Bottom line first: when line width drifts out of control on heavy-copper boards, the cause is usually not exposure or development — it is the etch factor collapsing as copper thickness rises while the compensation coefficient still reflects 1 OZ. Separating top width from bottom width, and building compensation tables by copper thickness and pattern density, is the precondition for getting heavy copper right the first time.
Side-etch is not a defect — it is what isotropic etching does
Cupric chloride etching is isotropic: as the chemistry attacks downward it also attacks sideways. The etch factor (EF) describes that anisotropy: EF = 2 × copper thickness ÷ (top width − bottom width). Higher EF means a more vertical sidewall; lower EF means a trapezoidal section and, in severe cases, a mushroom profile with overhanging copper on top and a pinched foot below.
Under a fixed recipe, EF moves inversely with copper thickness. At 1 OZ (35 μm) a shop typically holds 3.0–4.0, giving 4–6 μm of side-etch per edge; at 2 OZ (70 μm) EF falls to 2.0–2.5; at 3 OZ (105 μm) often only 1.5–2.0. At 13 OZ, subtractive etching can no longer hold fine line width: you need segmented etching (SEP) or multiple imaging passes, and the minimum achievable line width moves up.
A frequent misread: AOI reports top width as conforming while the cross-section shows bottom width short. Top width is set by development and exposure energy, bottom width by the etch. On thin copper the gap is negligible; on heavy copper it reaches 20–40 μm — enough to shift current capacity by a class.
Build the compensation table by copper thickness, not by board
- Design should specify bottom width as nominal, because that is the electrical dimension; manufacturing back-calculates dry-film width as top width = bottom width + 2 × side-etch + 8 μm margin.
- Per-edge compensation in practice: 8–12 μm at 1 OZ, 18–24 μm at 2 OZ, 28–36 μm at 3 OZ. At 13 OZ it must be measured against etch-stage count and conveyor speed; linear extrapolation fails.
- Index the table on copper thickness, line-width band and pattern density, covering the sub-3 mil, 3–6 mil and above-6 mil bands, calibrated per etch line.
- Re-calibrate quarterly with a step coupon; bath ageing and spray-pressure drift move EF by more than 10%.
Pattern density: the second variable in the coefficient
Etch rate depends on local copper area: an isolated line refreshes the chemistry quickly and side-etches more, while a dense area depletes the bath, etches slower and comes out wider. A 10–15 μm difference between the two on the same panel is normal, not an anomaly.
Control works on three levels: dummy copper in isolated areas to reach 40%–60% local coverage; oscillating spray with staged pressure, front-to-back differential below 0.5 bar; and a bath held at 130–160 g/L copper ion and 48–52 °C by automatic dosing. Compensate edge-to-centre variation with conveyor speed, never with concentration — raising it widens side-etch across the whole line.
Bottom width is the electrical dimension: two separate accounts
Impedance is computed from bottom width plus sidewall contribution. On a 50 Ω microstrip at 5 mil line width, a 10 μm change in bottom width shifts impedance by roughly 1.5–2 Ω, or 3%–4%: inside ±8% on its own, but stacking dielectric thickness tolerance and Dk lot variation pushes it past the limit. Check the coupon against a cross-sectioned bottom width, not the design value.
Current capacity follows the cross-sectional area at bottom width. On 13 OZ at 100 A-class current, losing 20 μm removes nearly 20% of the area and pushes temperature rise up a step. Enter the measured bottom width in any IPC-2152 calculation; nominal values are systematically optimistic on heavy copper.
Acceptance: cross-section, AOI, and coupon cannot substitute for each other
- Cross-section the first article and one panel per lot, recording top width, bottom width, sidewall angle (target 70°–85°) and copper thickness on one sheet.
- AOI judges top width, opens and shorts only; it cannot see bottom width. A clean AOI result is not evidence of a conforming etch.
- Distribute impedance coupons across layers — one set isolated, one dense — and attribute TDR deviations against cross-sectioned bottom widths.
- Hold line-width accuracy at ±8 μm and within-panel uniformity at ±10%; both must be met before the process is in control.
Line-width accuracy on heavy copper is not something the etch step can guarantee alone: nominal dimension, compensation logic, pattern-density distribution and cross-section feedback form one closed loop, and removing any link lets parameters drift by the next lot. AIPCB.cc supports up to 44 layers, a maximum panel size of 610×1100 mm, heavy copper up to 13 OZ, line-width accuracy of ±8 μm, impedance tolerance of ±8% (±10% standard), and aspect ratio up to 20:1 (25:1 for backplanes), with 24-hour rush prototypes. For engineering and quotations, contact info@ipcba.com.cn or +86 138 2310 9230.
Send your Gerber files and stackup requirements; a 24-hour expedite channel is available.