Soldering Defects and Desoldering
Diagnostics keeps sending you back to 'check for a cold joint' without ever showing what one actually looks like — this chapter is that missing visual reference: the six recognizable solder defects, how to tell them apart on sight, how to desolder cleanly without lifting a pad, and the heat-sensitive hardware (pots, switches, plastic jacks) that needs different technique than a bare PCB pad.
Every troubleshooting chapter on this site eventually says some version of “check for a cold joint” or “look for a solder bridge” without ever showing what those actually look like. Soldering and Tools covers the physical technique of making a good joint — this chapter is the other half: recognizing when a joint went wrong, undoing one cleanly, and the specific hardware (pots, switches, plastic-bodied jacks) that punishes generic technique with heat damage a bare PCB pad wouldn’t show.
Six defects, six distinct appearances
A cold joint looks dull, grainy, and matte gray instead of shiny and cone-shaped — it happened because the joint moved while cooling or never got hot enough to wet properly, and it shows up electrically as an intermittent connection that can pass or fail depending on temperature or vibration. A solder bridge is a thin, often barely visible trail of solder connecting two pads that should be isolated — too much solder, or pad spacing tight enough (fine-pitch ICs) that it happens almost by accident, and it reads electrically as a dead short, sometimes invisible to the eye and only caught with magnification or a continuity check. An overheated joint shows a discolored, browned or blackened pad, sometimes with the pad visibly lifting — caused by holding the iron on too long or running it too hot, and it risks the pad separating from the board entirely, not just a bad joint but a destroyed one. Insufficient solder leaves the pad only partially covered, the lead visible without a proper fillet around it — mechanically weak even if it happens to pass current today, prone to breaking loose from vibration or flexing. Excess solder buries the lead in a ball that obscures the pad and hides whether an adjacent bridge exists underneath it. A lifted pad — the copper itself separated from the board — is a genuine repair, not a reflow: it needs a jumper wire soldered to the exposed trace nearby, not just more heat and solder on a pad that no longer connects to anything.
Desoldering cleanly is a distinct skill from soldering well
Solder wick (braid) pulls molten solder away by capillary action: lay it over the joint, press the iron onto the wick so it heats both the wick and the joint together, and lift iron and wick away together while the solder is still liquid — pressing too hard is the main way to lift a pad here, since the wick itself is doing the heat-sinking and patience matters more than pressure. A solder sucker (vacuum desoldering pump) works faster on through-hole joints: heat the joint fully molten, then trigger the sucker directly over it — for multi-lug components like pots and switches, clear one lug at a time and gently work the lead out rather than trying to clear every lug before attempting removal, which is the fastest way to force a component against solder that hasn’t fully released and lift a pad in the process. ICs and other multi-pin parts come out the same one-pin-at-a-time way, alternating sides as each lead releases, watching for a pad that’s stuck to the lead rather than the board — that’s the early sign of a lift, and the fix is more heat on the remaining joints, not more force on the part.
Heat-sensitive hardware needs different technique than a bare pad
Potentiometers and switches both have heat-sensitive internals — a carbon track or spring contact that excess heat can damage permanently, with no reliable repair once it happens. The fix isn’t a different soldering process, just tighter discipline: heat the lug briefly (two to three seconds), apply solder from the opposite side rather than flooding the lug directly, and use a heat sink clip between the joint and the component body when the connection allows it. Plastic-bodied jacks (common on pedals) melt at temperatures a metal Switchcraft-style jack shrugs off — pre-tinning the wire and the jack terminal separately, then reflowing them together quickly, keeps total heat exposure to the plastic body far shorter than heating cold wire and cold terminal at the same time. A jack’s sleeve (ground) connection deserves particular care regardless of body material — it’s one of the single most common points of intermittent noise and hum when it’s mechanically weak, exactly the kind of fault Pedal Troubleshooting sends you looking for without necessarily naming it as a soldering-technique problem.
Wire selection matters more than it looks like it should
22 AWG stranded copper is the default for nearly all guitar and pedal wiring — stranded specifically, since solid-core wire work-hardens and snaps after repeated flexing inside a cavity or enclosure in a way stranded wire simply doesn’t. Heavier gauge (18 AWG or thicker) belongs on higher-current runs — power supply wiring inside an amp specifically — while thinner 26 AWG is reserved for genuinely fine-pitch connections most guitar work never needs. Speaker wire is its own category entirely, and it’s worth calling out because the connector can mislead you: speaker cable uses the same 1/4“ jacks as an ordinary instrument cable, but the wire itself needs real current-carrying capacity (16 or 14 AWG) that a standard instrument cable’s thin conductor doesn’t have — running a high-power amp’s output through undersized wire produces a voltage drop that measurably reduces power output and risks overheating the cable itself.
Common mistake: reflowing a lifted pad like it’s just a cold joint
A cold joint and a lifted pad can look superficially similar at a glance — both can present as “the connection isn’t reliable” — but they need completely different fixes. Reflowing a cold joint with fresh solder and letting it cool undisturbed solves the actual problem. Reflowing a lifted pad does nothing, because the pad’s copper is no longer connected to the board’s trace underneath it regardless of how much fresh solder sits on top. If a joint keeps failing after a careful reflow, check whether the pad itself has separated before assuming the reflow technique was simply wrong — a lifted pad needs a jumper wire to the nearest intact point on the trace, not another attempt at the same fix that already failed once.