what the filter really is, what CFM survives the ducting, how close the hood sits, and whether anything ever gets cleaned. Do that and the build earns its place as real protection. Skip it and you have built a very loud decoration.

Testing, Tuning, and Maintaining Your Homemade Extractor

A finished build is not a working build. Plenty of home-shop extractors look the part — hood, arm, fan, filter box — and still let fume drift past the welder’s helmet because nobody measured anything. Verifying your unit takes about fifteen minutes and a few dollars of gear. Do it before your first real project, then repeat it every time you change a filter.

Start With the Smoke Test

The fastest check needs no instruments. Strike an arc (or hold a smoke pencil, incense stick, or smoldering twist of tissue where the arc would be) and watch what the plume does. A correctly sized system pulls the smoke straight into the hood from roughly five inches away. If the smoke curls sideways, climbs above the hood line, or disperses past ten inches, your capture zone is too small for the distance — exactly the relationship OSHA’s local exhaust table describes, where required CFM climbs steeply as hood distance increases.

Turn Face Velocity Into a Real CFM Number

Smoke tells you whether airflow exists; an anemometer tells you how much. A basic vane-probe or hot-wire unit costs little enough to justify for a one-time check. Hold it at the hood inlet, take several readings across the opening, and average them. Then use the one formula that governs the whole build:

CFM = face velocity (fpm) × inlet area (ft²)

Example: a tapered hood with a 6 × 12 inch inlet measures 0.5 ft². If the probe averages 900 fpm, you are moving 450 CFM — comfortably above the roughly 350–370 CFM OSHA associates with a hood kept 7 inches from the work. A reading of 300 fpm puts you at 150 CFM, where the unit only captures at very close range. That single number converts guesswork into a design verdict and gives you a baseline to compare against next month.

Measure after every major change too: adding a filter stage, lengthening the flex hose, or switching from straight sheet metal to ribbed ducting all move the number. Home-shop builds commonly lose 15–30% of fan rating to real-world pressure drop, which is why a fan advertised at 450 free-air CFM may deliver noticeably less once the filter box, two elbows, and a blast gate are in the loop. If the manufacturer publishes a fan curve, find the static pressure your system produces and read off delivered CFM rather than trusting the box.

Adjust With the Damper and Speed Controller

Once you have numbers, tune rather than rebuild. Open the blast gate fully during welding; crack it down if hood noise is unbearable or the airflow is starving other tools. If you fitted a variable speed controller, run full extraction while the arc is live, then drop to a low background speed afterward to filter residual haze without burning filter life. This two-stage habit is the cheapest way to extend cartridge life during an Aluminum MIG Welding: The Complete Practical Guide-style session where you weld in bursts with long setup gaps.

Filter Change-Out Signals

Filters load gradually, and your ears notice before your eyes do. Watch for two signals:

• Visible discoloration on the pleated media surface, especially a grey or rust-brown band down the airflow path.
• Falling airflow — hood face velocity drops noticeably, and the fan sounds quieter or higher-pitched as it works against higher static pressure.

Loading raises system static pressure, which lowers delivered CFM, which shrinks your capture zone. By the time smoke testing fails, the filter has been underperforming for weeks. When either signal appears, replace the media; do not blow it out with compressed air, since that drives particles deeper into the pleats and can fracture HEPA media.

Maintenance Checklist

• Weekly: smoke or incense capture test at 5 inches; clear the hood inlet of spatter and grinding dust.
• Monthly: anemometer reading at the inlet, compared against your baseline CFM; inspect hose arm joints and clamps for leaks.
• Every session: empty the pre-filter or spark trap, check that the damper moves freely, and confirm the cord and switch are undamaged.
• Each filter change: re-run the CFM calculation and record the date on the filter housing with a marker.
• Annually: check motor mounts for vibration isolation sag, clean fan blades, and re-verify static pressure against the fan curve.

Finally, testing matters more on some jobs than others. Aluminum, stainless, and coated plate all produce fume you would rather not breathe. For those projects, verify the extractor and wear a P100 respirator as backup rather than treating the hood as the only line of defense. A unit that passes today’s smoke test still needs re-checking after you move it, extend the arm, or swap filters.

DIY Extractor vs. Commercial Unit vs. Respirator: An Honest Comparison

By this point you can build a working fume extractor, but should you? A homemade wheeled unit, a store-bought portable extractor, and a half-mask respirator solve three slightly different problems, and most home shops end up owning at least two of them. Here is how they compare before you spend a weekend or a paycheck.

The Head-to-Head Table

Factor	DIY Extractor	Commercial Portable Unit	P100 Half-Mask Respirator
Typical cost	$150–$500 in parts	$1,000–$4,000+ with certified filters	$30–$70 for mask plus cartridges
Airflow / capture	100–300 CFM, hood near arc	300–800 CFM with articulated arm	Captures nothing; filters what you inhale
Portability	Good, if built on casters	Excellent engineered arm and balance	Maximum; it walks with you
Filter standards	Your choice: HEPA, MERV pleats, carbon	Certified cassettes with tested curves	NIOSH-approved P100, 99.97% at 0.3 micron
Maintenance	Entirely on you	Set filter-change intervals	Daily seal checks, cartridge swaps
Protects bystanders	Partially	Yes	No

The pattern is simple: extraction is source control, protecting everyone in the shop and clearing the air itself. Respiratory protection is last-line defense, protecting only the wearer, and only if the mask seals. Neither replaces the other, which is why industrial hygiene programs pair them.

When the DIY Extractor Is the Right Call

A home-built unit makes sense when your work is light, intermittent, and mostly mild steel: welding brackets, repairing lawn equipment, fabricating a cart. Plain carbon steel fume is far less aggressive than stainless or galvanized, and a 100–300 CFM hood held 6–12 inches from the arc captures most of what the arc makes. For occasional hobby use, a $300 build you actually maintain beats a $2,000 machine that sits in the corner.

The trade-offs are the ones flagged throughout: no certified filter performance, no tested capture curve, and no one to blame but your own judgment. If you check airflow with an incense stick and swap filters on schedule, that trade is reasonable. If not, you get the feeling of control with none of the certainty, which is worse than no protection at all.

DIY builds also shine brightest on steel. If your project list is drifting toward aluminum, as covered in our Aluminum MIG Welding: The Complete Practical Guide, the fume is lighter but the ventilation discipline still applies.

When a Commercial Extractor Earns Its Price

Commercial portable extractors typically move 300 to 800 CFM through a certified filter cassette on an articulated arm, and that combination is genuinely hard to replicate at home. The arm holds position without sagging, the media has a tested efficiency curve rather than a marketing claim, and many units include filter-loading indicators. You are paying for engineering validation, not just sheet metal and a fan.

Spend that money when you weld stainless or galvanized regularly, where hexavalent chromium and zinc oxide are in play; when you weld several hours a week rather than a month; or when the shop doubles as a living space that never gets aired out. Hexavalent chromium is regulated specifically by OSHA’s hexavalent chromium standard, and stainless fume is the strongest single argument for certified filtration or a respirator, ideally both. The same applies to galvanized work and to processes like Aluminum Flux Core Welding: What Actually Works in 2026, where flux chemistry adds its own reaction products to the plume.

Where the Respirator Fits

A NIOSH-approved half-mask respirator with P100 filters is the standard respiratory backup for welders. It supplements source capture; it does not replace it. Extraction keeps fume out of the room, out of your lungs when you lift your hood to inspect a bead, and out of the air your family breathes if the shop is attached to the house. A respirator only protects you while the seal holds, and a beard breaks that seal instantly.

Reality check for anyone who welds for money: OSHA’s respiratory protection standard, 1910.134, requires a written program with medical evaluation, fit testing, and training whenever respirators are required in a workplace. A hobbyist welding alone on Saturday is not covered, but the medical logic does not change. If you reach for the mask every session, your ventilation is undersized, not the mask working correctly. NIOSH’s welding safety overview frames respiratory protection as one layer in a hierarchy of controls, with engineering controls like extraction ranked above it.

A Practical Decision Rule

Short version: if you weld mild steel a few times a month, build the DIY unit and keep a P100 half-mask on the bench for grinding, overhead welds, and anything you cannot capture cleanly. If you weld stainless, galvanized, or more than a few hours a week, buy a certified extractor and still wear the mask. If you will do neither, weld outdoors with the wind at your back, wear the respirator, and stop pretending a garage door is ventilation. The OSHA welding, cutting, and brazing requirements exist because the fume does not negotiate, and neither should you.

Conclusion: Clean Air Is a Build You Finish, Not a Fan You Buy

The real test of a welding fume extractor is not its price, its noise level, or the sticker CFM printed on the box. It is whether the fume reaches the inlet before it reaches you. Strike an arc, hold a smoke source or an incense stick near the hot zone, and watch: if the smoke bends toward the hood and disappears, your build works. If it drifts up into your face, nothing else you did matters. Every rule in this article exists to make that one smoke test pass — and to keep it passing a year from now.

The Six Rules That Decide Whether It Works

Strip away the component talk and a working DIY welding fume extractor comes down to six decisions you now already know how to make:

• Capture close to the arc. You need a minimum capture velocity of roughly 100 fpm right at the point where fume is generated. In practice, holding the hood 4 to 6 inches from the arc calls for about 150 CFM through a hood of ordinary size. Distance is the enemy: required airflow climbs steeply every few inches you move away, which is why an articulated arm beats a ceiling-mounted box every time.
• Size airflow to the distance table, not the fan’s marketing. OSHA’s local exhaust table in 1910.252(c) links hood distance to required CFM and duct diameter. Design to it, then subtract real-world pressure drop from bends, filters, and grilles before you believe the fan will deliver.
• Keep duct transport velocity at about 4,000 fpm. Below that, metal particulate settles out of the airstream, lines the duct, chokes your airflow over time, and creates a fuel source you do not want sitting near a hot process.
• Filter particulate with real HEPA-grade media. Welding fume is ultrafine metal oxide, not soot. Activated carbon is an odor polish, not a particulate control — carbon-only pads and “activated carbon mesh” let the harmful fraction straight through.
• Vent outdoors if at all possible. Discharging to the outside removes the filtration burden entirely. If your shop layout forces recirculation, HEPA-grade particulate filtration is non-negotiable, and a carbon stage only rides along behind it.
• Control sparks before the duct. A spark screen at the inlet, no plastic flex hose in the hot path, and a clear zone around the exhaust outlet keep a good build from becoming a fire investigation.

When DIY Is Not Enough on Its Own

A properly built source-capture unit is a legitimate engineering control, and it is far better than nothing. It is also not a blank check. Welding on stainless produces hexavalent chromium, and galvanized and coated steels push manganese and zinc oxide into the plume — these two jobs demand the strictest controls you can put in place. The IARC classifies welding fume as carcinogenic to humans, so treat the hood as your first line and a fit-tested P100 half-mask respirator as your backup, never a substitute.

The same caution applies in confined spaces, overhead work, and anything where the plume can outrun the hood. Material matters too: if aluminum is your regular project, our Aluminum MIG Welding: The Complete Practical Guide covers the technique side you should pair with ventilation; if you run stick electrode on alloy steel, see the Alloy Steel Welding Electrodes: Selection, Use & Safety guide for filler choices and their fume implications; and if you are moving into exotic alloys or Aerospace Welding: Processes, Standards, and Careers Guide, assume certified extraction and respiratory protection are required from day one.

Your Action Checklist

• Set the hood 4 to 6 inches from the arc and confirm roughly 150 CFM at that distance.
• Verify at least 100 fpm capture velocity at the joint with a vane probe or anemometer.
• Confirm 4,000 fpm transport velocity in the duct so particulate stays suspended.
• Install HEPA-grade particulate media; add carbon only as a second stage for odor.
• Vent outdoors where possible; document recirculation filtration if you cannot.
• Fit a spark screen, vibration isolation, a damper or speed controller, and a lockable arm.
• Run the smoke test, log the reading, and repeat it after every filter or duct change.
• Keep a P100 respirator on the hook for stainless, galvanized, overhead, and confined work.
• Track filter loading and airflow drop, and set a fixed filter-change schedule.

Clean air in a home shop is not something you buy once and forget. It is a build you finish, verify, and maintain — and that habit, more than any single fan or filter, is what lets you hang up your helmet at the end of the day without a scratchy throat and a headache to show for it.

Frequently Asked Questions

How much CFM does a DIY welding fume extractor need?

For light MIG or TIG work with the inlet 4 to 6 inches from the arc, OSHA’s local exhaust table calls for about 150 CFM with a 3-inch duct. Moving the inlet to 8 to 10 inches raises the requirement to roughly 425 CFM, so proximity matters far more than raw fan power.

Can a box fan with a furnace filter work as a welding fume extractor?

A box fan with a pleated filter can act as a background air cleaner to capture fume that escapes your main extraction, but it lacks the capture velocity to pull fume directly off the arc. Use it as a supplement, never as source capture.

Should a DIY welding fume extractor vent outdoors or recirculate?

Venting outdoors through a short wall penetration or window is the simplest and most reliable option, and it matches OSHA’s preference for exhausting contaminant-laden air out of the workspace. If you must recirculate, the unit needs true HEPA-grade particulate filtration; activated carbon alone captures neither metal fume particles nor carbon monoxide.

Does activated carbon remove welding fume?

No. Activated carbon adsorbs gases and odors, but welding fume is a particulate aerosol of metal and metal oxide particles that requires mechanical filtration such as HEPA media. Carbon can be added as a final stage for smell, but only after a particulate filter.

How close should the extraction hood be to the arc?

Four to six inches is the practical target. OSHA’s table shows required airflow quadrupling from about 150 CFM at that distance to 600 CFM at 10 to 12 inches, so a modest fan placed close beats a powerful fan placed far away.

Is welding fume really dangerous in a home garage?

Yes. Metal fume fever from zinc and galvanized steel, manganese exposure linked to neurological effects, and hexavalent chromium from stainless welding are all realistic risks, and IARC classifies welding fumes as carcinogenic to humans. Garages are worst-case environments because they are small, often enclosed, and have poor cross-ventilation.

Can a DIY extractor replace a respirator?

A well-built extractor is the primary engineering control, but a NIOSH-approved half-mask respirator with P100 filters remains essential backup, especially when welding stainless, galvanized, or in confined spaces, and whenever you are out of position and the hood can’t be placed close.

How often should I change the filters in my homemade extractor?

There is no fixed interval — it depends on arc hours and metals welded. Watch for visible darkening of the media and a noticeable drop in airflow at the hood; both indicate the pre-filter and HEPA stage are loaded and static pressure is climbing.

Is it safe to exhaust welding fume into my attic or through a window into the house?

No. Metal oxide fume deposits in attics and living spaces create corrosion, dust, and re-entrainment problems, and the fume simply returns to your breathing zone. Exhaust outdoors, discharging away from windows, doors, and the home’s fresh-air intake.