Solder Bridge

A solder bridge is an unwanted connection of solder spanning two or more conductors that are meant to be electrically isolated — adjacent IC pins, neighboring pads, or closely spaced traces. The result is a short circuit. Bridges are among the most common defects on fine-pitch work, where the gaps between pins are measured in fractions of a millimeter, and they are the classic self-inflicted wound of board repair: the fix you just made works, but two pins over, a shiny strand of solder is quietly connecting things that must never meet.

Why bridges form

Excess solder is the usual culprit, compounded by too much heat, too little flux, or pad geometry that leaves no room for error. In hand soldering, dragging surplus solder across a row of pins or letting the iron linger pulls bridges between neighbors. In reflow work, excess paste, paste smear from a misaligned stencil, or solder beading between pads produces the same outcome. The highest-stakes variant lives under BGA packages: during BGA rework or reballing of an ASIC, overheating or uneven pressure can cause adjacent balls to slump together beneath the chip, creating a bridge that no visual inspection from above will ever see. Factory test documentation for hashboards calls this out explicitly — boards that fail frequency testing get inspected for shorts, tin bridges, and foreign matter such as stray solder balls left over from rework.

What a bridge does to a hashboard

Consequences scale with what got connected. A bridge across two signals in the chip chain can prevent the chain from initializing — a board that enumerates zero chips or stops dead at one position. A bridge from signal to ground kills communication outright. The dangerous case is a bridge across a power boundary: shorting a hash domain's supply to an adjacent net stresses the DC-DC stage and can cascade into burned components the moment power is applied. This is why disciplined benches measure resistance across domain rails before the first power-up after any rework — a near-zero reading across nets that should be distinct costs ten seconds to catch with a multimeter and can save a board's power stage.

Detection and removal

Many bridges are visible under magnification as a shiny strand or blob linking two points that should be separate; a stereo microscope and raking light from a low angle make them jump out. The smallest hide between pins or under packages and are caught only by continuity testing between adjacent nets. Removal is a solved problem when done properly: flood the area with fresh flux and draw the excess away with solder wick, or reflow the row with a clean, well-tinned iron tip and let surface tension carry the surplus off — solder wants to sit on pads, and with adequate flux it will pull itself apart from a bridge. What does not work is chasing a bridge with more heat and no flux, which oxidizes the joint and smears the short around. After clearing, re-verify that the previously shorted nets read open, and inspect the neighboring pins you just heated.

The bench habit that prevents them

Bridges are cheap to prevent and expensive to power up. Use less solder than instinct suggests, keep flux generous, inspect every reworked area under magnification before it leaves the bench, and make the pre-power resistance check a reflex rather than a response to smoke. A repaired board should then re-run the full test sequence from the beginning — the standard that separates a proper repair from a patch, and the one a professional bench applies before any board goes back into service via a repair intake.