Stainless TIG Welding Polarity: DCEN, DCEP & AC Explained

The tungsten touches down for a split second, the arc snaps on, and instead of the tight, hissing bead you expected, the tip balls up like a melted candle and the stainless turns gray before you even add filler. Nine times out of ten the machine is fine, the gas is fine, and the tungsten is fine — the leads are simply plugged in backwards. Polarity is the quiet setting that decides whether stainless TIG feels effortless or fights you on every inch.

Stainless TIG Welding Polarity: The Short Answer Before the Deep Dive

A small shop outside Dayton takes in a repair job: a cracked 304 stainless exhaust manifold pulled off a diesel truck. The fabricator grinds out the crack, sets 65 amps on a dial that has never let him down, and strikes an arc. The tungsten balls up before he can even dip the filler. The bead lays down wide and gray, the arc hisses instead of hums, and he spends twenty minutes blaming gas flow, then the tungsten, then himself. What he never checks is the one thing that takes ten seconds: the torch lead is plugged into the positive terminal. Swap it to negative, ground to positive, and the same machine, gas, and electrode suddenly run a tight, quiet bead.

The short answer: for stainless steel TIG welding, use DCEN, which stands for direct current electrode negative and is also called straight polarity. Torch on negative, ground clamp on positive. That is the setting for sanitary tube, sheet metal, exhaust work, food-grade fittings, and just about every stainless joint you will ever run. In DCEN, roughly two-thirds of the arc heat lands at the positive pole, which is the workpiece, so the metal melts and the tungsten stays cool and pointed.

DCEN is not just the popular choice; it is the physics-friendly one. A cool tungsten emits electrons steadily, holds a sharp taper, and keeps the arc column narrow. Heat goes where you want it, penetration runs deep relative to bead width, and the heat-affected zone stays small. That combination matters more on stainless than on almost any other common base metal.

DCEP, or direct current electrode positive, flips the circuit. The torch becomes the positive pole and absorbs roughly two-thirds of the arc energy, so a 1/16 in. tip turns into a molten ball, sheds metal into the pool, and leaves you with a shallow, wide, ugly weld. For stainless TIG this is essentially never correct, though it has a narrow history in aluminum work. AC is the third option. Alternating current swaps polarity continuously, and on 60 Hz power the polarity reverses 120 times per second. Each cycle gives you a cleaning half and a penetration half, which is why AC dominates aluminum work and only occasionally shows up on stainless.

A Mental Map for the Three Polarities

• DCEN (straight polarity): heat into the work. Torch negative, ground positive. Deep fusion, narrow bead, long tungsten life. The stainless default.
• DCEP (reverse polarity): heat into the tungsten. Torch positive, ground negative. Broad, shallow weld, fast electrode erosion. Almost always the wrong answer here.
• AC: heat traded back and forth, with an adjustable balance between cleaning and penetration. A compromise mode that solves aluminum’s oxide problem but buys stainless little.

The terminology is where most confusion starts, because the same setting has three names. Straight polarity means DCEN. Reverse polarity means DCEP. Electrode negative means the torch is negative. ESAB’s introduction to GTAW polarity and Lincoln Electric’s AC/DC polarity explainer both spell out the same heat split, and it is worth reading once so the shop talk stops sounding like a foreign language.

The reason stainless tolerates no cleaning half-cycle is its oxide. Aluminum carries a thick, stubborn oxide that reforms in seconds and needs electrode-positive time to break up; that is why aluminum jobs lean on AC or dedicated cleaning procedures, a theme covered in guides like Aluminum MIG Welding: The Complete Practical Guide. Stainless wears a chromium-rich passive film only nanometers thick. It does not need cathodic bombardment, and paying for it with heat and tungsten erosion is a bad trade.

Everything past this point is detail on that foundation: how to set the machine and prep tungsten, what AC balance and frequency actually do on stainless, why DCEP destroys tips, how to read symptoms like balling, sooting, sugaring, and arc wander, what heat input does to sensitization and ferrite, how DCEN stacks up against DCEP and AC on real coupons, and why hexavalent chromium fume makes ventilation non-negotiable. If you only remember one line, make it this: stainless TIG is DCEN until proven otherwise.

What Polarity Actually Means in a TIG Circuit

Polarity sounds like electrical theory, but in TIG it is really a question of where the heat goes. Picture the complete circuit: the power source sends current through the torch lead to the tungsten electrode, across the arc column, into the workpiece, and back through the ground cable to the machine. Every part of that loop has to be connected, and polarity simply describes which side of the loop is positive and which is negative.

That sounds simple until the terminology gets involved. Electrons physically flow from negative to positive. Conventional current arrows, however, point from positive to negative. Both descriptions are correct, just from different viewpoints, which is why two welders can describe the same lead arrangement and sound like they disagree. The practical takeaway is to ignore the arrow debate and remember the connection: in DCEN, the torch and tungsten are negative, and the work clamp is positive.

The Same Setting, Three Different Names

Shop talk creates much of the confusion. Straight polarity and electrode negative describe the same DCEN connection. Reverse polarity and electrode positive describe DCEP. Manufacturers, manuals, and old-timers all use different labels for identical settings, so it helps to translate every term back to one question: is the torch negative or positive? The ESAB guide to GTAW polarity makes the same point in more technical language.

Term you hear	What it actually means
Straight polarity	DCEN: torch negative, work positive
Reverse polarity	DCEP: torch positive, work negative
Electrode negative	Same as DCEN
Electrode positive	Same as DCEP
AC	Polarity reverses every half cycle

Heat, Electrons, and Why the Tungsten Cares

In a DC arc, roughly two-thirds of the heat appears at the positive pole and about one-third at the negative pole. That single fact explains most polarity behavior. With DCEN, the work is positive, so the workpiece receives the majority of the heat while the tungsten stays comparatively cool. The electrons leave the tungsten and travel toward the work, and the arc remains stable because the electrode is not being bombarded by heavy positive ions on its way out. A focused, narrow arc and a tungsten that holds its shape are the visible results.

Switch to DCEP and the roles reverse. The tungsten becomes the positive pole, so it absorbs the bulk of the heat instead of the plate. That is why electrode-positive TIG demands a much larger tungsten for the same amperage, and why a small one will ball up or spit metal if you push it. The trade-off is that the ion bombardment that damages the tungsten is also what scrubs the workpiece: positive ions striking the surface break up aluminum oxide and other tenacious films. In other words, DCEP buys cleaning action at the cost of electrode life and joint penetration.

AC works the same circuit in both directions. The polarity swaps every half cycle, so neither side gets a permanent advantage. One half cycle behaves like DCEN and puts heat into the work; the other behaves like DCEP and breaks up surface oxides. The ratio between those halves is what a balance or cleaning-and-penetration control adjusts on a modern machine.

One more detail worth knowing: polarity is set at the machine terminals, not at the torch. On a dedicated TIG welder the leads are usually fixed internally, and you choose the process mode on the panel. On a stick machine or a multiprocess unit running scratch-start TIG, you physically move the leads. If DCEN is selected but the stinger holds the torch, you have created DCEP without meaning to, and the symptoms show up as a glowing, melting tungsten within seconds of striking an arc.

Why Stainless Steel Wants DCEN and Almost Nothing Else

The reason stainless behaves so differently from aluminum under the arc is not stubbornness or some quirk of the alloy. It comes down to a film you cannot see. Stainless steel protects itself with a chromium-rich passive oxide layer only a few nanometers thick, tightly bonded to the surface and able to rebuild itself in ordinary air within minutes of being scratched. That film is the whole point of stainless. It is also thin enough that a TIG arc punches through it without any dedicated cleaning phase.

Aluminum is the opposite story. Its oxide grows thick, refractory, and reforms almost instantly after you remove it. Aluminum oxide melts at roughly 3,700°F, far above the metal underneath, so the arc has to break it up before the puddle can even form. That is why aluminum TIG leans on the electrode-positive half of AC for cathodic cleaning, and why aluminum prep is a whole discipline of its own — you can read more about surface prep in this guide to aluminum cleaner for welding.

Stainless has no such problem. Its passive layer is thin, self-healing, and does not demand aggressive cleaning action. So when you run DCEP, or the electrode-positive portion of an AC cycle, on stainless, you are paying for a service you did not order. The arc spends part of its time blasting heat into the tungsten instead of the joint, and the only visible result is a wider etch band and a tip that starts to look like a blunt pencil.

What DCEN Actually Buys You

With the torch on negative and the work lead on positive, roughly two-thirds of the arc energy lands in the workpiece, a heat split that ESAB’s GTAW polarity primer describes as the foundation of efficient DC TIG work. On stainless, that translates into a set of practical advantages you feel on the very first bead:

• Concentrated heat-affected zone. Heat stays tight around the arc instead of spreading outward in a lazy halo.
• Deep fusion with a narrow bead. At the same amperage, DCEN penetrates further and produces a crowned, defined profile rather than a flat, wide puddle.
• A stable, quiet arc. Fewer arc jumps, less wandering, and a puddle that behaves predictably.
• Long tungsten life. A cool tungsten keeps its point, which means fewer balled tips, less erosion, and far fewer tungsten inclusions in the pool.
• Faster travel speed. More usable heat at the joint means you move sooner and finish sooner.
• Less distortion. Because the heat is concentrated and you travel faster, total heat input per inch of weld drops, so sheet panels and thin-wall tube pull and warp far less.

There is a second, quieter benefit that experienced hands notice right away: the weld comes out cleaner. No cleaning half of the cycle means no chalky white etch band flanking the bead, so less post-weld pickling and brushing to restore the passive layer. On sanitary tubing and architectural work, where appearance is part of the specification, that alone justifies DCEN.

What Happens When You Stray

DCEP on stainless concentrates heat at the tungsten, which is exactly where you do not want it. The tip overheats, balls up, and starts shedding into the puddle. Penetration goes shallow and wide at the same time, the HAZ balloons, and the part walks. You end up turning the amperage up to compensate for the loss at the joint, which only makes the tungsten problem worse. It is a loop with no good exit.

Polarity	Heat balance	Result on stainless
DCEN	About two-thirds to the work	Deep, narrow fusion; cool tip; tight HAZ
DCEP	About two-thirds to the tungsten	Balled tip, shallow wide puddle, excessive etch
AC	Split between both	Cleaning action stainless never needed

DCEN is not a magic setting, though. It simply puts the heat where fusion happens, and from there the usual rules apply: fit-up tight, keep the arc length short, and let travel speed rather than amperage govern the bead. Once polarity is right, everything else about stainless gets easier to read.

Setting Up the Machine: Leads, Tungsten, Gas, and Starting Amps

Once you know stainless wants DCEN, the rest of the setup becomes a sequence rather than a guessing game. Get the leads, mode, tungsten, gas, and amperage right in that order, and the machine stops fighting you. Skip a step, and you will spend the next hour blaming the gas or your technique for a problem that started at the front panel.

1. Connect the leads for DCEN

For DCEN TIG, the torch lead goes to the negative terminal and the work or ground clamp goes to the positive terminal. That is the whole rule, but it is worth checking twice. As ESAB’s GTAW polarity primer explains, the torch is the negative pole in DCEN, which is why the heat concentrates at the workpiece instead of the tungsten.

Do not trust the color of the connectors or the way the last welder left the machine. Terminal labeling varies between brands, and multi-process units often share sockets between TIG, stick, and MIG. Read the owner’s manual for your specific model, especially if the machine has a polarity switch or a dinse socket layout that looks ambiguous. Ten seconds with the manual beats a tungsten that balls up on the first tap.

2. Select DC and a start mode

Set the process selector to DC TIG, not AC. This sounds obvious, but many machines default to AC after an aluminum job. The difference between current type and polarity is a common source of confusion, and Lincoln Electric’s explanation of AC and DC polarity is a useful refresher: DC can be electrode negative or positive, while AC alternates between the two.

Next, choose high-frequency start for bench work. It gives you a clean, contact-free arc start and keeps the tungsten out of the puddle. Lift-arc is the better choice in the field, around sensitive electronics, or anywhere the HF noise is a problem. Kemppi’s control-panel documentation shows how these start modes are typically presented, though the exact labels differ by model.

3. Choose and prep the tungsten

For stainless on DCEN, 2% lanthanated, ceriated, or a rare-earth blend are the modern defaults. They hold a stable tip, start easily, and tolerate a wide current range. 2% thoriated tungsten still works well, but it contains a radioactive oxide. Grinding dust and inhalation controls matter, which is why many shops have moved to lanthanated or ceriated alternatives.

Grind the electrode longitudinally, not around its circumference. A taper roughly 2 to 2.5 times the electrode diameter is a good starting point, and the grind marks should run axially toward the tip. That orientation helps the arc stay where you point it instead of wandering. Leave a small blunt tip rather than a needle point, especially at higher amperage; a sharp point overheats and erodes faster.

4. Set argon coverage

Use 100% argon for stainless TIG. With a gas lens, 15 to 20 cfh is the usual working range, and the cup should be roughly two to three times the tungsten diameter. A larger cup gives better coverage on outside corners and tubing, while a smaller cup gets into tight joints.

Set a pre-flow of about one second so gas is already covering the joint when the arc starts. Then extend the post-flow until the weld stops glowing, which is often longer than welders expect. Stainless is sensitive to oxidation, and cutting the gas off while the bead is still hot is a fast way to get gray or sugared welds. On sanitary tubing, and in aerospace TIG welding, the gas side of the setup often matters more than the machine settings.

5. Set a starting amperage

A common rule of thumb for steel is about 1 amp per 0.001 in. of thickness. For stainless, trim that by roughly 10 to 20% because the material holds heat and distortion becomes a real risk. Tulsa Welding School’s overview of welding current covers the same relationship between thickness and amperage.

Material thickness	Baseline steel amps	Stainless starting range
0.030 in.	30 A	25–30 A
0.065 in.	65 A	55–65 A
0.125 in.	125 A	105–125 A
0.250 in.	250 A	210–250 A

Treat those numbers as a place to start, not a prescription. Joint type, fit-up, travel speed, and whether you are welding a corner or a butt all change the right setting. On thicker stainless, bevel the joint and plan for multiple passes rather than chasing penetration with maximum amps.

Pulse is a heat-control tool, not a polarity setting. It switches between peak and background current to cool the puddle between pulses, which helps on thin gauge and out-of-position work. It does not change the fact that you are welding DCEN.

With the leads correct, DC selected, the tungsten prepped, gas covering the puddle, and amps in the right neighborhood, stainless TIG becomes predictable. Most of the remaining problems trace back to one of those five steps, and the first one to check is always the lead connection.

AC on Stainless: When It Helps and When It Hurts

Here is the question that fills forum threads every week: if AC is the aluminum setting, can you run it on stainless? The honest answer is yes, you can. Whether you should is a completely different conversation. AC is not forbidden on stainless, but it is also not a default, and treating it like one is how welders end up with gray, sandblasted-looking beads and a tungsten that looks chewed instead of tapered.

Think of AC as a problem-solving tool that lives in the bottom of the drawer. Most stainless jobs never need it. A handful of specific situations genuinely benefit from it. Knowing which is which separates a welder who understands polarity from one who just memorizes settings.

What AC Is Actually Doing in a Stainless Puddle

AC TIG alternates between electrode-negative and electrode-positive half cycles many times a second. On the electrode-negative half, heat drives into the workpiece and you get penetration. On the electrode-positive half, the polarity flips: electrons bombard the tungsten, heating it instead of the plate, while positive ions scouring the surface break up oxides. That is the cathodic cleaning action, and it is why AC is the standard choice for aluminum.

Aluminum’s oxide layer is thick, refractory, and reforms almost instantly, so it genuinely needs that aggressive scrubbing. Stainless is a different animal. Its chromium-rich passive film is only nanometers thick, tight, and easy to disrupt with a wire brush and solvent. You can read more about why oxide removal dominates aluminum work in our guide to prepping aluminum for welding, and the contrast explains a lot.

Because stainless does not need the cleaning half-cycle, that portion of an AC wave is mostly overhead. It heats the tungsten, widens the etched band beside the bead, and steals time from the half of the cycle that actually fuses metal.

The Legitimate Cases for AC on Stainless

These are edge cases, not recommendations. But they exist.

• Magnetic arc blow on heavy or long joints. Arc blow is a DC phenomenon where magnetic forces deflect the arc, making it snake and wander. On thick sections or long runs, switching to AC can calm that deflection enough to finish the pass.
• Light surface contamination. A heavily electrode-negative balance, roughly 80 to 90 percent EN, gives you a small amount of cleaning without going full aluminum-style wave. Useful on a repair or cast part that will not come fully clean.
• Thin gauge where heat is the enemy. Since half the cycle heats the tungsten instead of the plate, total heat into the workpiece drops. On light sheet, that can be a way to keep distortion in check when DCEN wants to run away from you.

Where AC Quietly Hurts the Job

• The etched band. That frosted, gray halo on either side of the bead is the cleaning action at work, and on stainless it is cosmetic damage. The roughened surface is harder to passivate and traps contamination, which is bad news on food, pharma, or sanitary work.
• Tungsten erosion and inclusions. The positive half-cycle bombards the electrode. Tips ball, split, and spit, and a stray piece of tungsten in a stainless weld is a defect that shows up on radiography and in corrosion testing.
• Lost penetration and a lazy arc. With less heat reaching the plate, fusion gets shallow on anything thicker than sheet, and the arc tends to wander and crackle at low amperage instead of locking in.

The Bottom Line

DCEN remains the default for stainless for good reason: it puts the heat where the fusion is and leaves the passive film and the tungsten alone. AC is the exception you reach for to solve a specific problem, then put back in the drawer. If you are running AC on stainless as a habit rather than a fix, the gray bead and chewed tungsten are telling you exactly what to change.

DCEP and Reverse Polarity: Why It Melts Your Tungsten

If DCEN is the setting that makes stainless TIG feel effortless, DCEP is the one that makes it feel broken. Direct current electrode positive, which the old shop handbooks call reverse polarity, puts the torch on the positive side of the circuit. That single swap reverses the whole heat balance of the arc, and the tungsten pays for it.

The Physics: Heat Follows the Positive Pole

In any arc, roughly two-thirds of the total heat lands at the positive terminal. Under DCEN the work is positive, so the workpiece absorbs the bulk of the energy and the tungsten stays comparatively cool. Under DCEP the equation flips. Electrons boil off the workpiece and accelerate toward the torch, bombarding a tip that may be only 1/16 in. across. All that energy converges on a few square millimetres of tungsten.

ESAB’s introduction to GTAW polarity lays out the same split, and Lincoln Electric’s AC/DC polarity explainer makes the practical consequence clear: the electrode has to survive a heat load it was never sized for. A 1/16 in. lanthanated tip that runs all day at 120 amps on DCEN will ball, spit, and erode in seconds on DCEP at the same current.

What a DCEP Arc Does to a Stainless Bead

The damage is not subtle. A DCEP TIG arc on stainless produces shallow penetration, a wide bead, and rapid tungsten consumption. You get a puddle that spreads sideways instead of sinking in, a frosted etch band along both toes, and a tip that sheds tungsten directly into the pool. Those tungsten fragments show up later as hard inclusions under a grinder.

Penetration suffers for a simple reason: the energy that should have melted the root went into the electrode instead. Welders often respond by cranking the amperage, which makes the tungsten deteriorate faster while the bead stays stubbornly flat. The arc also wanders, because a molten, irregular tip gives the arc no stable emission point to anchor to.

Where DCEP Actually Came From

DCEP is not useless. It is just historical. In the pre-inverter era, welders ran aluminum on DCEP using large, balled pure tungsten electrodes. The electrode-positive half of the cycle blasted away the heavy aluminum oxide layer, which was the only practical way to clean the metal before square-wave machines existed. The trade-off was a massive, balled tip and limited amperage, and everyone accepted it because there was no alternative.

That legacy survives in some specialty aluminum and repair work, and in older procedure sheets written for transformer machines. It has essentially no place in stainless TIG. Here the oxide film is nanometres thick and chemically stable, so there is nothing for cathodic cleaning to remove. If you are chasing aluminum settings instead, the process behaves completely differently, and guides like our aluminum MIG welding practical guide cover the polarity logic that actually applies there.

The Crossover Confusion: MIG and Stick Habits

Most DCEP mistakes are not mysteries. They are carryover. Solid-wire GMAW normally runs DCEP, so a welder who spends the week running MIG walks to the TIG machine and plugs the torch into the positive terminal out of pure muscle memory. Stick welding compounds it: many SMAW electrodes run on DCEP or AC depending on their classification, so polarity habits learned on 7018 do not transfer to stainless TIG either. Our breakdown of why SMAW still wins in so many shops explains how electrode-dependent that logic really is.

Forums add a second layer of confusion. People use “reverse polarity” loosely to mean AC, or electrode positive in general, or sometimes just “the setting that isn’t working.” In TIG terms, reverse polarity means exactly one thing: torch positive, work negative. If someone tells you to run reverse polarity on stainless and does not mean DCEP, ask for the machine setting, not the label.

Switches, Cables, and Multi-Process Machines

Note: Some machines change polarity with a selector switch or a digital menu; others require you to physically move the torch and work leads between terminals. Multi-process units are the usual culprits, because the terminal labelled “torch” in MIG mode is not the one you want in TIG mode. Read the manual’s TIG page, not the front panel artwork, and confirm the torch is on negative before the arc starts.

The fix is embarrassingly cheap. Ten seconds with two cable ends prevents an hour of grinding out a shallow, contaminated bead, and it prevents a tungsten inclusion that could fail inspection. On stainless, DCEP is never the answer. It is the symptom.

Polarity Problems and Troubleshooting: Symptoms, Causes, Fixes

Most stainless TIG failures get blamed on gas or shaky hands. In reality, a small set of symptoms points straight back to polarity, torch setup, or shielding. Use the table to triage the problem, then work through the detailed fixes below.

Symptom	Most likely cause	First move
Tungsten balls instantly at low amps	Leads reversed, AC selected, or tungsten too small	Torch to negative, select DC, step up tungsten size
No arc, or arc only starts with a scratch	HF disabled, poor ground, contaminated tip	Clean and repoint tungsten, clamp bare metal, enable HF
Wide, flat, shallow bead with a sooty edge	DCEP or excessive electrode-positive AC balance	Verify DCEN; raise electrode-negative balance
Black soot or gray smut around the bead	Low flow, wrong gas, short post-flow, or draft	Set argon to 15–20 cfh, extend post-flow, block drafts
Sugaring inside tubing	No purge or insufficient purge	Stop, purge the joint, verify oxygen level
Blue, purple, or gray tint instead of straw	Excess heat input, slow travel, lost shielding	Reduce amps, increase travel speed, check gas coverage
Arc wanders and snakes	Arc blow, residual magnetism, poor ground, offset tip	Relocate ground, shorten arc, regrind tip
Porosity	Moisture or oil, gas leak, turbulent flow	Clean the joint, leak-test the torch, reset flow
Tungsten inclusion	Dipped tip, DCEP erosion, filler touching tungsten	Grind out the defect, repoint, shorten arc length
Cracking	Wrong filler, excessive restraint, sensitization	Remove the crack, change filler, control heat input

The Ten Symptoms in Detail

1. Tungsten balls instantly at low amperage. Fast fix: check that the torch is negative and the work clamp is positive, then confirm the machine is in DC mode. Root-cause fix: match tungsten diameter to current and use a longitudinal taper that suits the amperage. On a new machine, read the terminal labels rather than assuming; multi-process units are not always marked the way you expect. ESAB’s GTAW polarity primer is a useful refresher.
2. Arc will not start, or starts only with a scratch. Fast fix: repoint the tungsten, clamp the ground to clean bare metal, and confirm high-frequency start is enabled. Root-cause fix: repair damaged ground cables, use a solid work-clamp location, and learn the correct lift-arc gap if that is your start method.
3. Bead is wide, flat, and shallow with a sooty edge. Fast fix: switch to DCEN immediately. If you are intentionally running AC, raise the electrode-negative balance. Root-cause fix: learn whether your machine expresses balance as a percentage of cleaning or penetration, because conventions differ. Lincoln Electric’s AC/DC polarity guide explains the difference.
4. Black soot or gray smut around the bead. Fast fix: verify argon flow at roughly 15–20 cfh, extend post-flow, and shield the weld from drafts. Root-cause fix: check for a wrong gas, a leaking hose, or a clogged gas lens. More flow is not always better because excessive flow causes turbulence that aspirates air into the shield.
5. Sugaring on the back side of tubing. Fast fix: stop and purge before continuing. Root-cause fix: seal the purge chamber properly, allow enough pre-purge time, and measure oxygen. Sugaring means the root was not protected by adequate purge gas, and no polarity change will repair it.
6. Weld turns blue, purple, or gray instead of straw. Fast fix: reduce amperage, increase travel speed, and confirm gas coverage. Root-cause fix: lower total heat input with pulse or chill blocks and check for lost shielding. Straw or light gold tint is generally acceptable; deep blue, purple, or gray signals excessive oxidation and heat input.
7. Arc wanders and snakes. Fast fix: move the ground clamp closer to the joint, shorten the arc, and regrind a contaminated or offset tip. Root-cause fix: address magnetic arc blow. It comes from magnetic fields, often from DC current in long joints or residual magnetism in the part, and it appears as a wandering or snapping arc. Demagnetizing the part or changing the ground location often solves it.
8. Porosity. Fast fix: clean the joint with a dedicated stainless brush and solvent, leak-test the torch, and reset gas flow. Root-cause fix: eliminate moisture, oil, hose leaks, and turbulence from excessive flow. Drafts and a bad gas lens can create the same defect.
9. Tungsten inclusion. Fast fix: grind out the defect, repoint the electrode, and shorten arc length. Root-cause fix: stop dipping the tip, keep filler away from the tungsten, and correct any DCEP condition that is eroding the electrode. Tungsten inclusions occur when the electrode contacts the pool or filler, or when the tip overheats and erodes.
10. Cracking. Fast fix: remove the crack completely before attempting a repair. Root-cause fix: match filler to the base metal, reduce restraint, and control heat input. Unstabilized grades held too long in the sensitization range are more vulnerable, so travel speed and interpass temperature matter.

Metallurgy Matters: Heat Input, Sensitization, Ferrite, and Purge

A stainless weld can look beautiful and still fail six months later. The bead is clean, the toes wet in nicely, and the heat tint is barely straw. Then the part goes into a chlorinated washdown, a food line, or a steam header, and the corrosion starts at the grain boundaries just off the weld. Polarity is part of that story. Not because DCEN magically changes the metallurgy, but because it controls where the heat goes and how long the metal sits at damaging temperatures.

Heat Input: DCEN Travels Faster and Deposits Less Heat

With the torch negative, roughly two-thirds of the arc energy lands in the workpiece, which is exactly what you want for fusion. The arc is tight, the puddle is controllable, and you can move. That faster travel matters more than most welders realize. Heat input per inch is a function of amps, volts, and travel speed, so a concentrated DCEN arc that lets you run 15% faster can cut total heat input substantially compared with a wandering, wide arc that forces you to linger.

Lower heat input means a narrower heat-affected zone, less distortion, and less heat tint. The heat distribution in GTAW is the reason DCEN is the metallurgically friendly default for stainless, not just the convenient one. Stringer beads, skip welding, and allowing the part to cool between passes all build on that advantage.

Sensitization: The 800–1500°F Danger Zone

Unstabilized austenitic grades like 304 and 316 are vulnerable to sensitization. Hold them in the roughly 800–1500°F (427–816°C) range and chromium carbides can precipitate at the grain boundaries. That ties up chromium exactly where the passive film needs it, leaving a depleted zone that corrodes first. In a weld, the band just outside the fusion line is the usual suspect.

Low-carbon grades such as 304L and 316L limit carbon to about 0.03%, which dramatically reduces the amount of carbide that can form. Stabilized grades like 347 use niobium for the same purpose. Even with the right grade, heat input still matters: keep interpass temperatures down, avoid wide weaving, and let thin parts cool rather than dumping more amps at a joint that is already hot.

Ferrite: A Little Goes a Long Way Against Cracking

Fully austenitic weld metal can crack during solidification. A small amount of delta ferrite helps resist that cracking by changing how the solidifying grain boundaries behave. Weld procedures commonly target roughly 3–10 FN (ferrite number) in austenitic stainless weld metal. Filler selection and dilution from the base metal control it. Over-alloyed fillers and excessive dilution can push ferrite outside the desired window, which is one reason matching the filler to the actual base metal matters.

Filler Selection per AWS A5.9

AWS A5.9 covers bare stainless steel welding electrodes and rods. The classifications below cover most stainless TIG work, though the procedure and service environment always have the final say.

Base Metal / Joint	Typical Filler
304 / 304L	ER308L
316 / 316L	ER316L
Dissimilar, carbon steel to stainless	ER309L
Stabilized service (347 type)	ER347
Duplex stainless	ER2209

Use the L-grade filler on L-grade base metal, and do not assume more alloy is better. ER316L in a 304 joint adds molybdenum you did not ask for. ER309L is the workhorse for dissimilar joints because its higher alloy content tolerates dilution from the carbon steel side.

Back Purge: Oxygen Is the Enemy on the Back Side

On tubing and sanitary pipe, the back side of the root has no shielding gas unless you supply it. Argon purge displaces oxygen, and the specification is often strict: sanitary work commonly calls for below 100 ppm oxygen, while some pharmaceutical specifications require below 50 ppm. Measure with an oxygen analyzer rather than guessing by color.

This is where polarity and heat input connect directly. A hotter, wider arc produces a wider oxidized zone and a dirtier root, which demands more purge margin and more cleanup. A tight DCEN arc keeps the hot zone small and makes the purge easier to hold. Purge until the meter stabilizes at the target, seal the ends, and vent so the gas actually flows past the root.

Post-Weld Cleaning and Passivation

Heat tint is more than cosmetic. It is an oxide layer that sits on top of a chromium-depleted surface. Pickling removes the tint and the depleted layer, then passivation restores the passive film. ASTM A380 covers cleaning, descaling, and passivation of stainless steel parts, and it is the reference most specifications point to for post-weld restoration.

None of this replaces clean metal. Polarity, purge, and filler choice all assume the joint was degreased and brushed with a dedicated stainless brush. Get those basics right and the metallurgy takes care of itself.

DCEN vs DCEP vs AC for Stainless: Side-by-Side Comparison

Spec sheets and forum arguments make this harder than it is. Line the three modes up against the same criteria and the hierarchy on stainless becomes obvious.

Criterion	DCEN (Straight Polarity)	DCEP (Reverse Polarity)	AC
Lead connection	Torch negative, work positive	Torch positive, work negative	Alternates every half-cycle
Heat distribution	About two-thirds into the work	About two-thirds into the tungsten	Split, shifted by balance setting
Penetration	Deepest	Shallowest	Middle, adjustable
Bead shape	Narrow, crowned	Wide, flat, shallow	Medium with a wider etch band
Tungsten life	Excellent	Poor	Moderate to poor
Cleaning action	None to minimal	Strong cathodic cleaning	Adjustable
Arc stability	Very stable	Unstable at low amps	Stable on squarewave, noisier
Typical stainless use	Default for nearly everything	Rarely, if ever	Niche: arc blow, light contamination, thin gauge
Typical non-stainless use	Steel, titanium, most alloys	Legacy aluminum repair	Aluminum and magnesium
Skill demand	Low to moderate	High	Moderate to high

What the Table Means on Real Work

DCEN produces the deepest penetration and narrowest bead of the three modes on stainless at equal amperage. That is not a preference, it is physics: most of the arc energy lands in the workpiece instead of the electrode, as ESAB’s GTAW polarity primer explains. DCEP reverses that bargain, producing the shallowest penetration and fastest tungsten consumption, which makes it unsuitable as a stainless TIG default. AC sits between the two, offering cleaning at the cost of heat control and tungsten life. Lincoln Electric’s polarity guide is a useful refresher on how current type and polarity are separate choices.

• 20-gauge sheet metal enclosure: DCEN at roughly 40–60 A, with pulse around 100 Hz and 25% background current to keep the edges from curling.
• 1/8 in. sanitary tube: DCEN at roughly 90–120 A, a gas lens, and a full argon purge. Polarity is settled; fit-up and purge quality decide the result.
• Cast stainless repair: DCEN, preheat only if the procedure calls for it, and a nickel-based filler when dilution or cracking risk demands it.
• Thin stainless foil: DCEN at 5–20 A with a fine, sharp tip. DCEP is not a candidate here; it would ball the tungsten before the puddle forms. Work at this scale shows up in aerospace TIG welding, where low-amp stability matters more than headline output.
• Stainless to carbon steel: DCEN with ER309L filler to accommodate dilution from both sides.

AC’s Narrow Niche, With Numbers

AC is not forbidden on stainless, just rarely the best tool. It earns a place on long, heavy joints where magnetic arc blow makes DC wander. On a 16-gauge coupon, a typical AC starting point might be 70% electrode-negative balance at 60–100 Hz; raising electrode-negative time adds penetration and reduces cleaning, while lowering it cleans more but heats the tungsten and widens the etch band. Some machines present balance inversely, so test on scrap. Compare that with plain DCEN at the same 70 A and the DC arc is tighter, quieter, and leaves less tint. Pulse is usually the better answer for thin stainless than switching to AC.

Don’t Carry Other Processes’ Polarity Rules Into TIG

GMAW solid wire typically runs DCEP, flux-cored wire often does too, and SMAW polarity depends on the electrode classification. Plasma cutting is different again. Welders who move between processes often assume “reverse polarity” means the same thing everywhere, then wonder why their stainless TIG tungsten is melting.

Machine Features That Change the Comparison

Pulse alternates peak and background current to control heat input without changing polarity. Mixed AC/DC modes run a set percentage of AC cycles inside an overall DC output, blending cleaning and penetration. Waveform choice (sine, square, triangular) and high-frequency start shape the arc further. A DC-only TIG machine is sufficient for stainless, carbon steel, and most exotic alloys; AC capability matters mainly for aluminum and magnesium. If stainless is the target, prioritize pulse control and good low-amp stability over maximum amps. Entry-level DC units handle fabrication fine, mid-tier machines add better pulse and start control, and premium units buy you waveform and balance flexibility you may never use.

Safety, Fume, Practice Plan, and Final Takeaways

Lead connection, arc behavior, and heat distribution shape the weld you can see. What follows shapes the welder holding the torch. Stainless TIG looks like one of the cleaner processes in the shop. For air quality, it is one of the most deceptive.

Hexavalent Chromium: The Part You Can’t See or Smell

Stainless steel gets its corrosion resistance from chromium. The arc converts some of that chromium into a different chemical form: hexavalent chromium, Cr(VI). It rides the fume and is a known carcinogen, which is why it has a dedicated OSHA standard instead of being folded into general welding fume rules.

Under 29 CFR 1910.1026 for general industry and its construction counterpart 29 CFR 1926.1126, the permissible exposure limit is 5 micrograms per cubic meter as an 8-hour time-weighted average. The action level sits at 2.5 µg/m³. Those numbers are small enough that a welder on stainless indoors can pass them in a single shift without any visible haze in the air.

Engineering Controls Come First

PPE is the last line of defense, not the first. Before you reach for a respirator, fix the air.

• Local exhaust ventilation (LEV): A source-capture hood or extraction torch arm close to the arc is the most effective control you have. OSHA’s hexavalent chromium guidance treats LEV as a primary engineering control. Distance kills capture velocity, so put the hood near the plume, not across the bench.
• Dilution ventilation: A supplement, never a substitute. Open doors and shop fans move air but rarely capture fume at the arc.
• Confined spaces: Argon is heavier than air and pools in tanks, pits, and vessel interiors. It displaces oxygen silently with no warning smell. Treat any argon purge in a confined space as an asphyxiation hazard and follow atmospheric testing and attendant procedures.

For wider reading, the NIOSH welding topic page and OSHA’s welding safety fact sheet are solid starting points.

PPE That Actually Fits the Job

• Helmet and shade: Match shade to amperage, and check auto-darkening delay and sensitivity settings before low-amp work.
• Respirator: Where exposure assessment shows a need, use a fitted respirator with current fit testing. A leaking half-mask only gives false confidence.
• Skin and eyes: Flame-resistant sleeves, leather gloves, and safety glasses under the helmet for grinding and brushing.

Shops that also run aluminum will find some prep habits overlap. Our guide to prepping aluminum for welding covers contamination control in more detail.

A Four-Week Plan That Builds Real Skill

Polarity knowledge sticks once your hands have felt the difference. This progression costs a few pounds of scrap and teaches diagnosis rather than memorization.

1. Week one: Bead plates on 16-gauge stainless in DCEN at fixed amperage. Change travel speed only, and read heat tint for inches per minute.
2. Week two: Butt joints with a back purge. Judge every weld by the inside, on sugar, not on the outside bead.
3. Week three: Outside corners and tubing joints on pulse, keeping the arc on the joint and the heat under control.
4. Week four: Deliberate fault-finding. Swap the leads, run a bead, record symptoms, correct it. Repeat with AC mode and a reversed balance setting.

Pre-Weld Checklist

Check	Confirm
Polarity	Torch negative, ground positive (DCEN)
Mode	DC, not AC
Tungsten	Correct type and diameter, ground longitudinally
Gas flow	100% argon, roughly 15–20 cfh with gas lens
Purge	Oxygen measured, not guessed
Cleanliness	Dedicated stainless brush, solvent-cleaned joint

Final Takeaways

The core rule stays simple: stainless TIG runs on DCEN, torch negative, ground positive. Get that wrong and no gas tuning or technique will save the bead.

Three exceptions are worth keeping in your pocket. AC can be a legitimate tool for heavy sections where arc blow makes the arc wander, or for light surface contamination with a strongly electrode-negative balance. DCEP still has a place in specialty aluminum work but essentially none in stainless TIG. Borderline-thin material is usually better handled with pulse and a finer tungsten than with a cleaning-heavy AC setting.

Polarity is a five-second check. One glance at the terminals or one scrap coupon prevents hours of grinding, rework, and re-polishing heat tint. If stainless TIG ever feels impossible, check the leads before you doubt yourself. For problems beyond polarity, our aerospace TIG welding standards and technique guide covers arc control and cleanliness at a higher level of rigor.

Disclosure: some links on this site may be affiliate links. If you buy through them, we may earn a small commission at no extra cost to you. Recommendations are based on suitability for the work, not on commission rates.

Frequently Asked Questions

Is stainless steel TIG welding DCEN or DCEP?

Stainless steel TIG welding is done with DCEN (direct current electrode negative), also called straight polarity, with the torch on the negative terminal and the ground clamp on positive. This puts most of the arc heat into the workpiece, giving deeper penetration and a narrow bead while keeping the tungsten cool. DCEP is used only in rare specialty situations and will quickly melt a TIG electrode.

Can you TIG weld stainless steel on AC?

Yes, but it is rarely the best choice. AC adds an electrode-positive cleaning half-cycle that stainless does not really need, which widens the etch zone, increases tungsten wear, and makes heat control harder. AC is occasionally useful for reducing arc blow on long joints or adding light cleaning on contaminated material, but DCEN with proper cleaning and pulsing solves most stainless problems better.

Why does my tungsten ball up when TIG welding stainless?

The most common cause is reversed leads, putting the torch on positive (DCEP) so the tungsten absorbs most of the heat. Other causes include welding in AC mode, using too small a tungsten for the amperage, or a contaminated tip. Check polarity and mode first, then confirm tungsten diameter against your current setting.

What tungsten should I use for stainless TIG at DCEN?

Lanthanated (gold), ceriated (gray/orange), or rare-earth blend tungstens are popular modern choices because they start well and handle DCEN current without balling. Two-percent thoriated (red) tungsten also performs well but requires care with grinding dust because of its radioactive oxide. Grind a longitudinal taper about two to two and a half times the electrode diameter.

Does polarity affect heat tint on stainless welds?

Indirectly, yes. DCEP and cleaning-heavy AC settings spread more heat and produce a wider oxidized zone, which shows up as blue, purple, or gray tint instead of light straw. DCEN lets you travel faster at lower total heat input, which usually means less tint. Shielding gas coverage and post-flow also strongly affect tint.

What polarity is used for TIG welding aluminum versus stainless?

Aluminum is normally welded with AC because the electrode-positive half-cycle breaks up the thick aluminum oxide layer, while the electrode-negative half provides penetration. Stainless is normally welded with DCEN because its chromium oxide layer is thin and does not require cathodic cleaning. This difference is the single biggest reason welders get confused when switching between the two metals.

How do I know if my TIG leads are connected backwards?

Symptoms include a tungsten tip that balls within seconds at low amperage, a wide shallow bead, a harsh or unstable arc, and rapid electrode erosion. Check the machine’s terminal markings and owner’s manual, since multi-process units and some imported machines label terminals differently. When in doubt, test on a scrap coupon before starting the real joint.

Is stainless TIG welding fume dangerous even at DCEN?

Yes. Stainless welding fume contains hexavalent chromium, Cr(VI), a known carcinogen, and OSHA limits exposure to 5 micrograms per cubic meter as an 8-hour time-weighted average. Polarity does not change that hazard; local exhaust ventilation, good general ventilation, and respiratory protection where required are the controls that matter.