Reference · soft soldering, brazing, fluxes

Soldering

A solder joint is not glue. The tin dissolves a fraction of the copper and forms an alloy at the interface — the joint is metallurgical, not mechanical. That explains why a cold joint never becomes strong however much solder you pile on, why flux is mandatory rather than helpful, and why the right temperature matters more than the right iron.

Joint looks wrong? The troubleshooting table is arranged by what it looks like, not by cause.

Version 1.0 Alloys 8 Sections 11 SEEN Units °C, wt-%

The melting point decides everything else

Tap an alloy · logarithmic scale, °C

The 450 °C boundary is a definition, not physics: below it the process is called soft soldering, above it brazing. In practice the difference is that soft solder melts under an iron while brazing needs a flame, and that a brazed joint is roughly ten times stronger. Neither melts the base metal — that is what separates soldering from welding.

01

First principles

Five things. The first explains why soldering works at all, the third why most joints fail.

1 · The joint is an alloy, not a glue

When molten tin meets clean copper a little copper dissolves and a thin layer forms of an intermetallic compound, Cu6Sn5. That layer is the actual joint. The solder on top is only carrier and fill.

Two consequences that explain most things: the joint must get hot enough for that layer to form — melting solder on the tip and dabbing it on is not enough — and more solder never makes a joint stronger. A large blob on a cold joint is weaker than a thin, bright one.

The intermetallic layer also keeps growing slowly at room temperature and faster when hot. A thick layer is brittle, which is one reason old joints crack.

2 · The flux has exactly one job

Copper oxidises immediately, and molten solder does not wet oxide. The flux dissolves the oxide layer the moment the heat arrives, and keeps the surface clean until the solder gets there. That is why soldering without flux is not difficult but impossible.

Solder wire has a flux core, but it rarely stretches beyond a single joint. Reworking, oxidised surfaces and all SMD work need extra flux. Too much flux is almost never a problem; too little is the commonest fault.

3 · Heat the work, not the solder

Place the tip so it touches both the lead and the pad, wait a second, then bring the solder wire to the joint — not against the tip. If the solder melts against the work, the work is hot enough. If it only melts against the tip you get a cold joint that looks finished but has no intermetallic layer.

4 · Thermal mass beats wattage

A ground plane, a heavy lug or a large heatsink pulls heat away faster than a small tip can deliver it. The answer is rarely a higher temperature — that only chars the flux before it has done its work — but a larger tip with more mass and a better contact area. A 3 mm chisel tip solders heavy joints better at 320 °C than a needle tip does at 400.

5 · Time at temperature is what does the damage

Short and hot damages less than lukewarm and long. Board laminate, components and pads take 250 °C for a few seconds but lift off at 200 °C over half a minute. Poking at a joint for a long time is a sign of too low a temperature or too small a tip, not of needing more patience.

02

Alloys

Eight, covering everything from melting in your hand to joining bicycle frames.

Sn63Pb37 Eutectic tin-lead · the old standard

183 °C
Melting point183 °C, eutectic
Iron temp300–330 °C
WettingExcellent
AppearanceBright
RoHSNo
WhiskersNo

Why it is still the best alloy to learn on

Eutectic means the alloy goes straight from solid to liquid at a single temperature, with no pasty range in between. That makes the joint insensitive to movement at the critical moment and gives the characteristic bright, concave surface that is easy to judge by eye. It wets better than anything lead-free, works 35 degrees cooler and is more forgiving in every respect.

Lead

Banned in consumer electronics under RoHS, but permitted for private use, repair and hobby work. The risk in soldering comes from hand contact and ingestion, not from the fume — lead does not vaporise at 300 °C. Wash your hands before eating, do not solder where children eat, and never sweep up solder debris dry. The smoke you see is burnt flux, and it should be ventilated away regardless.

EasiestEutecticBrightNot RoHS

SAC305 Sn96.5 Ag3.0 Cu0.5 · the lead-free industry standard

217–220 °C
Melting range217–220 °C
Iron temp340–370 °C
WettingGood
AppearanceDull, grainy
RoHSYes
PriceThe silver costs

The lead-free alloy that wets best, thanks to the silver. The commonest source of disappointment is the appearance: a correct SAC joint is dull and slightly grainy, not bright. Judging it by the same standard as leaded solder leads people to rework joints that were already good.

Lead-free standardDull finishHotter

Sn99.3 Cu0.7 Lead-free without silver · cheapest, and the one in plumbing

227 °C
Melting point227 °C, eutectic
Iron temp350–380 °C
WettingModerate
AppearanceDull
Potable waterApproved
PriceLow

The standard choice for copper pipe in potable water systems, where lead solder is banned. For electronics it is worse than SAC305: it wets more sluggishly and needs a higher temperature, which wears both tips and components. Cheap enough to be common anyway. It is also the alloy on the budget shelf — Biltema's solder wire is Sn/Cu0.7 with a rosin core and states the same melting point, 227 °C. Fine for connectors and heavier wiring; for circuit boards the silver-bearing spool next to it is worth the extra.

PlumbingPotable waterCheap

Bi58 Sn42 Low-temperature solder · ”ChipQuik” and similar

138 °C
Melting point138 °C, eutectic
Iron temp220–250 °C
BrittlenessHigh
MixingLowers Sn-Pb sharply
UseDesoldering, delicate parts
RoHSYes

The tool for getting things off

Mixed into an existing joint, Bi58Sn42 drops the melting point of the whole blend dramatically, and the joint stays liquid for tens of seconds. That is exactly what you want when a board with fifty pins has to come off in one piece. Apply flux generously, work the low-temperature solder along every pin, and lift.

Clean up afterwards

Bismuth solder left in the pads makes future joints brittle and low-melting. Wick it all out and lay down fresh ordinary solder before the component is replaced. It is a step that often gets skipped, and it produces faults that surface much later.

DesolderingLow tempBrittle

Field's metal Bi32.5 In51 Sn16.5 · melts in your hand

62 °C
Melting point62 °C
ToxicityLow, cadmium-free
PriceHigh, indium
StrengthVery low
UseCasting, demonstration
AlternativeRose's metal, 98 °C

Melts in hot water. Used for impressions, low-temperature casting, thermal fuses and as a demonstration. Cadmium-free, unlike the old Wood's metal, which should be avoided entirely. As a solder it is too weak for anything that carries load.

62 °CCastingCadmium-free

Sn95 Sb5 Tin-antimony · for heat and vibration

235–240 °C
Melting range235–240 °C
StrengthHigh for a soft solder
Creep resistanceGood
UseRefrigeration, plumbing, sheet metal
Potable waterApproved
RoHSYes

Considerably stronger and more creep-resistant than ordinary tin solder, and used where the joint is loaded or gets hot: refrigeration systems, heat exchangers, sheet metal work. Wets less well and needs more heat.

StrongHeat-resistantRefrigeration

Silver braze Ag45–56 with copper, zinc and tin · brazing alloy

618–745 °C
Working temp620–750 °C
Tensile strength350–450 MPa
Heat sourcePropane or MAPP
Gap0.05–0.15 mm
FluxesBorax- or fluoride-based
CadmiumAvoid old rods

Brazing gives a joint that is often stronger than the base metal, and works on steel, stainless, copper, brass and dissimilar combinations. Capillary action is the whole mechanism: the gap should be a few hundredths of a millimetre, because then the filler is drawn in by itself. A slack joint does not fill — the capillary action disappears and the filler runs out instead.

Cadmium in old silver brazing rods

Silver brazing rods made before roughly 1990 often contain cadmium, which vaporises at brazing temperature and gives cadmium fumes — acutely dangerous and with symptoms that come hours later. Unmarked old rods from a garage clearance should be discarded, not used.

StrongestCapillary actionNeeds a flame

Brass braze CuZn · bronze welding, gas brazing of steel

875–895 °C
Working temp880–920 °C
SubstrateSteel, cast iron
Heat sourceOxy-fuel, induction
Gap fillingGood
FluxesBorax
Zinc fumeVentilation required

The classic for bicycle frames, cast iron repairs and steel structures. Fills gaps better than silver braze and costs a fraction as much, at the price of a considerably higher working temperature.

SteelCheapZinc fume
03

Fluxes

Four families. The choice is not about how well they solder but about what they leave behind.

Flux families
TypeActivityResidueUse for
Rosin, ROWeak to mediumSticky, hard, generally harmlessElectronics. The classic, smells of pine.
No-clean, REL0WeakSmall, dry, meant to stayElectronics in production. Does not need washing.
Water-soluble, OAStrongCorrosive, must come offElectronics where washing is planned. Never otherwise.
Acid-based, INVery strongStrongly corrosiveSheet metal, pipe, mechanical work. Never on circuit boards.

What the marking on the packet means

IPC J-STD-004 codes fluxes in three parts: origin (RO rosin, RE modified resin, OR organic, IN inorganic), activity level (L low, M medium, H high) and halide content (0 or 1). REL0 is therefore modified resin, low activity, halide-free — typically no-clean. ROM1 is rosin, medium activity, with halides, and needs washing.

Practical reading: a 1 at the end or the letter H means the residue has to come off.

04

What has to be washed off

The question almost everyone gets wrong, and in both directions.

Two opposite mistakes

Not washing off what has to go. Water-soluble and acid-based fluxes go on eating the copper for years. A board that works perfectly today can have green, corroded tracks in a year.

Washing off what is meant to stay. No-clean residue is deliberately encapsulated and inert. Wash it halfway — with too little solvent or without rinsing — and you break the capsule and spread the active residues out across the board, where they now sit exposed.

The decision
FluxesMust be washedHow
No-clean, REL0NoLeave it. If you want it neat: wash completely, never partially.
Rosin, ROL0OptionalCosmetic. IPA and a brush, see Chemicals, recipe 20.
Rosin, ROM1 or ROHYesIPA, or hot water with surfactant followed by IPA and drying.
Water-soluble, OAYes, alwaysHot deionised water, brush, then IPA and oven drying.
Acid-based, INYes, alwaysNeutralise with dilute soda — sodium carbonate, 20 g/l water — rinse thoroughly, dry.

Drying is half the job. Water left under a component gives leakage currents and corrosion. At least 30 minutes at 50 °C, or a day in dry air.

One test: run a finger over the board after drying. If it feels tacky, residue is still there.

05

The iron and the tip

The tip matters more than the iron, and looking after the tip matters most of all.

Temperature and tip
WorkAlloyTemperatureTip
Fine-pitch SMDSn63Pb37300–320 °CSmall chisel 1 mm or knife tip
Fine-pitch SMDSAC305330–350 °CSame
Through-plated, normalSn63Pb37320–340 °CChisel 1.6–2.4 mm
Through-plated, normalSAC305350–370 °CChisel 1.6–2.4 mm
Ground plane, lugsAny370–400 °CChisel 3–5 mm or a solder pot
Heavy cable, sheet metalSn95Sb5400–450 °CHigh-power iron, large mass

The tip must always be tinned

The iron plating on the tip oxidises the moment it is left clean and hot. A thin layer of solder protects it, and a tip that has oxidised through stops taking solder at all.

  • Tin it immediately when a new tip is heated for the first time.
  • Wipe it just before you solder, not after. A wiped tip left to sit oxidises in seconds.
  • Leave a blob of solder on it before switching off. That step decides tip life more than anything else.
  • Brass wool rather than a wet sponge. The sponge cools the tip by 50–100 degrees every time and wears the plating through thermal shock.
  • Never file a tip. The iron plating is a few micrometres thick, and under it is copper, which the solder dissolves in minutes.
Black tip that will not take solder

The oxide has won. Drop to 250 °C, rub it in flux-cored solder shavings or in dedicated tip tinner — a paste of sal ammoniac and tin. If that fails, the plating is broken through and the tip is finished. The usual cause: the iron left switched on and hot with no solder on the tip.

06

Technique

Four seconds, in the right order.

  1. Clean and fluxed. Oxidised leads or pads get pickled or scraped. Flux on.
  2. Tip against both parts. Lead and pad at the same time, so the heat goes to the joint and not just to one of them.
  3. One second. Let the joint take heat.
  4. Solder to the joint, on the opposite side from the tip. It should melt off the work.
  5. Remove the wire, then the tip. In that order.
  6. Do not move it until it has solidified. Movement during solidification gives a grainy, cracked joint — worst with non-eutectic alloys that have a pasty range.

The whole sequence should take two to four seconds on a normal joint. If it takes ten something is wrong: too low a temperature, too small a tip, or too little flux.

What a good joint looks like

  • Concave — the solder creeps up the lead and forms a fillet. A round, convex blob means the solder has not wetted.
  • All the way through on plated-through holes — solder should show on both sides.
  • Even surface. Bright on leaded solder, dull but smooth on lead-free. Grainy or crystalline means the joint moved while solidifying.
  • Thin. The outline of the lead should show through the solder.
07

SMD and hot air

Surface mount is not harder, only different. Two tricks do most of it.

Drag soldering

For closely spaced pins: flux generously, lay down a generous amount of solder, and drag the tip slowly along the row of pins. Surface tension pulls the solder where it belongs and leaves the bridges behind. It looks wrong while you are doing it and right when you have finished. The flux is what makes it work — without it the bridges stay.

Solder paste and hot air

  • Paste is solder powder in flux gel. Keep it refrigerated and use it within a few months — old paste separates and gives solder balls.
  • A stencil gives an even amount. Without one, a thin bead along the pads is enough; surface tension sorts out the rest.
  • Hot air 300–350 °C with a wide nozzle and low airflow. High airflow blows small components away.
  • Heat the whole area, not one spot. Move the gun in circles until the entire field melts at once.
  • Self-centring is real: surface tension pulls the component into place as the solder melts. Perfect placement beforehand is not necessary.
Neighbouring components and plastic

Hot air heats everything within a few centimetres. Mask with Kapton tape or aluminium foil, and remove plastic parts, batteries and electrolytic capacitors nearby. An electrolytic capacitor that has boiled looks intact and is ruined.

08

Desoldering

Harder than soldering, and the thing that most often ruins boards.

  • A solder sucker for single holes. Heat until the solder is fully liquid, then suck fast. Half-melted solder clogs the sucker.
  • Braid for surfaces and pads. Flux the braid, lay it on the joint, heat through the braid. Dry braid wicks badly.
  • Low-temperature solder for multi-pin components, see Bi58Sn42. The only thing that works well on a 40-pin chip without a hot air station.
  • Hot air for SMD. Heat the whole component and lift it with tweezers once everything has melted.
  • Always add fresh solder first when removing old. Fresh flux in the new solder makes the old release. Trying to wick dry, oxidised solder is the commonest cause of lifted pads.
Pads lift from pulling, not from heat

The copper foil is held down by adhesive that softens at around 120 °C. Pull on a component while the joint is hot and the pad comes with it. Heat alone almost never lifts a pad.

09

Brazing

Above 450 degrees. The joint is an order of magnitude stronger than soft solder, and the technique is a different craft.

The gap is the design

Brazing filler is drawn into the joint by capillary action, and capillary action is strongest in a narrow gap. 0.05 to 0.15 mm is right. A slack joint does not fill — the filler runs straight through. An interference fit does not fill either, because the filler cannot get in. Design the joint as a lap, not a butt, exactly as you would an adhesive joint.

The sequence

  1. Clean mechanically and degrease. Oxide and grease stop everything.
  2. Fluxes over the whole joint area, borax- or fluoride-based depending on the metal.
  3. Heat the whole part evenly, not the joint alone. The filler travels towards the hottest place, so heating from the wrong side puts it in the wrong place.
  4. The right temperature is visible: the flux turns clear and watery just before the filler melts. That is the signal.
  5. Touch the rod to the joint, not to the flame. If it melts off the metal, the part is hot enough.
  6. Let it cool slowly. Quenching in water cracks the joint and hardens the steel.
  7. Remove the flux residue — it is glassy and corrosive. Hot water while the part is still hot, or citric acid.
Three risks
  • Cadmium in old silver brazing rods. See the alloy card.
  • Zinc fume from brass braze causes metal fume fever. Ventilation.
  • Fluoride fluxes cause burns that are not felt until hours later, exactly like hydrogen fluoride. Gloves and rinsing.
10

Material table

What can be soldered, and what needs something else.

Material against method
MaterialSoft solderBrazeFluxesNote
Copper✓✓RosinThe ideal substrate. Everything solders to it.
Brass✓✓Rosin or acidZinc oxide on the surface needs a more active flux.
Tinned sheet✓✗RosinEasiest of all.
Steel and cast iron!✓Acid-based / boraxWorks with acid flux, but brazing is the right method.
Stainless!✓Phosphoric acid or fluorideThe chromium oxide has to go. Needs an aggressive flux.
Nickel and nickel-plated✓✓RosinSolders well. A common finish on connectors.
Silver and gold✓✓RosinGold dissolves into the solder — thin layers vanish completely.
Aluminium✗!Special Al fluxThe oxide reforms instantly. Needs Al-Si filler or a friction method.
Titanium✗✗—Needs vacuum or shielding gas. Not a hobby job.
Chromium and chrome-plated✗✗—Grind the chrome layer off first.
Carbide✗✓Fluoride-basedThis is how saw tips are attached to a blade.
Glass and ceramic✗✗—Metallise first, see Plating.
11

Troubleshooting

Arranged by what the joint looks like. Appearance is nearly always enough to make the diagnosis.

When the joint did not come out right
SymptomLikely causeRemedy
Round, convex blob sitting looseCold joint — the solder melted against the tip, not the workMore heat and a larger tip. Melt the solder against the work.
Grainy, crystalline surface on leaded solderThe joint moved while solidifyingHold still until it has set. Reheat.
Dull surface on lead-freeNormal for SAC305No action. The joint is probably fine.
The solder draws back and leaves bare copperDewetting — contamination or oxide under the solderWick it off, clean mechanically, flux generously, do it again.
The solder will not wet at allOxide layer, too little flux, or the wrong substrateMore flux. Check in the material table that the material can be soldered.
Bridges between pinsToo much solder, too little fluxFlux and drag the tip along the row, or remove it with braid.
The component stands on one endTombstoning — the pads melted at different ratesHeat the whole area evenly. The same amount of paste on both pads.
Small solder balls around the jointOld or damp solder paste, or heating too fastFresh paste. Preheat the board before reflow.
The pad came off the boardMechanical pulling on a hot jointCannot be repaired neatly — a copper wire as a bridge. Never pull on hot joints.
The board went brown or blisteredToo long at too high a temperatureShorter time, higher temperature, larger tip.
The joint cracked after a period in serviceThermal fatigue, or too thick an intermetallic layerRelieve it mechanically. Avoid reheating the same joint many times.
Green, corroded tracks after monthsFlux residue that should have been washed offToo late for what has corroded. See section 4 next time.
The tip is black and will not take solderOxidised iron platingTip tinner at 250 °C. Always tin before switching off.
The tip has pitted and gone thinThe plating is breached and the solder is dissolving the copperThe tip is finished. Never file a tip.
The braze ran straight through the jointToo large a gap — no capillary action0.05–0.15 mm. Redesign the joint.
The braze ended up in the wrong placeUneven heating — the filler travels towards the hottest pointHeat the whole part, not the joint alone.
Thin whiskers growing out of a joint after yearsTin whiskers on clean, bright tinThey can short. Leaded solder, or a duller and thicker tin layer, counteracts it.
Fume and ventilation

The smoke from soft soldering is burnt flux, not metal, and it irritates the airways and can cause sensitisation on repeated exposure. Use a fume extractor or a fan that draws the smoke away from your face. Brazing adds metal fume and demands real ventilation.