Lead Frame: Etching vs Stamping

Pick the wrong process for a lead frame and the damage shows up three ways at once. The tolerance budget burns out before you reach the inner leads, so units fail solderability and registration at the bond pad. The tooling invoice lands before the first production order is signed, and the per-part cost never recovers because the volume was never there to amortize a progressive die. The thickness you specced - 0.15 mm strip, half-hard C19400 - turns out to sit exactly where stamping fights work hardening and etching shrugs it off. Thirty years of quoting these parts tells me the etching-versus-stamping question is not a preference. It is a numbers problem, and the numbers are predictable if you know where to look.

How the Two Processes Actually Remove Material

Photochemical etching, sometimes called chemical milling, starts from a flat strip of copper alloy - typically 0.1 mm to 0.3 mm thick for lead frames. A photoresist is laminated to both faces, imaged with your CAD artwork, and developed. The exposed metal goes into a ferric chloride or cupric chloride spray etchant that dissolves it isotropically from both sides. What remains is your lead frame pattern. There is no mechanical force on the part. The strip does not see a punch, a die, or a press tonnage rating.

Stamping runs the same strip through a progressive die on a high-speed press. Each station cuts, forms, or bends a feature. By the last station the leads are blanked free of the carrier. The material is sheared by compressive and tensile stress at the cutting edge, and that shearing is what creates the burr, the work-hardened shear band, and the tool-wear drift that limits long-run accuracy.

The difference that matters most is not speed or cost. It is where the material is stressed. Etching removes metal atom by atom with no deformation, so the crystal structure of the strip arrives at your assembly line exactly as the mill delivered it. Stamping deforms the shear zone on every cut, bends the material at every form station, and leaves residual stress that only heat treatment can remove. For a part whose leads must stay flat and springy to 0.01 mm, that distinction is the whole argument.

Tolerance: Where the ±0.01 mm and ±0.02 mm Numbers Come From

The often-quoted figures hold up under production conditions. Photochemical etching holds a dimensional tolerance of roughly ±0.01 mm (±10 µm) on features at these thicknesses, because the etch front is controlled by the phototool and the etch factor, not by a worn cutting edge. Stamping holds about ±0.02 mm (±20 µm) on blanked features in the same thickness range - and that number assumes a fresh, well-maintained die. As the punch and die clearance opens up through wear, a stamping run drifts toward the loose end of that band, while an etching run stays flat because there is no contact tool to wear.

For lead pitch and lead width the gap matters more than the absolute number. Etching resolves lead widths down to about 0.1 mm and pitches near 0.2 mm without special effort. Pushing a stamped lead below roughly 0.2 mm width at 0.15 mm thickness starts a fight with burr height and lead straightness that eats your yield.

Parameter	Photochemical Etching	Precision Stamping
Dimensional tolerance	±0.01 mm	±0.02 mm
Material thickness range	0.02-0.5 mm (0.1-0.3 mm typical for LF)	0.1-0.3 mm practical; up to several mm
Min lead width / pitch	~0.1 mm / ~0.2 mm	~0.2 mm / ~0.4 mm
Inner corner radius	~0.05 mm (near sharp)	≥0.3x thickness (≈0.05-0.1 mm)
Burrs	None	Present, needs deburring
Tooling cost	Phototool only (hundreds of USD)	Progressive die ($5k-$50k+)
Break-even volume	Economical to ~50k pcs	Wins above ~50k-100k pcs

One tolerance subtlety buyers miss: etching's ±0.01 mm is a feature-to-feature number within one frame, and it holds on both the front and back of the strip because both faces are etched from the same artwork. Stamping's ±0.02 mm applies to the blanked silhouette, but the formed features - downsets, standoffs, bends - carry their own separate tolerance that stacks on top. If your drawing shows a ±0.01 mm pitch on a frame that also has a 0.2 mm downset, the downset tolerance is part of the stack, and stamping has to hold both or the part fails coplanarity at the bonder.

Thickness Envelope: 0.1 to 0.3 mm Is the Sweet Spot for Both, but for Different Reasons

Lead frames in semiconductor and LED packaging live in the 0.1 mm to 0.3 mm thickness band. Etching handles this entire range with no change to the process - the etch time simply scales. A 0.1 mm foil and a 0.3 mm strip run the same line. Stamping also covers 0.1-0.3 mm, but below 0.1 mm the strip gets hard to feed and flatness control at the press becomes the bottleneck, and above roughly 0.3 mm the lead frame has usually outgrown the fine-pitch role where etching wins and moved into a structural bracket where stamping's speed pays.

One thing stamping cannot do that etching does trivially: hold a constant thickness across a pattern with mixed feature densities. In etching, the untouched lands stay at full strip thickness; in stamping, every bend and relief is a separate die operation with its own springback variable.

There is a second thickness consideration that rarely makes it into the comparison table: what happens at the edges. Etching removes material from both faces, so a through-etched lead has a slightly tapered cross-section - the etchant attacks sideways as it goes down, and the sidewall angle lands around 70-85 degrees depending on etch factor and thickness. That taper is uniform and predictable, and for solderability and plating adhesion it is actually favorable, because it removes the sharp vertical shear zone. Stamping produces a straight shear wall with a burr on one side, which is exactly the geometry that traps plating chemistry and creates stress risers at the bond. If your lead frame sees wire bonding or soldering on the edges, put the edge cross-section on the drawing, not just the top view.

Volume and Economics: The Break-Even Is a Real Number, Not a Feeling

Etching carries almost no upfront cost. Your only tool is a photographic artwork - a few hundred dollars, revised in a day. That makes it the default for NPI, qualification builds, and any run under roughly 10,000 pieces. It also makes it the safe choice when the design is still moving, because changing the pattern is a new phototool, not a new die.

Stamping front-loads cost into the progressive die, typically $5,000 to $50,000 depending on stations, material, and precision. That spend only pays back when volume is there. The crossover - where stamping's lower per-part cost overtakes etching's zero tooling - usually sits between 50,000 and 100,000 pieces for a typical fine-pitch frame. Below that line, etching wins on total cost of ownership. Above it, stamping wins on unit price. Quote both before you commit; I have seen buyers lock a die at 20k units and never reach break-even.

Cost component	Photochemical Etching	Precision Stamping
Tooling (one-time)	Phototool, hundreds of USD	Progressive die, $5k-$50k+
Tooling revision	New phototool in days	Die modification, weeks
Per-part cost at 10k pcs	Low, no amortization burden	High, die not yet paid back
Per-part cost at 200k pcs	Higher per unit, plateaus	Low, die fully amortized
Deburring step	None	Tumble/brush/electrolytic
Setup time per run	Minutes (phototool change)	Hours (die setup, feed, pilots)

The break-even math has a second layer that people forget: the cost of a design revision. A lead frame that changes after tooling is built costs a die modification plus downtime plus re-qualification. The same change on an etching program is a phototool redraw and a new etch run. If your program has more than one revision planned - and most semiconductor programs do - fold the expected revision cost into the comparison, not just the tooling invoice. For a frame with a real chance of a design change inside 12 months, etching frequently wins the total-cost argument even past the nominal break-even volume.

Material Behavior: CuFe2P and C19400 Are the Same Animal, Specced Two Ways

Most lead frames are a copper-iron-phosphorus alloy. In Europe you will see CuFe2P (EN CW107C); in North American sourcing the same grade lands as C19400 under the UNS system. Both sit around 2.1-2.6% iron and 0.015-0.15% phosphorus, giving the high strength and thermal conductivity the package needs. Other common grades are C7025 (Cu-Ni-Si) for higher strength and C15100 (Cu-Zr) where creep matters.

Here the two processes diverge in how they treat the metal. Etching is non-mechanical, so it leaves the grain and temper exactly as received - no work hardening from the process itself, which protects the spring properties of the leads. Stamping work-hardens the shear zone and bends; you must design the temper and the die sequence around that, and you will anneal or stress-relieve after forming on the harder tempers. For C19400 in half-hard, stamping is routine. For C7025 in high-strength temper, the springback math gets serious and etching's neutrality starts to look attractive.

Alloy	System	Key properties	Typical lead frame use	Process note
CuFe2P / C19400	EN CW107C / UNS	High strength, good conductivity	Standard IC and LED frames	Half-hard stamps routinely; etches without temper loss
C7025	Cu-Ni-Si	Higher strength, good relaxation resistance	High pin-count, fine-pitch frames	Springback-sensitive in stamping; etching preserves temper
C15100	Cu-Zr	High conductivity, creep resistance	Power and high-current frames	Soft; stamping needs careful tooling, etching is gentle
C19210 / C19200	Cu-Fe-P variants	Conductivity-focused grades	Battery and power lead frames	Similar behavior to C19400

The temper question also changes the plating story. A stamped lead frame that gets work-hardened at the shear zone and then stress-relieved has a different surface energy and a different oxide state than an etched frame that was never touched. Etched copper surfaces plate more uniformly, which is one reason etching programs tend to see fewer plating adhesion rejects in the solderability test. It is not magic; it is simply that the surface has no smeared shear layer to hide under the deposit.

Corner Radius and Geometry: Sharp Is Etching's Home Turf

Chemical etching removes material from all exposed surfaces at once, so inner corners stay as sharp as the artwork resolves them - practically down to about 0.05 mm radius, sometimes less. There is no cutting edge that needs relief. Stamping punches cannot pull an infinitely sharp inner corner out of sheet without cracking the material or the tool. The rule of thumb is a minimum inner corner radius of roughly 0.3 times the material thickness. At 0.15 mm strip that is about 0.05 mm, but in practice you specify 0.1 mm or more to keep the die alive and the part crack-free. Any design with tight internal radii, fine slots, or asymmetric cutouts leans hard toward etching.

Geometry freedom is not just about corners. Etching can produce features that a progressive die cannot reach at all: internal slots narrower than the strip is thick, islands fully surrounded by removed metal, varying feature density across one frame, and patterns that change thickness locally through partial etching (half-etch). Stamping's geometry is bounded by what a punch can enter and strip; a die can cut a slot, but the slot width is limited by punch strength, and a fully enclosed island needs a separate stage with its own slug control. If your frame has any of these features, the process decision is already made for you.

Surface Quality and Burrs: The Hidden Yield Killer

Etching produces a clean, burr-free edge - the etchant attacks uniformly and leaves no shear lip. That single fact removes a whole process step. Stamping leaves a burr on every cut edge; height scales with clearance and wear and can run 5-10% of thickness. At 0.15 mm that is a 0.01-0.015 mm burr standing on a lead that must register to a 0.2 mm pitch. You deburr - tumble, brush, or electrolytic - and every deburring step is cost, handling, and a new source of variation. For plating adhesion and wire-bond reliability, the burr-free etched surface is the lower-risk path.

Burr direction matters even when burr height is controlled. A stamped part has a punch side and a die side, and the burr sits on one of them; if your assembly stacks two stamped leads face to face, the burrs can interlock or abrade plating. Etched parts have no such asymmetry, which simplifies assembly tooling and stacking. On high-frequency or high-voltage parts, burrs and sharp shear zones also concentrate electric fields and can become partial-discharge sites - a failure mode that no dimensional inspection will catch, because the part measures fine and the discharge only appears under stress testing.

Plating and Finishing: What Changes Between the Processes

Lead frames are plated after patterning, usually selectively - gold, silver, or tin on the contact and bond zones only, in the 2-8 µm range typical of reel-to-reel selective plating. The process choice changes how plating behaves in three ways.

First, adhesion. Plating bonds to a clean, uniform surface, and the etched edge profile (tapered, no shear smear) gives the deposit a mechanical key that a straight shear wall lacks. Stamped edges need a clean step - deburring plus a chemical clean - before plating, or the deposit lifts at the edge in the solderability test.

Second, masking and registration. Both processes use the same selective plating masks, but etched parts have no burr to shadow the mask edge, so the plated band lands where the artwork says it lands. On stamped parts, a burr at the band edge can wick plating past the mask or shadow a few microns of the band - small, but exactly the kind of variation that shows up as a void in cross-section.

Third, flatness through the plating line. Plating is a wet, heated process with strip handling at tension; a stamped frame with residual forming stress may bow when the stress relaxes in the plating bath. Etched frames, having no forming stress, hold flatness through the line. If coplanarity is a printed requirement on your drawing - and it usually is for fine-pitch frames - the process that keeps residual stress out of the part protects your coplanarity spec through every downstream wet step. For more on specifying the plated zones and thickness, the terminal plating guide covers the selection logic.

Quality Control and Inspection: How to Verify a Lead Frame

Whichever process you pick, the inspection plan decides whether the part performs. The checks below are the ones that catch lead frame failures before the bonder does:

• Coplanarity. All leads must sit in the same plane - typically within 0.05-0.1 mm for fine-pitch frames. Measure on a granite plate with a height gauge or CMM; this catches downset and forming errors that top-view dimensions miss.
• Pitch and lead width. Optical measurement across the full frame width, not lead-to-lead. Pitch error accumulates across the frame, so measure frame-wide and compute Cpk on the worst dimension.
• Burr height. Comparator or optical profilometer, sampled at defined intervals. Stamped frames need this as an SPC-charted dimension; etched frames can skip it.
• Solderability. Wetting balance or dip-and-look test per lot. This is the test that catches plating contamination, and it must be done on the finished plated part, not the bare strip.
• Plating thickness. X-ray fluorescence at the contact and bond zones, plus a cross-section of the band edge at lot changes.
• Material verification. Mill certificate check on grade, temper, and thickness at incoming; a stamped frame that arrives with the wrong temper will fail bend and springback in the field.

A capable supplier runs these on a documented sampling plan with first-article reports that match your drawing datums. The quality lab at a full-service stamper - CMM, optical measurement, and inline vision - should be able to reproduce your own inspection results, because when the customer's fixture and the supplier's fixture disagree, the part is fine and the spec is broken.

Lead Time and Design Iterations: The NPI Math

Time to first parts is often the real decision driver, and it favors etching hard. A phototool is artwork: your CAD file becomes a working phototool in days, and first etched parts follow within a week or two of approved artwork. A progressive die for a fine-pitch lead frame is a machine built to tenths: design, wire EDM, grinding, assembly, tryout, and adjustment typically take 4 to 8 weeks, and the first article is not guaranteed to pass on the first tryout.

Design iteration is where the gap compounds. Every revision to a stamped frame means die modification - new punches, new inserts, new tryout runs, and downtime on the press schedule. Every revision to an etched frame means a new phototool and a rerun. For a program in active development, etching lets you run qualification builds while the design is still moving, then convert to stamping once the design freezes and volume proves out. That staged strategy is common in LED and discrete semiconductor programs, and it is the cheapest way to use both processes: etch through the noise, stamp the winner.

Hybrid Approaches: When One Process Is Not Enough

The question is not always either/or. Three hybrid patterns show up in production lead frames:

• Etch the fine features, stamp the carrier. The fine-pitch inner leads are etched into a strip, and a stamped carrier frame provides the outer structure, downsets, and handling rails. You get etching's resolution where it matters and stamping's speed and strength where it does not.
• Half-etch plus form. Etching removes material selectively from one face to create thin flex regions or coined relief, then the frame goes through a light forming pass. This is how some clip and lead frame designs get local flexibility that neither process alone produces cleanly.
• Etched prototype to stamped production. Qualify on etched parts, then transfer the frozen design to a progressive die. The etched qualification parts match the final geometry within the tolerance gap, and the stamped production parts inherit the same artwork-derived datum.

Hybrid parts cost more to plan but often beat either single process on total cost of ownership. If your frame has 0.1 mm inner leads and a structural outer frame, a pure stamping quote will fight the fine features and a pure etching quote will pay for resolution it does not need. Ask your supplier to quote the hybrid before you default to one process.

Which One Do You Pick: The Decision Checklist

Run the decision on five inputs:

• Annual volume against the 50k-100k break-even. Below it, etching; above it, stamping - unless the design is still moving.
• Feature geometry against the corner-radius and slot limits. Sharp inner corners, fine slots, half-etch, or enclosed islands - etching.
• Thickness and temper of your CuFe2P/C19400 stock. Hard tempers and thin strip - etching protects the temper; routine half-hard 0.15-0.3 mm - stamping is viable.
• Design stability. More than one expected revision in 12 months - etching or a staged etch-to-stamp plan.
• Surface and burr requirements. Burr-free edges, plating adhesion, or stress-test cleanliness - etching.

The tolerance gap (±0.01 vs ±0.02 mm) and the burr question usually decide the borderline cases. Low volume, moving design, fine pitch, sharp internal corners, or a temper that fights springback - etching. High stable volume, simple geometry, and a need for the lowest possible unit cost - stamping. And when the answer is genuinely close, the staged hybrid gets you both.

Two related reads before you finalize: the lead frame stamping guide covers tolerances and alloys for the stamping route in depth, and the lead frame surface mount guide covers the package-level requirements that drive both processes.

Need a Process Call on Your Lead Frame

If you are sitting on a lead frame drawing and not sure which way to send it, send us the Gerber or DXF, the grade and temper, and your expected annual volume. We will run both a photochemical etching and a stamping feasibility on it, give you the real per-part numbers at your quantity, and tell you straight which process protects your yield and your margin. That is a thirty-year quote, not a sales script.