What does wire feeding rate actually control in laser soldering? Wire feeding rate sets how much alloy — and with flux-cored wire, how much flux — arrives at the joint per unit time, so it directly controls the balance b…
What does wire feeding rate actually control in laser soldering?
Wire feeding rate sets how much alloy — and with flux-cored wire, how much flux — arrives at the joint per unit time, so it directly controls the balance between energy delivered by the laser and energy consumed by melting and heating the incoming solder. It is not an independent parameter: the same feed rate produces a sound fillet at 60 W and a cold, balled joint at 35 W, or a bridged joint at 150 ms dwell and a properly wetted one at 300 ms. Experienced process engineers treat feed rate, laser power, irradiation time and wire entry position as a single four-variable set, and they tune feed rate last, after the thermal profile of the pad and lead has been established.
Why the energy balance matters more than the “amount of solder”
The energy needed to melt solder alloy is surprisingly small compared with the energy needed to heat the pad, lead and surrounding laminate to wetting temperature. For a lead-free alloy such as SAC305, roughly 0.5 J is enough to take about 4 mg of wire from room temperature to 250 °C, including latent heat of fusion. A 40 W absorbed beam delivers that in around 12 ms. What actually limits the process is the interface: the copper pad and component termination act as a heat sink that is one to two orders of magnitude larger than the wire volume.
This is why an over-fast feed rate fails in a specific way. The extra alloy does not simply produce a bigger fillet; it arrives cold, absorbs laser energy at the melt pool surface, and effectively shields the wetting interface from the beam. The result is a joint that looks adequately filled but has poor wetting, a dull grainy surface, and a thin brittle intermetallic layer in cross-section. Conversely, an under-fast feed starves the fillet: the beam heats the pad well, the alloy wets and spreads, but there is not enough volume to form a proper concave fillet, and the joint fails mechanically at low pull force rather than by wetting.
As a working number, a 0.5 mm wire feeding at 30 mm/s delivers about 5.9 mm³/s. A typical SMT pad joint needs 0.5–1.5 mm³, which corresponds to 85–250 ms of feeding. Use this kind of calculation to set a starting feed length per joint, then verify by cross-section rather than by eye.
Starved, balanced and flooded: three joint formation regimes
- Starved (feed 20–30% below optimum): sharp, concave but undersized fillets; visible pad area left unwetted; wetting angle often above 60°; strength limited by fillet cross-section. Often mistaken for “insufficient laser power” and wrongly corrected by raising power, which oxidises the pad and darkens the flux residue.
- Balanced (within roughly ±15% of optimum): smooth concave fillet, wetting angle below 30–45°, fillet height around 75% or more of the lead or termination height, bright fine-grained surface, no solder on the beam-facing side of the component body.
- Flooded (feed 25% and above optimum): convex or bulbous joints, solder balls and spatter around the joint, bridging between adjacent pads at pitches below 0.5 mm, wire stub remnants when the tail is not fully melted, and a higher void rate because flux volatiles are trapped under a large molten volume.
Note that flooding also degrades the wire feed itself: when the wire tip contacts an already-large melt pool, heat conducts back up the wire, softening it above the nozzle and causing buckling, curling or a “birdnest” at the feed tip. That is a mechanical symptom that shows up several boards before the joint quality drifts.
Matching feed rate to wire diameter, laser power and joint geometry
The values below are practical starting windows for diode-pumped fibre-delivered laser soldering with closed-loop temperature control; they assume flux-cored wire and a spot diameter 1.5–2× the pad width.
| Wire diameter | Typical feed rate | Laser power / spot | Irradiation time | Typical joint |
|---|---|---|---|---|
| 0.3 mm | 8–25 mm/s | 15–35 W, 0.6–1.0 mm | 120–350 ms | Fine-pitch SMT pads, 0402/0603, sensor terminals |
| 0.5 mm | 15–45 mm/s | 25–60 W, 0.8–1.5 mm | 150–500 ms | Connector pins, 0.5 mm pitch flex, optical module leads |
| 0.8 mm | 20–60 mm/s | 40–90 W, 1.2–2.5 mm | 250–800 ms | ECU through-hole pins, relay and terminal blocks |
| 1.0–1.2 mm | 25–70 mm/s | 60–150 W, 2.0–3.0 mm | 300–1200 ms | Power terminals, busbar and wiring harness joints |
Three selection rules apply across the table. First, thinner wire buys resolution and control but costs cycle time, because more feed length is needed for the same deposited volume — halving the diameter quadruples the length required. Second, as joint thermal mass rises, increase laser power and dwell first, and only then raise feed rate, otherwise the additional alloy cools the interface. Third, when wire diameter changes, re-derive feed length from the fillet volume rather than scaling rate linearly, since volume scales with the square of diameter.
Feed position and timing matter as much as feed rate
A correct feed rate delivered to the wrong place still produces defects. The wire should enter at the trailing edge of the melt pool, at 30–60° to the joint plane, offset from the beam centre by roughly 0.3–0.8× the spot radius, so that the beam heats the pad and the wire is melted by conduction from the pool rather than by direct irradiation. Direct irradiation of the wire tip causes balling: the tip melts into a sphere before it touches the pad, and surface tension then holds the ball away from the joint.
Feed timing is a second lever. Pre-feeding 50–200 ms before the laser fires places cold alloy on the pad and delays wetting; feeding synchronously with the laser, with a short post-feed of 20–60 ms to let the tail detach cleanly, is generally the more stable sequence. A retract of 0.5–2 mm after each joint prevents the softened wire from sticking to the solidified fillet and dragging it. Some systems use a closed-loop feed motor with encoder feedback plus a wire-end sensor, which is worth specifying if you run 0.3–0.5 mm wire at high duty cycle.
Failure symptoms and troubleshooting steps
| Symptom | Likely feed-related cause | Action |
|---|---|---|
| Solder ball beside joint | Wire tip irradiated before contact; feed rate too fast for the dwell | Shift wire entry to trailing edge; reduce rate 10–15%; add 20 ms post-feed |
| Bridging at fine pitch | Excess deposited volume; over-feed plus over-long dwell | Reduce feed length per joint first, then dwell; check wire straightness |
| Dull, grainy, poorly wetted joint | Over-fast feed chilling the interface | Reduce rate 15–20% or raise power 10%; verify with cross-section, not visually |
| Wire stub left in fillet | Feed continues after beam off; retract missing | Advance feed stop by 30–50 ms relative to laser off; enable retract |
| Inconsistent volume board to board | Feeder roller slip, nozzle wear, wire kink | Command 100 × 3 mm feeds, measure total length; accept ±3%; replace rollers/nozzle |
| Buckling or curling at tip | Heat conducted back up wire; feed too slow relative to power | Raise feed rate or shorten dwell; add wire cooling or shorten nozzle-to-joint distance |
Setup and acceptance checks
Run a one-factor-at-a-time sweep on feed rate with power, dwell and spot fixed, stepping in 5% increments across ±25% of the nominal value, and evaluate three outputs: fillet height relative to lead height, wetting angle, and pull or shear force on a sample of at least 30 joints. Cross-section 5–10 joints per setup to check intermetallic thickness and voiding; for automotive work, X-ray void acceptance is typically set below 15–25% of joint area depending on the customer standard. Then verify feed repeatability mechanically, because a 5% feed error is invisible on the bench but shows up as a fillet height trend over a production shift. Finally, confirm that the AOI recipe is tuned to the settled feed rate — a recipe built around flooded joints will pass starved ones.
FAQ
Does a faster wire feed rate shorten the cycle time?
Only marginally, and usually at the cost of quality. Cycle time in laser soldering is dominated by pad heat-up and dwell, not by wire delivery. Cutting feed time by 50% typically saves 50–100 ms while pushing the joint into the flooded regime.
Can I fix a cold joint by increasing the wire feed rate?
No — that is the classic wrong correction. More cold alloy removes heat from the interface and makes wetting worse. Increase power or dwell, or reduce feed rate, then re-verify by cross-section.
What wire diameter should I start with?
0.5 mm covers most connector and SMT work with a comfortable process window. Go to 0.3 mm for pitches below 0.5 mm or very small thermal mass, and to 0.8 mm or above for through-hole pins and terminals where deposited volume per joint is the limiting factor.
Why do the first few boards of a shift show different joints?
Cold machine and fixture, plus a partially straightened wire path, change the effective feed length. Purge 5–10 joints into scrap, verify the 100-feed length check, and only then release production.
How do I know feed rate is the cause and not laser power?
Set feed rate to nominal and vary power ±10%: if joint appearance tracks power smoothly, the issue is thermal. If defects appear abruptly at a specific feed value, the feeder or wire path is the cause.
If you are calibrating a feed window for a specific pad geometry or alloy, share the joint dimensions and wire specification with our process team and we can propose starting parameters for your ULiLASER system.