Robotic Soldering: Programming Tip Temperature, Dwell and Wire Feed
Robotic soldering sits between hand assembly and wave or selective soldering: a programmed iron, a motorised wire feeder and a positioner make the same joint the same way on every board. It is chosen when a product has a small number of through-hole or connector joints that a wave cannot reach, when tall parts or heat-sensitive neighbours rule out mass soldering, and when the volume is steady enough that hand work starts to drift. Getting it right means controlling iron tip temperature, dwell time, solder wire feed and the mechanical repeatability of the machine, then proving all four on the first article.
Where Robotic Soldering Fits in Assembly
Robotic soldering is normally reserved for joints that a wave cannot reach, such as a connector at the board edge whose pins face inwards, a part whose body cannot survive a wave, or a small number of through-hole components on an otherwise surface-mount board. The machine brings repeatability of position, time and solder volume that hand soldering cannot match, but it does not remove the need to define the joint in numbers.
The decision usually comes down to joint count and geometry. Arrays of connectors at a repeatable pitch benefit most, because every pin receives the same treatment whether it is the first or the last on the board. Panels with two or three isolated joints and a wide mix of part types are often faster to hand solder, because programming and fixturing cost more time than they save.
Iron Tip Temperature and Thermal Recovery
The set point on the controller is not the temperature at the joint. A tip that is too small for the pad loses heat faster than the heater can replace it, so the joint sees a lower temperature, and the programmer compensates by extending the dwell, which spreads heat into the laminate and the component instead. An iron tip temperature between 320 C and 370 C covers most lead-free work, with the lower end preferred where the component is heat sensitive.
Thermal recovery is the property that actually matters. A tip with good recovery returns to set point within a second or two of touching a heavy joint; a poor one stays low through the whole dwell. Measuring the tip with a contact thermocouple at the start of a shift shows whether the controller reading is honest, because worn tip plating, a loose tip and an oxidised surface all shift the true temperature away from the displayed value.
Dwell Time and Joint Thermal Mass
Dwell time is the period during which the tip stays in contact with the joint. It has to be long enough for the solder to wet both the pin and the pad and for the fillet to form, and short enough that the flux is not consumed before wetting is complete. Ground planes and thick copper raise thermal mass sharply, so a pin tied to an internal plane may need two or three times the dwell of a pin on an isolated pad.
The practical window is narrow. Below about one second the fillet may be incomplete and the solder may sit as a blob rather than a curve; above five seconds the laminate around the pad begins to discolour and the risk of a lifted pad or a cracked barrel rises. For a given product the range is established on the first article and then fixed, rather than adjusted board by board by whoever is running the machine.
Solder Wire Feed Rate and Volume
The feeder decides how much solder arrives at the joint. Solder wire feed is programmed as a length per joint, and that length is derived from the volume needed to fill the barrel and leave a concave fillet, not from the appearance of the first attempt. A 0.8 mm wire feeding at 1.5 mm per second delivers roughly the same volume as a 1.0 mm wire at 1.0 mm per second, so a change of wire diameter requires the program to be recalculated.
The angle at which the wire approaches the joint matters as much as the volume. Feeding into the tip so that the wire meets hot iron before it reaches the joint gives a repeatable transfer; feeding onto a cold pad leaves unmelted wire and a rough, cold-looking joint. The delay between tip contact and the start of the feed belongs in the program, because starting too early pushes unmelted wire against the pin.
Tip Geometry and Joint Access
Tip shape decides how much of the joint is heated at once. A chisel transfers heat through a broad face and suits pins, tabs and small terminals; a conical tip concentrates heat on a small area and is easier to place when neighbouring parts are close; a hoof or drag tip carries a reservoir of solder and can finish a row of pins in one pass. For connector rows at 2.54 mm pitch a chisel between 1.6 mm and 2.4 mm is typical.
Access is often the limiting factor rather than thermal performance. A tall electrolytic capacitor beside a connector can block the approach angle, forcing the use of a longer, thinner tip that recovers heat poorly. Where access is genuinely tight, the answer belongs in the layout review rather than in the soldering program, because no tip shape recovers performance that the geometry has taken away.
Flux and Atmosphere Around the Joint
Robotic soldering normally uses cored wire, so the flux is delivered with the solder. No-clean cored wire leaves a small residue that is acceptable on most assemblies, while water-soluble cored wire wets faster and needs a flux residue cleaning step afterwards. Flux activity has to match the surface finish, since a heavily oxidised finish needs more activity than a freshly plated one, and a mismatch shows up as poor wetting at one corner of the board.
Adding nitrogen around the joint reduces oxidation and widens the process window, particularly for lead-free alloys and for joints with long dwells on heavy copper. It does not repair a tip that is too small or a temperature that is too low, and the flow has to be aimed at the joint rather than flooded over the whole assembly to give any benefit.
Position Repeatability and Fixturing
The machine repeats the coordinates it was taught, so accuracy depends on the fixture holding the board in the same place for every cycle. A panel located by two tooling holes and a spring-loaded pin repeats better than one pushed against a machined stop, because the stop wears and the wear is invisible until joints start to miss. Fixture wear of 0.1 mm moves every joint by that amount, which is enough to miss a pad on a fine-pitch connector.
Teaching should be done from CAD data where the machine supports it, with a camera correcting for the actual position of each board. Where vision is not available, the first board of each lot verifies that the taught coordinates still land on the pads, and a shift is traced back to the fixture rather than corrected pin by pin in the program.
Programming the Joint Sequence
The order in which joints are soldered affects the temperature of the whole board. Soldering a ground pin early makes the following joints easier, because the surrounding copper is already warm; soldering every pin of one connector consecutively keeps the tip in place but heats the connector body, which is a risk for plastic housings. A practical compromise alternates between connectors and returns to the heavy joints later in the cycle.
Every step should be recorded as part of the soldering robot programming: approach height, contact time, feed length, retract height and the delay between them. When a joint fails in the field, that record is what makes the cause visible, because the machine will have repeated exactly what it was told to do, cycle after cycle.
First Article and In-Process Verification
The first article is where the program is confirmed. Visual checks for fillet shape and wetting angle, a pull test on a representative joint, and a cross-section of one joint of each type give more information than any amount of inspection at the end of the line. Where a joint cannot be sectioned without destroying the assembly, a coupon built with the same components can be sectioned in its place.
Once the line is running, the process is checked by measuring tip temperature at the start of each shift, by inspecting the first and last board of every lot, and by recording the program revision in use. A tip change, a spool change or an alloy change is a process change and should be treated as one, with the first board inspected and released before the remainder of the lot.
FAQ
Which iron tip temperature suits lead-free robotic soldering? Most lead-free programs run between 320 C and 370 C, with the lower end used where the component or the laminate is heat sensitive. The set point matters less than the measured tip temperature under load, so a contact thermocouple check at the start of a shift is more useful than the controller display.
How is dwell time decided? From the thermal mass of the joint. A pin connected to a ground plane needs a longer dwell than a pin on an isolated pad, so heavy and light joints are grouped separately in the program. The window is confirmed on the first article rather than guessed.
Is nitrogen necessary for robotic soldering? No, but it widens the window on heavy joints and on oxidised surface finishes. Where it is used, the flow should be directed at the joint, since flooding the whole assembly wastes gas and cools the board unevenly.
