Proper MIG Welding Technique: Settings, Angles, and Steps

Two beads off the same machine, same wire, same gas roll off the feeder, and one gets cut off the frame while the other gets a gold-sticker stamp. The difference was not the welder, the wire, or the machine. It was how the gun was held, how fast it moved, and how much stickout was left hanging out of the nozzle. That gap between “looks welded” and “is welded” is exactly what proper MIG welding technique closes, and it is a set of learnable numbers and habits rather than some secret handed down by old hands.

What Proper MIG Welding Technique Actually Means (and Why Good Beads Get Cut Off)

Two beads off the same machine, the same wire, the same gas roll. Same amperage, same voltage, same joint. One came off the plate with a clean, flat, slightly convex profile and fused nicely at both toes. The other looked like a cold, lumpy caterpillar that a wire brush could peel off in pieces. When the second one was cut out of the frame, the inspector stopped asking about the machine and started asking about the welder.

That gap between “looks welded” and “is welded” is exactly what proper MIG welding technique closes, and it is a set of learnable numbers and habits rather than some secret handed down by old hands.

Here is the thing worth understanding up front: the machine owns the parameters, and the operator owns the technique. Voltage, wire feed speed, amperage, and gas flow come off the machine. But the work angle, travel angle, travel speed, stickout, contact-tip-to-work distance, body position, and joint prep come off you. And the operator-side variables are what turn a passable bead into a sound one.

What MIG Welding Is, in Plain Terms

MIG welding, formally gas metal arc welding (GMAW), feeds a continuous solid wire electrode through a welding gun into the weld pool while shielding gas protects the pool from contamination, according to Miller Electric. That is the whole process in one sentence, and the reason it took over shops is range. The MIG process is practical from 24-gauge material up to 1/2-inch thick, which is a large part of why it dominates hobby, farm, and general fabrication work.

Need a side-by-side comparison before you commit? Our look at where stick welding still wins is worth a read, but for most people choosing a first process, MIG wins on speed and on material thickness coverage.

The Unifying Idea: Heat Input

Almost every defect in this article traces back to one number. If you can calculate it, you can predict the bead.

Heat input is calculated as (60 x Amps x Volts) / (1,000 x Travel Speed in IPM) = kJ/in, and amperage combined with travel speed has the largest effect on it, as explained by The Fabricator. Read that formula as a statement of consequence: more amps or more volts means more heat dumped into the joint, and faster travel means less heat per inch.

So the arc sounds wrong, the bead is too flat, the toes do not wet out, the plate distorts, and there is porosity in the root. All four are heat-input symptoms wearing different clothes. Once you see that, the rest of this article becomes a diagnostic tool instead of a list of rules to memorize.

Two more practical anchors from Miller’s parameter guidance: a sound MIG bead should be no larger than the thinnest section of metal being welded, and the arc should stay on the leading edge of the puddle so molten metal does not run ahead of the heat source. Those two habits alone fix more bad beads than any change to the machine.

“Proper” Means Reproducible, Not Impressive

Here is the professional standard in one breath: a qualified welder under AWS D1.1 works to a written Welding Procedure Specification (WPS) with recorded parameters, not to feel. The procedure says what voltage, what wire, what gas, what travel speed, what position. The welder executes it. Another welder with the same procedure should get the same result.

That is the gap between an attractive bead and a sound one. Under AWS D1.1, Table 8.1 sets visual acceptance criteria including no cracks, complete fusion at the toes, and a limit on undercut of roughly 1/32 inch on most joints. Confirm current values against the code edition in force in your area, since acceptance criteria get revised.

A pretty bead can be convex, slightly rippled, and shiny, and still have incomplete fusion at the toes or an undercut that eats into the base metal. A sound bead might look plainer. The difference is not in the look, it is in the penetration profile and the parent metal.

What You Will Walk Away With

By the end of this guide you will have four things:

• A repeatable settings recipe for carbon steel, with the numbers written down.
• A way to read your own bead, so you can tell heat input problems from gas problems from angle problems.
• A defect-to-cause table you can screenshot and use at the bench.
• The safety rules that are not optional, anchored in the actual OSHA regulation.

Before any of that works, though, you need the framework. Let us start where most shops actually start: the machine settings, and why voltage and wire feed speed are not two separate dials.

Machine Settings: Voltage, Wire Feed Speed, and Amperage Working Together

Most MIG problems are not mysteries. They happen because two dials that look independent are actually joined together. On a constant-voltage wire-feed machine, voltage controls arc length, while wire feed speed controls how fast the wire melts into the puddle, and that feed rate is what determines your amperage. As ESAB explains, wire feed speed governs the amperage on this type of machine.

That single fact explains why so many operators chase erratic results. Turn the voltage up to flatten a rough bead, and the arc length changes. Turn the wire feed speed up to fill a gap faster, and now the amperage climbs with it, the puddle gets bigger, and the bead changes shape again. You adjusted two things while thinking you adjusted one.

The Starting-Point Math You Can Use Tonight

You do not need a spreadsheet to get in the ballpark. Miller Electric publishes a simple rule for steel: each .001 inch of material thickness needs roughly 1 amp of output. One-eighth-inch material needs about 125 amps. Quarter-inch plate needs about 250 amps.

From there, pick the wire that can carry that current:

Wire diameter	Typical amperage range	Best fit
.023 inch	30–130 amps	Thin sheet, light fabrication
.030 inch	40–145 amps	General purpose mild steel
.035 inch	50–180 amps	Heavier plate, structural work
.045 inch	75–250 amps	Thick plate, production shops

Next, convert amps into wire feed speed with Miller’s multipliers, which give inches of wire per minute for every amp of output:

• .023-inch wire: multiply amps by 3.5
• .030-inch wire: multiply amps by 2
• .035-inch wire: multiply amps by 1.6
• .045-inch wire: multiply amps by 1

Work the example: 1/8-inch steel at 125 amps with .030-inch wire. Multiply 125 by 2 and you land on 250 IPM of wire feed speed. That is your number. Set it, strike an arc, and adjust from there.

Thin Material and Thicker Stock

Thin sheet behaves nothing like plate, and Hobart Brothers publishes the numbers to prove it. 24-gauge material welds at 30–50 amps with a travel speed of only 10–20 IPM. On that same thin material, .024-inch wire runs at 13–15 volts and 130–160 IPM, while .030-inch wire runs at 15–17 volts and 75–100 IPM. Notice the flip: the thicker wire needs more volts but far less feed speed.

Step up to a 3/16-inch joint and the numbers jump again. With .035-inch wire, Hobart lists 140–175 amps, 19–22 volts, 240–290 IPM of wire feed and 14–19 IPM of travel.

Polarity and the Voltage Drop Trap

Polarity is a setting people forget, and getting it wrong fails quietly. Solid MIG wire runs on DC electrode positive (DCEP), which produces deeper penetration than DCEN. Some self-shielded flux-cored wires require DCEN, the same convention used in stick welding, which is one reason stick welding still earns its keep on site work and heavy restoration jobs.

Then there is cable length. The Fabricator points out that a power source set to 25 volts may actually deliver only 23 volts at the arc on a long lead. That shortfall shows up as cold lap, where the filler metal does not fully fuse with the base metal at the weld toes.

Read Your Machine by Its Symptoms

Rather than guessing, use what the arc tells you. Miller’s symptom guide is the fastest diagnostic in the shop:

Symptom	Likely cause	First move
Poor arc control, inconsistent penetration, turbulent pool	Voltage too high	Turn voltage down 1–2 volts
Poor arc starts, ropey bead, excessive spatter, cold lap at the toes	Voltage too low	Raise voltage, check cable length
Wide bead, burn-through, heavy spatter	Wire feed speed too high	Reduce WFS, recheck amperage
Narrow convex bead, poor toe tie-in	Wire feed speed too low	Raise WFS

Heat Input and Travel Speed: The Two Numbers That Decide Everything

Strip away the wire, the gas, the gun angle, and the machine settings, and nearly every MIG defect still traces back to one measurement: how much heat your arc delivers to the metal for each inch of weld. Get that number right and the rest of your technique has solid ground to stand on. Get it wrong and you will spend an afternoon chasing symptoms that all share the same root cause.

The measurement is called heat input, and the math is short enough to run in your head at the bench. The Fabricator publishes the working version most shops use:

Heat input (kJ/in) = (60 × Amps × Volts) ÷ (1,000 × Travel speed in IPM)

Amps and volts come off the machine. Travel speed is the one number you control second by second, which makes it the most powerful lever you have and the easiest one to get wrong.

Three Times the Heat, Same Machine

Watch what happens when only one variable changes. Same wire, same gas, same 200 amps and 20 volts, two different travel speeds:

Settings	Travel speed	Heat input	Likely result
200 amps, 20 volts	5 IPM	48 kJ/in	Wide, soft bead, deep melt, burn-through risk on thin steel, heavy distortion
200 amps, 20 volts	15 IPM	16 kJ/in	Narrow bead, quick weld, but light penetration and poor toe tie-in

Same machine, same wire, same gas. Crawl along at 5 IPM and you pile roughly three times the heat into every inch of joint. That is the difference between a sound weld and a melted, twisted mess on the same part.

According to The Fabricator, amperage combined with travel speed has the largest effect on heat input of any variable pair. Voltage contributes as well, which is why Hobart Brothers warns that setting voltage and amperage too low drops heat input and produces a ropey bead appearance, lack of penetration, and spatter, while too high voltage gives you undercut and a concave bead. Neither end of the dial is free.

How to Pick a Travel Speed That Works

You do not need a calculator at the torch. Three habits keep you in the right window:

• Match the pool to the joint. The size of the puddle you are dragging along should correspond to the thickness of the metal you are joining. Bigger pool, bigger heat, deeper melt.
• Keep the bead no larger than the thinnest section of the joint. On a lap joint or a transition in thickness, the smallest member sets the limit. A bead wider than the thinnest section is a bead with no place to land.
• Keep the arc on the leading edge of the puddle. If the heat source trails behind the molten pool, metal runs ahead of you, the bead gets ahead of the arc, and you lose control of both penetration and shape.

Too Fast, Too Slow

Travel too fast and you get a narrow, convex bead with inadequate toe tie-in, insufficient penetration, distortion, and an inconsistent bead from one pass to the next. Miller Electric and Bernard and Tregaskiss both describe the same narrow, humped profile that refuses to wet out at the edges. The toes will undercut or leave a cold gap, and the bead sits on top of the joint instead of fusing into it.

Travel too slow and the opposite happens: you force too much heat into the joint. The bead widens, the penetration profile flattens out, and on thinner material you invite burn-through. Miller points out that slow travel combined with excess heat pushes distortion up as well, because all that energy has to go somewhere.

Why Distortion Is the Expensive Mistake

A cosmetic bead can be ground off. A crooked part cannot. Hobart Brothers notes that excessive heat input from high voltage, high amperage, slow travel speed, or long welds leads to burn-through and distortion, and that distortion degrades fit-up, which leads to downtime and rework. Controlling heat on the front end is almost always cheaper than straightening a warped frame on the back end.

A Worked Example

For 1/2-inch A36 mild steel, The Fabricator gives a published recipe of roughly 250 amps with 90% argon / 10% CO2 in spray transfer, at about 26 to 28 volts and 375 to 420 IPM wire feed speed. Run that recipe, hold your travel speed steady, and the heat input becomes a predictable number instead of a gamble.

Want every related number gathered in one place, from machine settings to gas and wire choices? Our welding equipment and shielding gas guide covers the hardware that feeds the arc.

Gun Geometry: Work Angle, Travel Angle, and Push vs. Pull

Most of what you tune on a MIG machine is a number you dial in. These two are numbers you hold. Work angle and travel angle decide where the arc lands relative to the joint, and they shape the bead more than most welders expect. Get them right and filler flows into the toe of the joint. Get them wrong and you get spatter, undercut, or a bead that looks like it landed on the plate instead of melting into it.

Work Angle vs. Travel Angle: Two Different Measurements

Two terms cover gun geometry, and mixing them up is common. The work angle is the relationship between the electrode axis (the line running down the center of the wire and contact tip) and the surface of the workpiece. The travel angle is the position of the gun relative to perpendicular, which in practice is the decision to push the gun or drag it. This is the convention the American Welding Society uses in the Welding Handbook, and it is explained cleanly in Bernard and Tregaskiss’s MIG welding basics.

In plain terms, the work angle answers “what direction am I pointing the torch at the joint?” The travel angle answers “am I walking the torch into the weld or dragging it out of the weld?”

Work Angles for Common Joints

Miller’s guidance for MIG welding basics on mild steel breaks it down by joint type in the flat position:

Joint (flat position)	Work angle	What you are aiming at
Butt joint	90 degrees	Torch square to the plate, splitting the joint equally
T-joint	45 degrees	Bisecting the fillet corner between the two plates
Lap joint	About 60 degrees	Pointed into the corner of the lap; the angle grows toward 70 degrees as the thicker member gets thicker

That last row is worth remembering. A lap joint made of two thin sheets and a lap joint made of a thick base plate are not the same problem. As the thicker member gets thicker, you rotate the gun further over so you are still driving heat into the joint rather than melting the face of the top sheet.

Travel Angle and the Push vs. Pull Decision

For normal welding in all positions, the travel angle runs between 5 and 15 degrees. Push beyond 20 to 25 degrees and the arc starts to misbehave: more spatter, less penetration, and general arc instability. That is not a fine-tuning window, it is the point where the arc starts spitting the wire and fighting you.

Once you settle on 5 to 15 degrees, you have two options for which way the torch leans:

	Drag (backhand)	Push (forehand)
Penetration	Deeper	Shallower
Bead shape	Narrower, with more buildup	Wider and flatter
Visibility	Arc sits ahead of the joint, harder to see	Better view of the joint and the wire
Wire control	Less control of the wire into the joint	Better control, wire stays aimed where you want it

The honest rule of thumb: push for thin material and aluminum, drag for thicker steel and self-shielded flux-cored wire where you want the slag to run out of the weld rather than sit on top of it. New to aluminum? Our complete guide to aluminum MIG welding covers why the flat, wide push bead is easier on that material than a narrow drag bead. And whatever you are welding, the surface still has to be clean first, so prepping aluminum properly decides whether any of this holds together.

How the Angles Change Out of Position

Flat is the reference. Everything else is an adjustment to that.

Horizontal: Keep the same travel angle you used flat, but drop the work angle by 0 to 15 degrees. That slight downward tilt stops filler metal from sagging out of the joint or rolling onto the bottom side. Voltage and amperage usually stay the same as flat or drop slightly.

Vertical: On a vertical T-joint, use a work angle of just over 90 degrees to the joint. The torch stays nearly upright but leans enough that the wire feeds into the root of the joint rather than straight down onto the puddle.

Ergonomics: Technique That Does Not Wreck Your Body

Gun geometry is also a body-position problem, and Bernard and Tregaskiss make the case that technique which hurts your body does not survive the week. Their recommendations are worth building into your setup:

• Locking trigger to prevent trigger finger, the compressed tendon condition welders pick up from squeezing the trigger for hours.
• Rotatable gun neck so you can reach awkward joints without contorting your wrist.
• Hands at or just below elbow height, with the work positioned between your waist and shoulders.
• Rear swivels on the power cable so the cable does not tug your arm back as you move.
• Wrist-neutral handle and neck angles, keeping the forearm lined up with the hand instead of bent back at the joint.

Wire Stickout, Contact-Tip-to-Work Distance, Shielding Gas, and Polarity

Proper MIG welding technique often hinges on three numbers that get buried in almost every troubleshooting conversation: wire stickout, contact tip to work distance, and gas flow. None of them show up on the machine’s display, and none of them change when you turn a knob. Yet any one of them, out of spec, will wreck a bead that your voltage and wire feed settings would otherwise handle perfectly.

Stickout: How much wire is doing the work

Stickout, also called electrode extension, is the length of wire sticking out past the contact tip. Miller Electric recommends about 3/8 inch for mild steel and gives welders a trick for checking it: listen for a sizzling bacon sound as the wire feeds, which is the wire melting at the contact tip. An irregular or chattering arc is very commonly caused by stickout that is too long.

Different transfer modes want different lengths. Bernard and Tregaskiss split it cleanly: 1/4 to 3/8 inch for short circuit transfer to limit spatter, and about 3/4 inch for spray transfer. ESAB gives you a diameter-based approach instead: 3/8 to 1/2 inch for small-diameter wire such as .030 inch, and 5/8 to 3/4 inch for larger diameters such as 1/16 inch.

Length matters because of plain physics. A longer stickout adds resistance to the circuit, which drops the current your machine is trying to deliver and fights the voltage setting you dialled in. Hobart Brothers spells out the consequences for small-diameter wire: keep stickout at 5/8 inch or shorter, because a long stickout lowers amperage, cuts penetration, prevents the weld from wetting out and filling the toes (cold lap), and robs the joint of shielding gas coverage. Three problems, one cause.

Contact tip to work distance is a separate measurement

CTWD is the gap between the contact nozzle tip and the workpiece, and it is not the same thing as stickout. Keep the gun at a consistent distance: about 3/4 inch suits spray transfer, while 3/8 to 1/2 inch suits short circuit. Too short in spray causes burnbacks, and too long invites weld discontinuities from poor gas coverage. The nozzle is doing the work of directing argon over the puddle, so the moment you drift out of position, you have weakened the shield. If you are shopping for a new gun setup or consumables, the welding equipment guide from Airgas is a good place to see what nozzle lengths and gun options are actually out there.

Shielding gas: the cheapest performance lever you have

Set shielding gas flow to 20 to 25 cubic feet per hour and check for hose leaks with a soapy water solution. Every leak is shielding you do not have, and a cylinder reading fine at the gauge can still deliver an unshielded weld if the hose is cracked or a fitting is loose.

For carbon steel, Miller calls 75% argon and 25% CO2 (C-25) the best all-purpose shielding gas: the least spatter, the best bead appearance, and it will not promote burn-through on thinner metals. Pure CO2 gives deeper penetration but noticeably more spatter and a rougher bead. Hobart Brothers adds a practical argument for 90% argon and 10% CO2: one gas can be standardised across spray, pulsed-spray, and short circuit applications, so a shop runs one cylinder instead of juggling three. If you are still deciding what to order, the AllGas welding supply guide covers cylinder and service options.

Polarity: a setting you cannot see

Solid MIG wire runs DC electrode positive (DCEP), which gives deeper penetration than DCEN. Many self-shielded flux-cored wires require DCEN. Get it backwards and you get erratic results that look like a machine fault rather than a setup error.

Keep the gun clean or forget every number above

Spatter on the contact tip, a worn liner, and a clogged nozzle will destroy a good setting faster than any wrong number. Removing spatter from contact tubes, replacing worn contact tips and liners, and discarding rusty wire are routine setup steps, not optional extras.

Symptom	Likely Cause	First Change to Make
Chattering or irregular arc	Stickout too long	Trim to 3/8 in and re-check
Cold lap, unfilled toes	Long stickout, low amperage, dirty metal	Shorten stickout, clean to bare metal
Burnbacks in spray	CTWD too short	Hold about 3/4 in from the work
Porosity, pinholes	Low gas flow, leak, wind, clogged nozzle	Set 20-25 CFH, soapy-water test the hose
Excessive spatter, rough bead	Pure CO2, gas mix wrong, long stickout	Switch to C-25, shorten stickout
Poor penetration on solid wire	Polarity set to DCEN	Switch to DC electrode positive

Joint Preparation, Fit-Up, and Getting Metal Clean Enough to Weld

Most failed MIG welds were never arc problems. Voltage, wire, and machine settings were fine. What failed was the prep work that happened before the trigger got pulled: rust left on the plate, a clamp biting into mill scale, tacks made with the wrong settings, a root opening too tight for the thickness. Because the symptoms show up later, in the finished weld, operators usually blame the machine.

Miller’s MIG welding basics guide puts it bluntly: unlike stick electrodes and flux-cored wire, solid MIG wire does not combat rust, dirt, oil, or other contaminants very well. The shielding gas protects the molten pool from the air, not from the scale you left on the steel. Clean down to bare metal before you strike an arc.

Clean metal and a clamp that can make contact

The work clamp deserves its own attention, because beginners almost always misdiagnose it. A clamp on mill scale, paint, rust, or a dirty machine surface adds impedance to the circuit. Arc voltage sags, current delivery drops, and the wire starts feeding unevenly. The welder swaps drive rolls, liners, and guns. The actual fix was a scuff pad on the clamp face and a move to clean metal a few inches from the joint, clear of heat and spatter.

Two more items from the same Miller guidance: check that all cable connections are tight and free of fraying or damage, and keep gun consumables clean. Spatter on the contact tip, a worn contact tip, a gunked liner, and a half-blocked nozzle will destroy good settings every time. Remove excess spatter from the contact tube, replace worn contact tips and liners, and toss any spool of wire that looks rusty. For aluminum the prep rules change completely, and this guide to prepping aluminum for welding explains why a dedicated aluminum cleaner matters instead of a wire brush.

Joint design, bevel, and root opening

On thin material, a square butt joint with a small, consistent gap is all you need. As the material thickens, the joint has to be opened up. Miller’s guidance is to bevel thicker metal so the weld fully penetrates to the base metal, and that matters most on butt joints where two edges meet face to face. Without a bevel, you are asking a narrow bead to melt through the entire thickness, which realistically means burnthrough or a cold, unfused center.

Groove angle, groove face (the flat land at the bottom of the bevel) and root opening (the gap between the two edges) are not guesswork in a coded shop. AWS D1.1 Table 8.2 specifies minimum groove face and root opening dimensions for each joint type. Use the numbers in your own procedure and check the code edition in force where you work. For aluminum, where heat conductivity punishes a sloppy bevel, the aluminum MIG welding guide covers how joint design changes by alloy and gauge.

Tacks are welds too

Fit-up mistakes become permanent at the tack stage. Tacks get ground out before the real pass, and they are where undercut and slag traps start. Tack with the same machine settings you plan to use for the production bead instead of a quick low-voltage stab to save a minute. Keep tacks small, chip off any slag, and clean the tack area before welding through it. A wire wheel or stiff brush gets you close, but what you really want is a smooth, bright surface where the final bead will sit.

Multipass discipline

On multipass welds, weave beads may be used to fill the joint, and a slight hesitation at the top toe helps prevent undercut and ensures proper tie-in to the base metal. That hesitation is deliberate: you slow down just enough at the top corner for the puddle to wet out into the toe rather than racing past it and leaving a sharp notch.

The trap is overdoing it. A wide weave spreads the arc over more area, which lowers the effective heat input per inch of joint. Push it too far and lack of fusion shows up instead of undercut, because the puddle is never concentrated enough to fuse the sides of the groove. Keep the weave to what your wire diameter can wet out, usually a slight side-to-side motion rather than a wide zigzag.

That drive roll tension line deserves emphasis, because it is the classic cause of erratic feeding and bird-nesting, where the wire coils and kinks as it leaves the gun. Too much or too little tension on the drive rolls or the wire spool hub leads to poor wire feeding performance, so adjust them according to your machine’s owner’s manual rather than by feel. Loosen the drive roll if the wire is being flattened and stripping the liner. Tighten it if the wire is slipping and feeds slower than the dial says. If the feed is smooth but the arc stutters, suspect the spool hub instead.

Welding in All Four Positions Without Losing Your Mind

Four positions and a completely different feel to the puddle. Position does not change the physics of the arc, but it changes what gravity does to the molten metal. Settings that lay down a textbook flat bead can spit, droop, or burn clean through the moment you turn the work over. The good news is that position changes are small, predictable, and easy to learn once you know which two things to adjust.

Flat (1G): Your Baseline

Flat position is where you build your reference settings, so learn it first. A butt joint gets a 90 degree work angle, a T-joint about 45 degrees, and a lap joint roughly 60 to 70 degrees, with the angle increasing as the thicker member gets thicker. Your travel angle stays at 5 to 15 degrees of push for normal work in all positions. Make a consistent bead here and you have a baseline every other position is measured against.

Horizontal (2G): Beat the Rollover

In horizontal position the joint is vertical, so fillet metal tends to sag, roll over the bottom side of the joint, or slump into a lumpy teardrop. The fix is the work angle. Drop it 0 to 15 degrees from what you would use flat, which pushes arc force back into the joint and keeps the puddle where you want it. Travel angle stays the same as flat.

On settings, Miller Electric notes that voltage and amperage in horizontal position are usually the same as, or slightly less than, settings for the same weld in flat position. A small drop is plenty.

Vertical (3G): Down or Up

This is where most people quit, so let us make it less dramatic. The rule of thumb from Miller is simple: reduce voltage and amperage by 10 to 15 percent from the settings for the same weld in flat position. You are fighting gravity, and every extra unit of heat is heat the puddle does not need to hold its shape.

Vertical down

Start at the top of the joint and weld downward. Travel speed is faster than vertical up, so the arc penetrates less, which makes vertical down ideal for thin material and helps reduce melt-through. Keep the wire on the leading edge of the puddle, and on thin gauges direct the wire away from the puddle so the arc is not blasting straight down into molten metal.

Vertical up

Start at the bottom of the joint and weld upward. You get better penetration, which is why vertical up is standard for thicker material, typically 1/4 inch or more. Hold the travel angle 5 to 15 degrees down from perpendicular and use a slight weave to control puddle size and cooling. A big puddle hanging off the joint is how you get slag, undercut, and burn-through in one pass.

Aluminum in particular rewards this restraint. Because it conducts heat and melts so quickly, keeping the puddle small is the whole game, and our aluminum MIG welding guide covers the gas, gun, and cleaning side in more depth.

Overhead (4G): Gravity Wins the Race

Overhead is the position where everything wants to fall out of the joint. Miller’s guidance: use faster travel speeds so weld metal does not fall out, keep weave beads narrow, lower voltage and amperage to keep the puddle small, and consider a smaller diameter wire. Drag, push, or perpendicular all work overhead, so pick what feels controllable rather than what someone told you is correct.

The trick is to let the puddle barely exist. A small puddle that travels fast leaves a bead you can live with. A large puddle hanging under the torch drips onto your helmet and leaves a joint that needs grinding.

What Certification Means Here

Positions are not just a skill ladder, they are a paperwork ladder. Under AWS D1.1, a 3G vertical position plate test covers the flat (1G), horizontal (2G), and vertical (3G) positions, so a welder who passes vertical has already earned the flat and horizontal tickets. One catch: qualification is employer-bound. When you change jobs, your new employer has to accept your existing qualification or re-test you. Keep your test records and know who signed them.

Per-Position Settings Summary

Position	Work Angle	Travel Angle	Settings vs. Flat	Main Technique Focus
Flat (1G)	90° butt, 45° T-joint, 60-70° lap	5-15° push	Baseline	Reference recipe
Horizontal (2G)	Drop 0-15°	Same as flat	Same or slightly less	Stop rollover
Vertical down (3G)	Just over 90° on a T-joint	5-15° from perpendicular	10-15% lower	Thin material, less melt-through
Vertical up (3G)	Just over 90° on a T-joint	5-15° down from perpendicular	10-15% lower	Thicker material, more penetration
Overhead (4G)	Per joint type	Drag, push, or perpendicular	Lower volts and amps, faster travel	Keep the puddle small

The Most Common MIG Technique Mistakes (and How to Fix Each One)

MIG defects are honest: they show up in a predictable order, they point back to a short list of causes, and nearly all of them are fixable by the person holding the gun. Read the symptoms backwards and the fix usually collapses to one dial, a shorter stickout, or ten minutes with a wire brush.

The table below is the reference worth screenshotting. The symptom descriptions follow published troubleshooting guidance from Miller Electric on setting MIG parameters, with heat input and joint prep notes from The Fabricator and Hobart Brothers.

Symptom, Cause, First Fix

Symptom	Likely Cause	First Fix
Porosity and pinholes, face and interior	Shielding compromised: low flow, leak, wind, clogged nozzle, long stickout	Set flow to 20-25 CFH, soap-test hoses, clear nozzle, shorten stickout
Toes not wetting out (cold lap)	Heat input too low: long stickout, dirty metal, low voltage or feed speed	Clean to bare metal, stickout to about 3/8 inch, add volts or amps
Visible unfused edges, lack of fusion	Travel too fast, wrong work angle, not enough heat at the toe	Correct work angle, slow travel, add heat
Groove along the weld toe (undercut)	Arc too long, travel angle too steep, travel too fast to fill the toe	Shorten stickout, travel angle to 5-15 degrees, slow down
Melt hole straight through the sheet	Too much heat on thin material	Drop wire feed speed and voltage, shorten stickout, slow travel
Narrow convex bead, poor toe tie-in	Wire feed speed too low	Raise wire feed speed in small steps
Excessively wide bead, burn-through, heavy spatter	Wire feed speed too high	Lower the wire feed speed
Poor arc control, inconsistent penetration, turbulent pool	Voltage too high	Reduce voltage a little at a time
Poor arc starts and control, excessive spatter, convex bead, poor toe tie-in	Voltage too low	Increase voltage a little at a time
Narrow convex bead, inadequate toe tie-in, insufficient penetration, inconsistent bead	Travel speed too fast	Slow the travel speed
Excessively wide bead, poor penetration, burn-through on thin metal	Travel speed too slow	Speed up the travel
More spatter, less penetration, generally unstable arc	Travel angle pushed past 20-25 degrees	Return travel angle to 5-15 degrees
Cold lap and a cold-looking bead despite correct settings	Voltage drop over long power cables	Shorten or re-route work lead, tighten connections
Erratic feeding, bird-nesting, liner slip	Drive roller or spool hub tension off	Set roller tension to manual spec, adjust spool brake
Rust, paint, oil or moisture on the joint	Contaminated metal	Clean and degrease back to bare metal before striking an arc

Reading the Classic Defects Backwards

Cold lap is filler metal that did not fully fuse with the base metal at the weld toes, with three causes at once: heat input low, stickout long, metal dirty. Hobart Brothers links poor bead wetting from a long stickout to cold lap, and The Fabricator lists cold lap, ropey or humped welds and excessive spatter as signs of low voltage. Fix the sequence, not the symptom: clean the metal, pull the stickout back to roughly 3/8 inch, then add heat in small steps. On coated or oxidized stock, especially aluminum, cleaning aluminum properly before you weld removes a whole class of cold lap and porosity problems.

Porosity and pinholes almost always mean the shielding was compromised, as Miller Electric notes: low flow, a leak in the hose or regulator, wind across the work, a clogged or undersized nozzle, or a stickout so long the arc reaches past the gas coverage. A spatter-blocked nozzle is the sneaky one: it looks fine and quietly pulls air into the gas column.

Undercut is a fusion problem as much as a heat problem. It appears when the arc is too long, the travel angle too steep, the work angle is wrong for the joint, or the bead moves too fast to fill the toe. Too high a voltage flattens the bead into a concave profile that can cut a groove near the toe.

Burn-through is the opposite problem, pure excess heat on thin material: change wire feed speed and voltage first, shorten the stickout, and slow down. Lack of fusion is the mirror image, where the toes did not merge, so the remedy is more heat in the right place, not more speed.

When the Technique Is Right and the Equipment Is the Limit

Once your settings are repeatable, gear upgrades stop being optional. A gun with a better liner and a properly sized nozzle reduces spatter and gives steadier wire. For aluminum, a spool gun or rotary gun solves problems no setting change can; our complete aluminum MIG welding guide shows where that swap pays off and where it is overkill. An affordable fume extraction unit moves the visible plume out of your breathing zone, which helps on a workbench or in a garage, but is no substitute for proper local exhaust with coatings or stainless. Gear is not a substitute for technique: a new gun will not fix a cold lap caused by dirty metal, nor a travel angle held at 40 degrees.

Any gear link we add later is labeled clearly as an affiliate link at the point it appears, and we tell you when we earn a commission.

Safety and Fume Control: The Parts You Cannot Skip

Most welding safety advice arrives as folklore: wear a helmet, clear the area, keep an extinguisher nearby. The enforceable version lives in the Code of Federal Regulations, and for arc welding it is 29 CFR 1910.252. Learn the regulation and you can run a safer shop, train new welders faster and answer an inspector without guessing.

There is a link between technique and fume that most guides skip. Fume depends on heat input, arc time and what coats the metal. A lazy bead puts more vapor into the air than a tight, controlled one, and bare steel is nothing like painted, galvanized or stainless at the same settings. Safety is part of proper MIG welding technique, not an appendix.

Fire Prevention Is a Regulation, Not a Suggestion

Under 1910.252(a)(1), if the object to be welded cannot be moved, movable fire hazards must be removed to a safe place or confined with guards. If neither is possible, welding and cutting must not be performed. That is not a judgment call, it is a stop. Combustibles should be relocated at least 35 feet where practicable, or protected with flameproofed covers or metal guards (1910.252(a)(2)(vii)). NFPA Standard 51B, Standard for Fire Prevention in Use of Cutting and Welding Processes, is incorporated by reference into 1910.252(a)(1) and carries the detail on hot work permits.

A fire watch is required where appreciable combustible material sits within 35 feet of the point of operation, where wall or floor openings fall inside a 35-foot radius, or where combustibles exist on the far side of metal partitions. Extinguishing equipment must be at hand, and the watch maintained for at least half an hour after work finishes (1910.252(a)(2)(iii)).

Confined Spaces: Machines and Cylinders Stay Outside

Welding inside a tank, hopper or pipe changes the rule set. Under 1910.252(b)(4), welding machines and gas cylinders stay outside the space, an attendant with a preplanned rescue procedure observes the welder at all times, and lifelines are rigged for quick removal. Hot metal is marked so nobody walks into a tank still cooling inside.

Technique matters more here than anywhere else. A low-heat, controlled bead is a safety measure, not just a quality measure: fewer amps, shorter stickout, more stops and less spatter mean less fume in a space with poor airflow. The AllGas welding supply guide to gases, cylinder services and safety is a useful refresher on cylinder staging, regulators and hose inspection outside the space.

Ventilation, Hazard Communication and Chromium (VI)

Under 1910.252(c), local exhaust or general ventilation must keep toxic fumes, gases and dusts below the maximum allowable concentration in 29 CFR 1910.1000, and Hazard Communication (29 CFR 1910.1200) covers fluxes, coatings and filler metals. Control what the welder breathes, and tell them what is in the material.

Chromium (VI) needs specific attention, because stainless, chrome-plated parts and some coated steels generate it. OSHA’s permissible exposure limit is 5 micrograms per cubic meter as an 8-hour time-weighted average, with an action level of 2.5 micrograms per cubic meter (29 CFR 1910.1026). Engineering and work practice controls come first; respiratory protection is added only where those controls are insufficient or not feasible (1910.1026(f)).

OSHA’s fact sheet on controlling hazardous fume and gases during welding calls Cr(VI) fume highly toxic, damaging to eyes, skin, nose, throat and lungs, and a carcinogen. It advises cleaning coatings, solvent residue and paint off surfaces before welding, which is where stripping to bare metal pays off twice. OSHA’s April 10, 2017 letter of interpretation adds that air from carcinogen areas should be filtered and exhausted to safe, non-occupied locations so it cannot re-enter through fresh-air inlets, since recirculating filtered air can push exposures back up.

Personal Protective Equipment and Helmet Shades

1910.252(b)(2) covers body protection: flame-resistant clothing, leather gloves, a leather apron or sleeves, and safety footwear. OSHA’s filter-shade guide lists shade 12 for gas-shielded arc welding on ferrous material and shade 11 for nonferrous material (1910.252(b)(2)(ii)(H)). Shade 12 is the default for mild steel MIG; nonferrous runs one step lighter.

Situation	Requirement	Reference
Hazards cannot be moved or guarded	Welding and cutting must not be performed	1910.252(a)(1)
Combustibles nearby	Move 35 ft where practicable, or guard them	1910.252(a)(2)(vii)
Combustibles, openings or partitions within 35 ft	Fire watch with extinguisher, held 30 minutes after work stops	1910.252(a)(2)(iii)
Confined space	Machines and cylinders outside; attendant and lifelines rigged	1910.252(b)(4)
Fume and chromium (VI)	Ventilation below the 1910.1000 limit; Cr(VI) PEL 5 µg/m³, action level 2.5 µg/m³	1910.252(c), 1910.1026

Short Checklist Before the Arc Strikes

• Area: combustibles relocated or guarded, openings checked, fire watch assigned.
• Material: paint, rust, solvent residue and galvanizing stripped to bare metal.
• Machine: correct polarity for the wire, ground clamp on clean metal, cables intact.
• Gas: right cylinder and flow rate, regulator and hose leak-free, cylinder upright and secured.
• Ventilation: local exhaust running close to the arc, or general ventilation verified.
• PPE: shade 12 ferrous MIG, shade 11 nonferrous, flame-resistant clothing, leather gloves and apron, safety boots.
• Post-work: hot metal marked, fire watch held 30 minutes, area checked for smoldering material.

Put the checklist in the job folder next to your settings log. Safety habits fail the way technique habits fail: not because a welder does not know the rule, but because nobody made the rule part of the routine.

Building a Consistent Practice (and What to Take Away)

The welders who look consistently good are rarely guessing better than everyone else. They run a written recipe, and they can tell you afterward why the bead looked the way it did. That is the real payoff of proper MIG welding technique: turning what your hands know into something you can repeat.

Start With a Settings Log

Notebook, spreadsheet, or a photo of a tag taped to the work order. The format does not matter, as long as every job gets one. For each weld, record:

• Wire type and diameter
• Shielding gas and flow rate
• Wire feed speed and voltage
• Travel speed, and how you judged it
• Stickout
• Work angle and travel angle
• Position, joint design, and material thickness

Travel speed is the field most often left blank, and it explains most surprises later. Miller’s guidance is to match weld pool size to joint thickness, keeping the bead no larger than the thinnest section of metal being welded. If you never log it, you cannot tell later whether a bad bead came from heat or from motion.

Log the small things too. An irregular-sounding arc is very often a stickout that is too long, which Miller calls an extremely common error. And long power cables with loose connections are a leading cause of voltage drop and erratic feeding, according to The Fabricator and Miller, so record the setup you actually used, cable run included.

This habit mirrors how AWS D1.1 paperwork works. A WPS is required for every production process and joint. A PQR must be on file for every non-prequalified WPS. Welder qualification records must be current for each process used, and filler metal certificates must be traceable to the lot numbers consumed on the job. A personal log is not a PQR, but it is the same discipline that makes heavier documentation easy to keep.

When You Need Certainty, Cut a Coupon

Visual inspection tells you what the bead looks like, not whether it fused all the way through. Under AWS D1.1 test methods, destructive testing on a cut coupon is the definitive way to verify fusion and penetration, so for critical work, build a procedure you can afford to break.

A 30-Day Practice Plan

Week	Focus	What You Do
One	Flat butt and T-joint coupons	One material, one wire diameter, one gas. Change one variable at a time to learn what each one does.
Two	The same joint out of position	Start horizontal, then vertical up. Keep the same settings and record what changes.
Three	Thin gauge and a new gas	Try 75% argon / 25% CO2 for carbon steel with minimal spatter, then compare with 100% CO2, which gives deeper penetration and higher deposition on thicker plate but more spatter.
Four	A real project	Write the recipe down before you strike an arc, then inspect the finished work.

Four Questions After Every Weld

1. Did the arc sound right? Steady and busy, or irregular and popping?
2. Did the bead tie in at both toes, with no cold lap?
3. Was the bead profile right for the joint?
4. Did the heat stay under control, with no burn-through and no new distortion?

Ask these while the gun is still hot. Fifteen seconds, and the log fills itself.

What to Take Away

The framework is simple: parameters come from the machine, technique comes from the welder, heat input ties every defect together, and the symptom table is your diagnostic tool. Learn those four things and you are ahead of most people who have owned a MIG gun for a decade.

Then keep the supporting gear and knowledge current. Our airgas welding supplies buyer’s guide covers wire and shielding gas basics, the welding equipment guide covers choosing a gun and accessories worth buying, and the aluminium welding course breaks down what structured training looks like. For shielding gas handling and cylinder safety, see the gases and cylinders safety guide.

Frequently Asked Questions

What are the three main MIG welding settings I should start with?

Wire feed speed, voltage, and travel speed. Wire feed speed sets your amperage on a constant-voltage machine, voltage sets your arc length, and travel speed controls how much heat you put into each inch of joint. Start with the rule of roughly 1 amp per .001 inch of material thickness, then set voltage by listening to and watching the arc, and adjust travel speed to keep the bead sized to the joint.

How much wire stickout should I use for MIG welding?

About 3/8 inch for typical mild steel work, 1/4 to 3/8 inch in short-circuit transfer, and closer to 3/4 inch in spray transfer. Longer stickout adds electrical resistance, which drops current, reduces penetration, causes spatter, and can rob the joint of shielding gas coverage. A good trick is to listen for a steady sizzling-bacon sound – an irregular arc often means the stickout is too long.

Should I push or drag when MIG welding?

Push (forehand) gives a wider, flatter bead with slightly less penetration but a better view of the joint and better control of the wire into the joint. Drag (backhand) gives deeper penetration, a narrower bead and more buildup. Push suits thin material and aluminum; drag suits thicker steel and self-shielded flux-cored wire, where you want slag out of the weld.

What travel angle should I use for MIG welding?

Between 5 and 15 degrees for normal welding in all positions. Going beyond 20-25 degrees increases spatter, reduces penetration and destabilizes the arc. The work angle is a separate value that depends on the joint: 90 degrees for a flat butt joint, about 45 degrees for a T-joint, and 60 to 70 degrees for a lap joint.

How do I fix a MIG weld with undercut?

Undercut is a groove at the weld toe caused by the arc not filling the base metal, and it usually points to too much heat or a badly set angle. Shorten the stickout, drop voltage slightly, reduce travel speed so the puddle fills the toe, and check that your work angle suits the joint. On multipass welds, a slight pause at the toe as you weave helps tie-in.

What causes porosity and pinholes in a MIG weld?

Almost always compromised shielding. That can mean low or no shielding gas, a leak in the hose, wind or a draft, a clogged gas nozzle, or stickout so long the wire reaches outside the gas coverage. Check gas flow, hunt leaks with soapy water, clean the nozzle, and shorten the stickout. No shielding gas shows up as porosity and pinholes in both the face and the interior of the weld.

What voltage and wire feed speed do I need for 1/8-inch mild steel?

Roughly 125 amps, which is about 1 amp per .001 inch of material thickness. With .030-inch wire, the common starting point is 250 IPM of wire feed speed using the 2 inches-per-amp multiplier, with voltage set by ear and eye until the arc is stable and the bead has a slight crown that ties in at both toes.

Can I MIG weld outdoors in the wind?

Yes, but you need to shield the arc from the breeze, because wind breaks up the shielding gas and causes porosity. A simple windscreen around the work area, or working in a sheltered corner, solves most of it. Also check that the cylinder is upright and secured, and that wind is not carrying fumes toward you or anyone else.

Do I need a written WPS for a home project?

For a shop doing structural or pressure work, yes – AWS D1.1 requires a written Welding Procedure Specification, a supporting PQR for non-prequalified procedures, and current welder qualification records. For hobby work, the practical substitute is a personal settings log: write down wire, gas, wire feed speed, voltage, travel speed, stickout, angles and position for every joint you are happy with, so you can repeat it instead of re-learning it.

How do I get good penetration without burning through thin metal?

Shorten the stickout, drop the wire feed speed and voltage, and slow the travel speed so the pool has time to wet out at the toes. On thin material, drag or push angle, keep the bead narrow, and consider a smaller diameter wire such as .023-inch. On the other hand, if you are seeing cold lap instead of burn-through, you need more heat and a cleaner joint, not less.