Aluminum TIG Welding Polarity: AC, DCEN & DCEP Guide

A fabricator once spent an entire afternoon fighting a bubbly, black-dotted 6061 fillet on a trailer ramp, convinced his amperage and travel speed were wrong. The real problem was simpler: his machine was still set to DC electrode negative from the stainless job he ran the day before, so nothing was stripping the aluminum oxide off the joint. Polarity is the first decision in aluminum TIG welding, and getting it wrong makes every other setting irrelevant.

The Short Answer: What Polarity Should You Use for Aluminum TIG?

Use AC — alternating current. That is the correct polarity for the overwhelming majority of aluminum GTAW work, whether you are welding 0.040-inch sheet or a heavy 6061-T6 frame joint. Set the machine to AC before you strike the arc, and every other setting on the front panel — amperage, balance, frequency — has a chance of working. Get this first decision wrong and no amount of amperage tuning will save the weld.

What “polarity” actually means

Polarity is simply the direction in which welding current flows between the tungsten electrode and the workpiece. Think of it as traffic on a one-way street: current either leaves the electrode and goes into the plate, comes out of the plate and goes into the electrode, or flips back and forth many times per second. That gives you three choices:

• DCEN (DC electrode negative, sometimes called “straight polarity”) — current flows from the tungsten into the work.
• DCEP (DC electrode positive, or “reverse polarity”) — current flows from the work into the tungsten.
• AC (alternating current) — the direction reverses automatically, hundreds of times per second.

Polarity	Oxide cleaning	Penetration	Heat at the tungsten
DCEN	Essentially none	Deep and focused	~20%
DCEP	Strong etching action	Shallow	~80%
AC	Good, every positive half cycle	Medium	Split roughly evenly

Those heat numbers come from the European Aluminium Arc Welding manual, and they explain the whole trade-off in one line. On DCEN, roughly 80% of the arc’s heat goes into the base metal and only about 20% lands on the electrode — great for fusion, useless for breaking up oxide. Reverse the polarity and the split flips: about 20% of the heat reaches the work while roughly 80% cooks the tungsten. That is why DCEP cleans beautifully but melts electrodes if you hold it.

For most metals, GTAW runs on DC electrode negative — steel, stainless, copper, titanium, all of it. Aluminum is the famous exception. As TWI notes, AC must be employed for materials such as aluminium that carry a tenacious oxide film on the surface. The electrode-positive half of each AC cycle blasts that oxide off; the electrode-negative half drives heat into the plate and fuses it. You get both jobs done in a single arc instead of choosing one and sacrificing the other.

What the rest of this guide covers

Polarity is only the first dial. Once you are on AC, three more decisions shape the bead, and this article walks through all of them:

1. AC balance — the percentage split between cleaning and penetration, and how to read it off the puddle.
2. AC frequency — how tightly the arc is focused, and what 120 Hz does differently from 60 Hz.
3. Tungsten selection and prep — which electrode alloy and tip geometry suits each polarity.
4. Troubleshooting — decoding sugary beads, black peppering, and lack of fusion back to their polarity or settings cause.
5. Safety — UV, ozone, HF ignition risk, ventilation rules, and a practical settings workflow.

One warning before you go further: polarity interacts with cleanliness. Even a perfect AC setup will reject a joint that is oily, humid, or mechanically contaminated, so pre-cleaning belongs ahead of every setting change — our guide to aluminum cleaner for welding covers degreasing, etching, and drying in detail. And if you are weighing processes, our aluminum MIG welding guide explains where wire feed makes more sense than TIG. If you are here because a bead already went wrong, start with the troubleshooting table in section 9; if you are starting a fresh job, read on from the oxide problem that makes polarity matter in the first place.

Why Aluminum Is Different: The Oxide Problem Polarity Has to Solve

Steel rusts slowly and predictably. Aluminum oxidizes the instant oxygen touches it. Expose a freshly ground 6061 surface to air and a new oxide layer starts forming within milliseconds, growing to a protective film only atoms thick — and thickening further in hot or humid conditions. That skin is why polarity matters: every setting on the machine answers one question. How do we get through this oxide before the arc can weld the metal underneath?

The Refractory Skin Over Soft Metal

Here is the number that explains aluminum TIG in one line. Aluminum oxide melts at roughly 3,600°F (about 2,000°C), while the aluminum beneath it melts at around 1,200°F (about 660°C), as Miller Welds explains in its AC balance guidance. The oxide is three times more refractory than the metal it covers. As the puddle forms, the oxide does not soften and flow away with the base metal the way mill scale does on steel. It sits on top as a solid raft, physically holding the filler out of the joint and resisting the fluidity of the pool.

The symptoms are classic: the puddle refuses to wet out, the arc wanders instead of digging in, and the finished bead looks sugary or chunky. These are not amperage problems. They are oxide problems, and more heat will not cure them.

An Oxide Layer That Acts Like a Sponge

The oxide creates a second, sneakier problem. Its structure is porous at a microscopic level, and Miller warns that aluminum oxide effectively acts like a sponge, holding oil, moisture and shop grime in ways clean-looking metal does not. A plate that was wiped down and looks spotless can still carry hydrocarbons buried in the film. Under the arc those contaminants break down and dump hydrogen into the puddle, showing up later as porosity, black peppering and brittle beads.

This is why prep comes before polarity. The cleaning recipes in Aluminum Cleaner for Welding: How to Prep Aluminum Right are the foundation the whole weld sits on: wipe the joint with acetone or a dedicated aluminum degreaser, then brush with a stainless steel wire brush used only on aluminum and never on steel. That work removes the contaminated outer layer and cuts how much electrode-positive cleaning time you will need from the arc later. Clean metal means less aggressive balance settings, less tungsten heat and a calmer puddle.

What “Cleaning Action” Actually Is

Look at a well-run aluminum TIG bead and you will see a bright, frosty, etched stripe along both toes of the weld. On typical sheet it measures roughly 1/16 to 1/8 in. wide, as The Fabricator describes in its aluminum TIG tungsten guidance. That white band is not spatter or burn. It is oxide electrically stripped and vaporized during the electrode-positive portion of the AC cycle, leaving a surgically etched path. Welders call it cleaning action or the etched zone, and it is the visual proof that polarity is doing its job.

The width of that stripe is diagnostic. Too narrow or missing entirely, and the arc never removed the oxide, which means fusion is compromised. Too wide and fuzzy, and you are over-cleaning — burning tungsten and wasting penetration. You will learn to read it as a gauge in Aluminum MIG Welding: The Complete Practical Guide too, where surface prep drives spatter and porosity exactly as it drives TIG quality.

Alloys and Series: Why 5xxx and 6xxx Behave Differently

Alloy chemistry changes the difficulty. The common 5xxx series (aluminum-magnesium) work-hardens and carries more magnesium, which adds its own oxide and can make the puddle noticeably less fluid. 6xxx series (6061, 6063) are the general-fabrication workhorses and respond predictably to proper AC settings. On magnesium-bearing alloys the oxide issue is compounded, which is why Miller’s common TIG problems guide repeatedly traces bad beads back to contaminated or inadequately cleaned surfaces rather than machine settings. Specialty work such as Alloy Wheel Welding: A Comprehensive Guide to Safe and Effective Aluminum Wheel Repair compounds the challenge further, because cast aluminum wheels hide porosity and old repair material under the oxide.

In short: the oxide is the enemy, it is harder than the metal, it hides contaminants, and it has to go before fusion is possible. That single fact is what eliminates plain DC from everyday aluminum work and hands the job to alternating current — the subject of the next sections.

DC Electrode Negative (DCEN) on Aluminum: Deep Penetration, Zero Cleaning

Switch your TIG machine to DC electrode negative and run a bead on 6061, and you get something convincing for about two seconds: a narrow, deep, confident-looking line of fusion. Then the puddle starts acting strange. It refuses to wet out, it darkens, and the finished bead takes on the sugary, black-dotted texture everyone dreads. DCEN is not the wrong polarity because it is weak. It is the wrong polarity because it skips the half of the job aluminum demands most: destroying the oxide skin.

That said, DCEN deserves a proper explanation. Understanding exactly what it does and does not do is the fastest route to understanding why AC is mandatory for almost everything else.

The Physics: Electrons Traveling Into the Plate

Polarity simply describes the direction of current flow across the arc. With DC electrode negative, electrons stream from the tungsten, across the arc, into the workpiece. Because electrons carry most of the arc’s energy, that energy lands in the plate rather than in the electrode. The result is a concentrated, stable arc with deep, narrow penetration and a comparatively cool tungsten. This is the same reason DCEN is the default polarity for steel and stainless steel GTAW, where no oxide problem exists. The TWI overview of TIG equipment explains how arc characteristics shift with electrode polarity.

On aluminum the trade-off is absolute: DC-EN produces deep, narrow penetration with essentially no arc cleaning action, as European Aluminium notes in its welding guidance. Nothing in the electrode-negative half of the cycle blasts the oxide film apart. The arc simply dumps heat on top of a contaminant layer and expects fusion anyway.

Where DCEN on Aluminum Actually Gets Used

You will not find DCEN on an aluminum job by accident in most shops. It shows up in a few deliberate situations:

• Thick sections with helium shielding. Helium’s higher arc voltage pushes far more energy into the joint. European Aluminium reports that DC-EN with helium can be suitable for thicknesses up to about 25 mm, compared with roughly 6–10 mm for conventional AC TIG. That is a dramatic capability jump.
• Mechanized and automated work. Because the technique demands discipline that is hard to hold by hand, DC-EN aluminum welding is generally recommended as a mechanized process, with torch angle, arc length and travel speed held constant by a carriage or robot.
• Very thick, heavily pre-cleaned plate where a fabricator has mechanically removed every trace of oxide in the joint zone and wants maximum penetration without the heat spread of AC.
• Back-chaining or closing runs on thick material after the root has already been cleaned and established by another method.

Note what all of these have in common: none of them is a hand-held trailer ramp on a Tuesday afternoon.

The Pre-Cleaning Discipline You Cannot Skip

On AC, the electrode-positive half of the cycle scrubs the oxide for you, so a slightly lazy prep job often still welds. On DCEN there is no safety net. Every surface the puddle will touch has to be mechanically or chemically cleaned first: fresh stainless wire brushing with a dedicated aluminum brush, solvent degreasing, and removal of the oxide layer immediately before welding. The Aluminum Cleaner for Welding: How to Prep Aluminum Right guide walks through the full prep sequence, and it matters more on DCEN than on any other polarity.

Timing matters too. Aluminum rebuilds its oxide film almost instantly, so a joint brushed an hour earlier in a humid shop is already carrying a fresh skin.

The Failure Mode: Oxide Inclusions and Sugary Beads

Without cleaning action, fusion depends entirely on breaking up the oxide mechanically by arc force and heat. When that does not happen, the oxide stays put, and the AWS aluminum welding reference documents the predictable outcome: oxide inclusions and fusion defects trapped along the weld line.

Typical signatures of a DCEN mistake on aluminum include:

• A dull, sugary or crumbly bead surface with black dots and soot rather than a shiny, fluid puddle.
• A complete absence of the frosty etched band that should flank a good AC bead.
• Porosity and lack of fusion that x-rays or a bend test will find even when the surface looks acceptable.
• Filler that balls up and refuses to tie into the joint sidewalls.

This is why most welders who wander onto DCEN from a stainless job diagnose their problem as “amps too low” or “travel speed too fast.” The machine settings are rarely the issue. Polarity is.

The Sub-Millimeter Arc Length Rule

If you do run DCEN on aluminum, arc length becomes critical: the arc generally needs to be held under 1 mm. Any longer and the shielding envelope breaks down, the arc wanders, and you lose both penetration and gas coverage. That tight standoff is another reason DCEN favors mechanized setups, where arc length is fixed by a slide rather than an operator’s wrist. For comparison, our Aluminum MIG Welding: The Complete Practical Guide covers the transfer modes and gun handling that make MIG more forgiving of operator variation on aluminum.

Bottom line: DCEN is a powerful, specialized tool for thick, meticulously pre-cleaned, mechanized aluminum work, sometimes with helium pushing it to 25 mm. For everything else, it explains exactly why AC exists.

DC Electrode Positive (DCEP): The Cleaning Polarity You Rarely Use

Flip your polarity switch to DC electrode positive and you essentially reverse the whole engine of the arc. Electrons now stream from the workpiece up into the tungsten instead of shooting out of it, and conventional current flows the opposite way. The result is an arc that cleans beautifully and penetrates barely at all. It is the mirror image of DCEN, and for most welders it is the polarity they meet once, try for a afternoon, and never select again.

That reputation is earned, not accidental. According to the arc physics summarized by TWI’s TIG equipment knowledge base, reversing electrode polarity shifts the balance of arc heat toward the electrode rather than the plate. The European Aluminium arc welding reference puts a number on it: roughly 80% of the heat generated in DCEP is generated at the electrode, not in the workpiece. Think about that for a second. Four fifths of your welding heat is going into a rod of tungsten that is already one of the highest-melting-point materials on earth. The plate gets the remaining fifth, which is why the weld is shallow, wide, and slow to tie in.

Why the Cleaning Is So Good

The flip side of all that electrode heat is genuinely excellent oxide removal. In DCEP, positive ions in the arc column slam into the workpiece surface, physically blasting apart the stubborn aluminum oxide skin that causes so much trouble. On a properly set DCEP bead you will see a wide, frosty, etched zone flanking the puddle that looks almost over-prepped. For a demonstration of what cleaning action actually looks like on aluminum, DCEP is unmatched, which is exactly why many instructors use it to teach the concept before moving students onto AC.

The catch is that etching and fusion are not the same thing. A beautifully frosted DCEP bead can still be sitting on top of the plate with barely any root penetration, because the base metal simply never got enough heat to melt deeply. On anything structural, that is a reject waiting for a cut-out.

The Tungsten Problem

All that heat landing on the electrode is what really limits DCEP. The tungsten balls up, the tip geometry becomes unstable, and the arc starts to wander and sputter. To run DCEP at all you need to size up: a larger-diameter electrode than you would use on AC or DCEN, ground to a wide included angle rather than the usual sharp taper, so the tip has enough mass to soak up the punishment. Even with that discipline, most operators still see excessive electrode heating and shortened electrode life, which is why DCEP is widely treated as the least-used GTAW polarity in production work. Tungsten selection guides for aluminum are explicit about this: tip geometry and diameter follow polarity, not personal preference.

Where DCEP Actually Gets Used

Despite everything, DCEP is not useless. The realistic applications are narrow:

• Tack welding thin sheet, where a quick clean, shallow tack is all you need and full penetration is irrelevant.
• Very light gauge work where even AC would burn through, though this is rare and requires real control.
• Teaching and demonstration, to show students what oxide cleaning looks like before introducing AC balance.
• Brief pre-pass cleaning on heavily oxidized material, switching to AC or DCEN immediately afterward for the real weld.

Even in these cases, a short DCEP pre-pass does not replace proper preparation. Mechanically and chemically cleaning the joint first, as covered in Aluminum Cleaner for Welding: How to Prep Aluminum Right, will always beat trying to burn contamination off with polarity alone.

DCEP vs DCEN at a Glance

Characteristic	DCEP	DCEN
Electron flow	Workpiece to tungsten	Tungsten to workpiece
Oxide cleaning	Excellent	None
Penetration	Very shallow	Deep, narrow
Heat location	~80% at the electrode	Concentrated in the plate
Tungsten prep	Large diameter, wide angle, balled tip	Smaller diameter, fine pointed taper
Typical use	Tacks, pre-passes, demos	Thick sections, mechanized work

One last warning. Polarity mistakes cut both ways, and Miller lists incorrect polarity as a top-ten TIG problem: running DCEN on aluminum lets the filler metal mix with partially melted, still-solid oxide, producing a contaminated, sugary bead that looks terrible and tests worse. So before you strike an arc on aluminum, glance at the front panel and confirm what you are actually running. DCEP is a specialized tool, not a default, but knowing exactly what it does and does not give you makes the choice obvious: for nearly every aluminum GTAW joint, AC wins.

Why AC Is the Standard Polarity for Aluminum TIG Welding

If DCEN and DCEP each do half the job and ruin the other half, the engineering answer is to alternate between them. That is exactly what alternating current does. In one AC TIG arc, the current swings between electrode positive and electrode negative many times per second, so the joint gets a cleaning pulse and a fusion pulse in quick succession rather than a choice.

Two Jobs, One Arc

During the electrode-positive (EP) half-cycle, electrons reverse direction: they travel from the workpiece into the tungsten, and heavy argon ions smash into the plate surface. That bombardment lifts aluminum oxide and the contaminants bonded to it out of the joint. It is the same etching action DCEP produces, but it lasts only half of each cycle, so the tungsten survives it.

During the electrode-negative (EN) half-cycle, electrons leave the tungsten and drive into the plate, concentrating heat in the base metal. That is the penetrating, fusion-producing half. Miller summarizes the split: the EP half-cycle removes surface oxides while the EN half-cycle melts and fuses the base metal. Switch fast enough and the two effects blend into one continuous, fluid puddle.

On a standard 60 Hz machine the polarity reverses 120 times a second, giving 60 cleaning half-cycles every second. To the eye it looks like one steady arc; chemically it is a rapid scrub-and-melt sequence running along the joint.

The Zero-Crossing Problem

The trick is what happens at the moment of reversal. Alternating current must pass through zero to change direction, and as current falls toward zero the arc plasma cools and starts to de-ionize. On an old sine-wave transformer machine the arc would extinguish at every crossing with nothing to re-establish it, because welding voltage cannot jump a de-ionized gap.

The fix on legacy equipment was a continuous high-frequency spark that re-ionized the gap at every reversal. That is why classic AC buzz boxes hummed loudly, and why welders linked AC with radio interference, stray HF sparks and poor stability at low amperages. As TWI’s equipment overview explains, sine-wave AC needed HF assistance to keep reigniting.

How Square Wave Solved It

Modern inverter power sources do not follow a smooth sine curve. They switch current between polarities almost instantly and hold it at a steady plateau, so the arc never lingers near zero. The plasma column stays ionized through the transition and reignites itself, so high-frequency start can be reduced to a single ignition pulse at strike-up or switched off entirely while welding.

TWI notes that square-wave and switched-DC sources made arc reignition so reliable that HF could be reduced or eliminated, and the same switching architecture added two controls sine-wave machines never had: EN/EP balance and reversal frequency.

Characteristic	Sine-wave AC (transformer)	Square-wave AC (inverter)
Current at reversal	Gradual, passes slowly through zero	Near-instantaneous switch
HF requirement	Continuous during welding	Start-only or none
Arc stability at low amps	Poor, tends to stutter	Good, supports thin sheet
Balance control	Limited or absent	Fine adjustment, typically 60–90% EN
Interference and noise	Noticeable HF buzz	Quiet and low-interference

What a Correct AC Bead Looks Like

Run AC properly and the result is unmistakable: a shiny, wet-looking puddle that wets into the edges, flanked by narrow, frosty-white etched bands where the oxide has been lifted away. Those pale stripes are not contamination; they prove the cleaning half-cycle worked. A well-set weld holds them to a consistent width, about one to two millimetres either side of the bead. Black peppering inside the ripples means oxide is trapped rather than removed, and the balance needs attention.

Why AC Is the Default

Almost every manual aluminum job lands on AC for the same reason: the operator must clean and fuse at the same time, with no chance to stop and scrub between passes. The 5xxx (aluminum-magnesium) and 6xxx (aluminum-silicon) alloys that dominate shop and marine work both grow a stubborn oxide film, and AC handles them without special procedure changes. It works on thin sheet, where a shallow EN-heavy puddle gives control, and on heavier stock, where more EP can be dialed in for cleaning.

That is why AC is the assumed setting for everything from hand-fabricated aluminum boat hulls to alloy wheel repair, where sections are thin, the alloy is heat-sensitive and the joint geometry changes every few inches. ESAB recommends pure, zirconiated or ceriated tungsten for AC aluminum work because those electrodes hold a stable arc across both half-cycles.

AC Balance Control: Tuning Cleaning Against Penetration

AC balance is the single most useful dial on an aluminum TIG machine, and it is also the one most often misunderstood. In plain terms, balance is the percentage split between the electrode-negative (EN) half of the cycle and the electrode-positive (EP) half. Because the arc alternates back and forth many times per second, those two portions always add up to 100%. If one side goes up, the other comes down by exactly the same amount. Think of it as a seesaw: more EN means more penetration and less cleaning, more EP means more oxide etching and a hotter tungsten.

Get that split right and the puddle behaves. Get it wrong and you are back to fighting black dots, sugary beads, or a tungsten that keeps balling up no matter how careful you are with your arc length.

Read Your Machine’s Display Before You Touch the Dial

Here is where plenty of welders sabotage themselves without realizing it. Manufacturers do not agree on which number to show you. Miller, Lincoln and HTP display the electrode-negative percentage, so a reading of 75 means three-quarters of the cycle is spent penetrating. Everlast and Thermal Arc machines display the opposite, the electrode-positive percentage, so the same physical setting on those brands reads 25. The arc is identical; only the label is flipped.

Where to Start: Typical Balance Numbers

For most shop work on clean aluminum, a starting point of 75% EN is a reliable place to strike an arc. From there, the practical adjustments look like this:

• Clean, freshly deoxidized material: 68–75% EN is the preferred working band for clean workpieces, and very clean parts can go close to 80% EN for maximum penetration (Miller).
• Older, dirtier or mill-finish aluminum: drop toward 65% EN so the EP half can burn off more contamination before it reaches the puddle.
• Adjustable range: many machines allow roughly 50–75% adjustment, though inverter units often give you a wider span than transformer rigs (ESAB University).

One number matters more than any other: 60% EN is the floor. If your weld will only come out clean after you dial below 60% EN, the part is telling you it needs mechanical cleaning rather than more arc cleaning. Chasing contamination with EP is a losing game — you burn tungsten, widen the heat-affected zone, and still weld over oxide that was never removed in the first place. A stiff stainless brush, a proper aluminum cleaner for welding, and a wiped joint will almost always let you move back up into the 70% range.

Adjusting by What You Actually See

Balance is not set by the numbers alone; it is confirmed in the puddle. Use the weld itself as your gauge:

What you observe	What it means	Adjustment
Black peppering or dark dots in and around the bead	Not enough cleaning (too much EN)	Increase EP, lower the EN percentage
Wide frosty etched zone, dull gray wash	Too much cleaning (too much EP)	Increase EN, narrow the etch
Tungsten balling, arc wander, glowing tip	Electrode overheating from excess EP	Add EN, or step up tungsten diameter
Narrow bead with good tie-in at the toes	Healthy balance for the material	Hold the setting
Shiny molten dot appears but sits on top of the plate	Oxide film still blocking fusion	Add cleaning (more EP) or stop and degrease

Increasing the EN portion does four things at once: it narrows the bead, deepens penetration, reduces tungsten balling, and shrinks the etched zone. That is why clean material tolerates high EN so well, while dirty material fights it.

Material condition	Starting balance (EN)	Expected result
New, clean 6061 or 5052 sheet	75–80%	Deep fusion, tight etched band
General fabrication, clean plate	70–75%	Good cleaning with solid penetration
Older, oxidized or mill-finish stock	65–70%	Wider etch, more forgiving on surface film
Greasy, heavily oxidized, or unknown alloy	Below 60% = stop	Clean the part mechanically first

The practical takeaway is simple: start at 75% EN, read the puddle, and move in 5% steps. Peppering means you need more EP. Excessive etching and a molten tungsten mean you have too much. And if you find yourself parked below 60% EN, put the torch down and clean the metal instead — that is almost always the real fix, and it is far cheaper than a rebuilt joint.

AC Frequency and Waveform: The Other Polarity Dial

Once you have set AC balance, there is a second dial that changes how the arc behaves without changing the EN/EP split at all: AC frequency. Frequency is simply the number of times per second the current reverses polarity. Balance decides how long the arc spends cleaning versus penetrating; frequency decides how often that handoff happens. On a traditional transformer machine you do not get a choice. As TWI’s equipment overview of the TIG process explains, line-frequency AC output is fixed at the supply frequency, which is 60 Hz in North America and 50 Hz in Europe. The arc simply extinguishes and re-strikes 120 times a second, and you live with whatever arc cone that produces.

Modern inverter power sources decouple the output from the wall supply, so the frequency becomes adjustable. Common inverter ranges run from around 20–40 Hz at the low end up to 120–150 Hz, and some machines go to 200–250 Hz. That range gives you a genuinely useful shaping tool.

What Higher Frequency Actually Does

Raising the frequency constricts the arc column. Because the current reverses before the ionized path has time to spread, the arc cone stays narrow and stiff, and the arc root becomes noticeably more resistant to wandering. In practice this means:

• Tighter heat zones. A focused arc puts heat in a smaller spot, which is exactly what you want on a narrow butt joint, a corner fillet on thin sheet, or a repair right next to a machined edge.
• Better directional control. A stiff arc is easier to aim along a joint, especially out of position or when the gap is inconsistent.
• Slightly higher effective heat input. This is the trade-off people forget. A constricted arc delivers its energy more efficiently into a smaller area, so the same amperage can feel hotter and dig deeper. Move from 60 Hz to 120 Hz without touching the amperage and expect a deeper, narrower bead, sometimes burn-through on thin material.

This is why fabricators welding thin aluminum, or repairing something like an alloy wheel where the parent metal is thin and heat-sensitive, often run 100–150 Hz and then dial the amperage down a few amps to compensate. The same arc-focusing logic applies if you ever step across to related aluminum processes, and our alloy wheel welding guide covers the heat-control side of that repair work in more detail.

What Lower Frequency Does

Dropping below 60 Hz has the opposite effect. The arc cone widens and softens, the etched cleaning zone broadens, and the arc becomes more forgiving on fit-up gaps, heavier sections and dirtier material. A wide, soft arc spreads heat over more surface area, which helps on thick plate and on castings where you want gentle preheat across a broad zone rather than a hot pinpoint. The cost is reduced control on thin work and a wider frost band, which on visible welds can look untidy.

Waveform Options on Modern Inverters

Adjustable frequency is only part of what inverter machines now offer. Many let you change the shape of the AC output, not just its speed:

• Advanced (or “balanced”) square wave gives a fast, crisp polarity transition and a very stable arc at ordinary welding amperage. It is the general-purpose choice for most aluminum fabrication.
• Triangle wave ramps the current up and down smoothly instead of switching abruptly. The result is a softer arc with less flicker and less spatter, which is ideal for low-amperage work on thin sheet where a square wave would punch through.
• Unbalanced or modified sine wave options shift the effective cleaning/penetration relationship at very low amps, letting you keep cleaning action on material that would otherwise be destroyed by a full-power AC pulse.
• Pulse modes superimpose a pulsing peak current on the AC output, cutting average heat input substantially while retaining cleaning on every pulse. This is how welders run root passes on thin 5052 and 6061 sheet without a lot of distortion.

Miller’s guidance on AC balance control pairs frequency and balance as complementary adjustments: balance sets the cleaning-to-penetration ratio, frequency sets the arc shape. The two should be tuned together, and both are secondary to cleanliness. As ESAB University notes on GTAW of aluminum with alternating current, and as European Aluminium points out, running around 90% electrode-negative balance is aggressive enough that a properly prepared tungsten can stay pointed through filler runs rather than balling, which itself improves arc aim on narrow joints.

Note: Frequency and waveform cannot rescue a badly prepared joint. They change arc geometry, not metallurgy. A greasy or unetched surface will still reject the puddle at 150 Hz exactly as it does at 60 Hz, so chemical pre-cleaning comes first. See our aluminum cleaner for welding guide for the prep sequence that makes these settings worth tuning.

The practical workflow is simple: start at your machine’s default, usually 60–80 Hz with mid-range balance, run a test bead on scrap of the same alloy and thickness, then move frequency up if the arc wanders or the bead is too wide, and down if the arc feels harsh or the etched zone looks excessive. Adjust one dial at a time, and keep the amperage fixed until the arc behavior is right.

Tungsten Selection and Prep for Each Polarity

No polarity setting saves you from the wrong tungsten in the collet. The electrode lives directly inside the arc, and its alloy, diameter and tip geometry are dictated by the polarity you run and the machine driving it. A transformer sine-wave machine and a modern inverter do not even want the same tip shape on AC.

Why Tungsten Is the Only Electrode Material That Works

Tungsten has the highest melting point of any metal at roughly 3,410°C, which is why it can sit in an arc stream for hours without simply vaporizing (The Fabricator). Pure tungsten alone has limits, so manufacturers alloy it with small percentages of oxides that improve current carrying, arc starting and contamination resistance. Those additions are what the color stripes on the electrode body report, and they map almost perfectly to polarity duty.

Color Codes Matched to Polarity

Color / AWS	Alloy	Best polarity	Practical character
Green (EWP)	Pure tungsten, ~99.5%	AC	Soft, stable arc on transformers; balls easily; weak for DC penetration
Brown (EWZr-1)	≥99.10% W with 0.15–0.40% zirconium	AC only	Best pure-type for AC; tolerates contamination better than green; never for DC
Red (EWTh-2)	2% thoriated	DCEN mainly	Excellent arc starting and deep, focused penetration on steel and aluminum DC work
Grey (EWG/Ce-2)	2% ceriated	AC or DC	Easy low-amperage starts; a favorite on inverter machines for thin material
Blue (EWLa-1 / EWLa-2)	1–2% lanthanated	AC or DC	Most versatile all-rounder; runs on both transformer and inverter machines without complaint

The zirconiated brown electrode deserves special attention here: at least 99.10% tungsten with 0.15–0.40% zirconium, engineered specifically for alternating current on aluminum and magnesium. It tolerates the repeated polarity reversals of AC far better than thoriated grades and is never recommended for DC circuits (The Fabricator). Conversely, thoriated red is the classic DCEN electrode and a poor choice for sustained AC work on older machines.

Balled, Pointed or Truncated: Tip Geometry by Polarity

DCEN wants a ground, pointed tip; the concentrated point keeps the arc focused and drives heat into the plate. AC on transformer machines historically wants a balled end, formed by momentarily switching to DCEP and letting the tip melt into a hemisphere. There is a hard limit: the ball should never exceed 1.5 times the electrode diameter, or the arc becomes unstable and the molten ball can drop into the puddle (Miller Welds). A good ball is a slightly flattened dome, not a cherry.

Inverter machines change the rules. Because the square wave is electronically controlled, many inverters allow a pointed or lightly truncated tip on AC, skipping the balling step. That is one reason lanthanated electrodes took over in modern shops: they hold a point on AC without degrading the way thoriated tungsten does. See TWI’s TIG equipment overview for more on how power source type shapes electrode behavior.

Grinding and Prep Discipline

How you grind matters as much as what you grind:

• Grind lengthwise, with the grind marks running parallel to the electrode axis. Circumferential scratches spin the arc sideways and cause wander.
• Keep the taper short. A land or taper of no more than 2.5 times the electrode diameter keeps the tip cool and stable; long needles overheat and spit.
• Use a dedicated grinder, ideally with a diamond wheel, used only for tungsten. Steel grinding wheels embed iron in the electrode and poison the arc.

The same discipline applies to the joint. No tungsten choice compensates for an oxide-covered, oil-soaked surface, so mechanical cleaning and a proper aluminum cleaner still come before the arc strikes.

Sizing the Electrode for AC

AC puts more heat into the tungsten than DCEN, because the electrode-positive half of every cycle drives heat back into the electrode. A tungsten that carries 120 A on DCEN may run borderline on AC at the same setting, so start one size up when you switch. A 3/32 in (2.4 mm) tungsten handles most general aluminum AC work; step to 1/8 in (3.2 mm) for sustained high-amperage runs on thicker material. Keep stick-out short, since overhanging the tungsten beyond the gas cup raises its operating temperature. Remember the filler rod you choose still needs to match the alloy, not the tungsten.

Bottom line: match the electrode to the polarity you actually run most. Zirconiated brown or pure green for sine-wave AC, lanthanated or ceriated for inverter AC and DC versatility, and thoriated red for dedicated DCEN work. With the right electrode properly ground, AC balance and frequency become meaningful adjustments instead of guesses.

Common Aluminum TIG Polarity Mistakes and How to Fix Them

Aluminum TIG failures rarely announce themselves as polarity problems. They show up as sugary gray beads, black specks skating across the puddle, tungsten that keeps collapsing, or a fillet that looks welded until it pops under load. Amperage and travel speed usually get the blame, but the machine setting is more often the real fault. Below are the realistic failure signatures, what actually causes them, and the fix for each.

Polarity Failure Cheat Sheet

What you see	Likely cause	Fix
Sugary, gray, oxide-laced bead with no etched zone beside it	DC or DCEN selected instead of AC	Switch to AC and re-prep the joint
Black pepper flakes and brown haze swirling in the puddle	Too much electrode-negative time	Add cleaning (EP) time in small steps
Wide frosted band, balled tungsten, soft wandering arc	Too much electrode-positive time	Raise EN just to the point before peppering
Cleaning looks wrong the second you strike the arc	Balance display convention misread	Confirm whether the readout shows EN or EP
Porosity and black dots on an otherwise clean setup	Skipped degreasing, wet filler, dirty cup	Degrease, dedicated brush, dry filler
Erratic arc, tungsten spitting, gray inclusions	Wrong electrode or badly ground tip	2% lanthanated or ceriated, lengthwise grind
Cloudy bead, porosity at stops and starts	Gas coverage too short or a draft	Lengthen post-flow, block the airflow

Mistake 1: DC Left on From the Previous Job

This is the trailer-ramp story. When the machine is still on electrode negative, electrons flow straight from the tungsten into the plate, the arc stays narrow and hot, and nothing strips the oxide skin. The bead comes out gray and contaminated rather than shiny, with no frosty etched zone on either side of the weld. Miller’s TIG troubleshooting guide shows straight DC on aluminum with argon producing exactly that contaminated, oxide-laced result. The fix takes two seconds: confirm the display reads AC before you strike an arc. If the weld is already bad, cut it out, clean back to bright metal, and start over on AC. There is no setting that rescues a DC bead on aluminum.

Mistakes 2 and 3: Balance Points in the Wrong Direction

Two opposite symptoms come from two opposite balance errors, and both are visible the moment the puddle forms.

• Too much EN (not enough cleaning): brownish oxidation and black-pepper flakes appearing in the puddle mean the balance needs more electrode-positive time. Miller recommends adding cleaning in small increments until the pepper disappears.
• Too much EP (too much cleaning): an over-wide etched band, excessive puddle etching, and a tungsten that keeps balling or deforming mean the reverse. Raise EN to just before peppering appears, which gives you penetration without inviting contamination.

Machine makers do not agree on how to display this. Miller, Lincoln and HTP show the electrode-negative percentage; Everlast and Thermal Arc typically show electrode-positive. A welder who reads “65%” as cleaning on a machine that actually means 65% penetration has the balance almost backwards. Check the manual, then sanity-check the display against the weld: more frost than penetration means you are lower in EN than you think.

Mistakes 4 and 5: Skipping Prep and Contaminated Filler

Polarity cannot fix a dirty joint. Oil, mill scale and shop grime soak into the oxide layer and boil out as porosity later. Mechanical cleaning plus a proper degrease comes first, every time; our aluminum cleaner and prep guide covers the sequence in detail. Keep a dedicated stainless brush that never touches steel, wipe with acetone, and store filler rod clean. If TIG keeps fighting you on a grimy field repair, it may be worth stepping down to a flux-cored aluminum process that tolerates imperfect surfaces, rather than chasing the problem through balance settings.

Mistakes 6 and 7: Wrong Tungsten and Forgotten Gas

An electrode chosen for steel will misbehave on AC: pure or thoriated tungsten that erodes, spits, or wanders pulls the arc off the joint and drops tungsten particles into the bead. Switch to 2% lanthanated or ceriated, and grind it lengthwise with a taper no longer than 2.5 times the diameter. Gas is the last silent failure: too little flow, a long arc, or a draft gives a cloudy, porous weld that looks fine until it is x-rayed. Extend post-flow until the puddle dulls, and shield the area.

Finally, if the setup is right but the technique keeps slipping, it is usually a practice issue rather than a polarity issue. Run a test coupon, watch the etched zone appear, and adjust one variable at a time. Troubleshooting works best when only one thing changes per pass, which is also the fastest way to learn how an aluminum MIG setup or any other process behaves differently from AC TIG.

Safety, Fumes and Ventilation When Running AC on Aluminum

Polarity changes what the arc does to the metal, and it quietly changes what the arc does to you as well. The same electrode-positive half-cycle that etches oxide off a 6061 joint is also driving oxygen into the gas envelope and producing ozone, while the whole arc is throwing out ultraviolet radiation intense enough to burn skin and damage eyes in seconds. Aluminum GTAW is often described as a low-fume process, and compared with flux-cored or stick welding that is true, but low is not zero. Everything you have tuned so far, balance, frequency, tungsten, deserves a matching ventilation plan.

What the Arc Actually Puts in the Air

The fume from aluminum TIG is mostly fine particulate from the electrode, the plate surface and the filler, along with gases generated by arc chemistry. Ozone and nitrogen oxides come from the arc reacting with the air around it, and both are irritants that compound with exposure over a shift. Where AC is run with thoriated tungsten, or where an HF start is used, you also have to account for thorium dust during grinding and for small quantities of hydrogen fluoride. The relevant exposure limit for aluminum fume is an 8-hour time-weighted average PEL of 5 mg/m³ measured as aluminum.

Ventilation is not optional guidance here. OSHA 1926.353 requires local exhaust hoods to be placed as close as practicable to the work, with general mechanical ventilation used to keep fume concentrations within safe limits. AWS ANSI Z49.1, Safety in Welding, Cutting and Allied Processes, applies the same logic to every welding operation: provide adequate ventilation, and where that is impractical, provide respiratory protection instead.

Hazards Tied Directly to Polarity Choice

Three polarity-specific risks are easy to overlook:

• HF ignition risk. High-frequency start and AC maintenance circuits can ignite flammable vapors at a distance from the arc and can upset nearby electronics, computers, drives and fuel systems. Clear the area, ventilate before striking an arc, and keep ignition sources away from the tank bay.
• Oxygen displacement. Argon is heavier than air and pools in pits, hulls and tanks. Welding inside an aluminum boat or any low-lying enclosure demands a confined-space procedure, continuous gas monitoring and forced-air supply, not just a fan.
• UV and infrared exposure. AC and DCEP phases produce a broad, bright arc that will sunburn exposed skin behind your neck and inside your sleeves. Auto-darkening lenses do not replace covering up.

PPE Minimum for AC Aluminum GTAW

• A welding helmet with a shade appropriate for AC, typically shade 10 to 13, plus a gold or platinum filter if you are running high frequency all day.
• Leather jacket, gauntlets and closed cuffs, since arc burn from UV is a real shift-long exposure.
• Safety glasses under the helmet for grinding, mandatory when prepping tungsten or cleaning a joint.
• A respirator where local exhaust cannot reach the arc, as required under AWS Z49.1 when ventilation is impractical.

The Five-Step Polarity Setup Workflow

1. Clean first. Degrease, wire brush with a dedicated stainless brush, and remove the oxide. A aluminum cleaner for welding removes what the arc cannot.
2. Choose the tungsten for the polarity you intend to run, and grind it lengthwise with a taper no longer than about 2.5 times diameter.
3. Set AC balance near 70% EN for clean material, dropping toward 55-60% EN if the plate is dirty or has been sitting.
4. Run a test bead on scrap of the same thickness and alloy, and read it: peppering means more electrode-positive cleaning, a frosty wash-out or balled tip means back off toward EN.
5. Adjust one variable at a time and retest before touching the joint.

The Polarity Decision Tree

Everything in this guide reduces to a short chain of decisions. Is the part aluminum? Then the oxide layer decides the process. Do you need cleaning and penetration in one arc? Run AC. Is the material exceptionally thick, mechanized and perfectly pre-cleaned, with helium available? DCEN becomes worth considering. Is the work very thin tacks or a demonstration of cleaning action? DCEP has a narrow role. On everything else, the correct answer is alternating current with a sensible balance, a focused frequency, the right tungsten and clean metal underneath it.

Set polarity first, then amperage, then balance and frequency. Get that order right and the puddle tells you the rest, which is exactly where the finishing touches on any aluminum weld come from, whether you are on trailer ramps or on alloy wheel repair. Polarity is one setting; the habit of checking it before every job is what keeps the afternoon from disappearing.

Frequently Asked Questions

What polarity is used for TIG welding aluminum?

Alternating current (AC) is the standard polarity for TIG welding aluminum. The electrode-positive half of the AC cycle cleans away the stubborn aluminum oxide, while the electrode-negative half penetrates and fuses the base metal. DC is used only in specialized cases.

Can you TIG weld aluminum on DC electrode negative?

Yes, but only under controlled conditions. DCEN gives deep penetration and about 80% of the heat in the plate, but provides no oxide cleaning, so it demands immaculate pre-cleaning, a very short arc and usually helium shielding. It is mostly a mechanized technique.

What is a good starting AC balance setting for aluminum?

Start around 75% electrode negative on clean material. Use roughly 65% EN for dirty or heavily oxidized aluminum and move toward 80% EN on very clean parts. If you see black peppering in the puddle, reduce EN, but not below about 60%.

Why does my tungsten keep balling and the arc wander on AC?

You are spending too much time electrode positive, which drives excess heat into the tungsten. Increase the EN percentage of your balance, and verify you are running a pure, zirconiated, ceriated or lanthanated tungsten suited to AC rather than a thoriated one on a transformer machine.

What do the black dots in my aluminum weld puddle mean?

That is peppering, caused by incomplete oxide removal. Increase the electrode-positive cleaning portion by lowering the balance setting, but first improve mechanical cleaning with acetone and a dedicated stainless wire brush, since balance is only the last line of defense.

Does AC frequency change polarity?

No. AC frequency is how many times per second the current switches polarity, typically 60 Hz on older machines and adjustable up to 150 Hz or more on inverters. Higher frequency tightens and focuses the arc; it does not change the cleaning-to-penetration ratio, which balance controls.

Which tungsten is best for AC aluminum TIG welding?

On transformer or sine-wave machines, pure green or zirconiated brown tungsten with a balled tip. On modern inverter AC machines, ceriated, lanthanated or rare-earth electrodes ground to a point hold better because balance can run much further toward electrode negative.

Is DCEP ever used for aluminum TIG?

Rarely. DCEP cleans exceptionally well but sends about 80% of the arc heat into the tungsten, giving shallow penetration and rapid electrode failure. It is limited to very thin gauge work, brief cleaning passes or demonstration purposes.