The Repair LibraryRead · Learn · Master

IC Removal Techniques by Package Type

Once you have decided to do a component-level repair, taking the chip off is the next step — and there is no single method, because the right technique is set by the package. This section matches the method to the part: a small two-lead chip or SOT comes off with an iron in seconds; a gull-wing leaded IC like an SOIC or QFP can be removed by hot air, by an iron flood-and-drag, or with a low-melting-point alloy that keeps every lead molten at once; a no-lead QFN or a ball-grid array, with joints hidden underneath, needs hot air. Whatever the package, the goal is the same — get every joint molten together, or freed, so the part lifts straight up without tearing a pad — and the section closes on how to choose, including the important caveat that a low-melt alloy must be cleaned off completely before you solder the new part.

IntermediateMedium Risk22 min read

What You Will Learn

  • You will learn why the removal method is chosen by the package type.
  • You will learn to remove small two-lead and SOT parts quickly with an iron.
  • You will learn three ways to remove a gull-wing leaded IC — hot air, flood-and-drag, and low-melt alloy.
  • You will learn why no-lead QFNs and BGAs need hot air, and how their hidden joints behave.
  • You will learn how the low-melt-alloy method works and why it must be cleaned off before re-soldering.

What You Will Be Able To Do

  • You will be able to choose a removal method that suits the package.
  • You will be able to remove a small two-lead or SOT part with an iron.
  • You will be able to remove a gull-wing IC by hot air, flood-and-drag, or low-melt alloy.
  • You will be able to remove a no-lead QFN or a BGA with hot air, heating the hidden joints evenly.
  • You will be able to use a low-melt alloy and clean it off completely before re-soldering.

Required Tools

No physical tools required. This is a conceptual section.

Section Overview

You have decided to repair at the component level (9.1); now you take the chip off — and the first thing to know is that there is no single removal method, because the right technique is set by the package. What changes from part to part is the lead type, the pitch, and whether the joints are even reachable — and those decide whether you reach for an iron, hot air, or a special alloy. One goal stays constant across all of them: get every joint molten at the same time (or otherwise freed) so the part lifts straight up without tearing a pad (5.4). A small two-terminal chip or an SOT comes off in seconds with an iron — add a little solder to bridge both ends, then rock and lift (6.6). A gull-wing leaded IC like an SOIC, SOP, or QFP gives you three good options: hot air to reflow all the leads and lift (8.4); an iron flood-and-drag, where you lay a fat blob of solder along a row of leads and drag the iron back and forth to keep them all molten while you free that side; or a low-melting-point alloy (the tin-bismuth, ChipQuik-type material) added along the leads to drop the melt point so every lead stays liquid together and the part lifts with just an iron. A no-lead package — a QFN or DFN — hides its leads and a center thermal pad underneath, so hot air is the method (8.4), and that big center pad holds heat and releases last. A ball-grid array is hot air only: heat the whole package evenly and lift it straight up with a vacuum pen once the hidden balls are molten, and if you mean to reuse it, it needs reballing (8.5). The low-melt-alloy method deserves one caution: the bismuth alloy contaminates the joint, so it must be completely cleaned off the pads before you solder the new part, or the new joint will be weak (alloy compatibility). Identify the package, pick the method that melts all its joints at once, never force it, protect the neighbors (8.6), and lift straight up.

Why This Matters

Matching the removal method to the package is what gets a chip off cleanly instead of tearing the pads it sits on — and torn pads can turn a routine swap into a scrapped board. This matters because the wrong method fails or damages: trying to remove a fine-pitch QFP by heating leads one at a time, or a QFN with an iron it can't reach, leaves you prying at a half-stuck part and lifting pads (5.4), while the right method frees every joint together. It matters because the package tells you the tool: reachable gull-wing leads give you three options, but hidden joints under a QFN or BGA force hot air — knowing this before you start means you bring the right tool to the job (8.4). It matters because the low-melt-alloy method is a genuine capability boost: it lets you remove a multi-lead IC with just an iron, safely and with little heat to the neighbors, which is invaluable when you don't own hot airbut only if you know its one critical rule about cleaning it off. It matters because the "all joints molten at once" principle is universal: every method here is just a different way to achieve it, so understanding the goal lets you adapt when a part doesn't match a textbook case. It matters because the hidden-joint packages behave in ways you must respect: a QFN's center pad and a BGA's inner balls hold heat and release last, so heating evenly and never forcing is what keeps them from taking pads with them. And it matters because this is the removal half of the whole chip-swap: a clean removal onto intact pads is what makes the pad assessment, site prep, and installation that follow even possible (9.3 through 9.5). Pick the method the package calls for, and the chip comes off ready for a new one to go on.

Required Prerequisites

  • When to Remove a Chip vs. Replace the Board — Section 9.1 decided that a component-level repair is the right call; this section takes the part off. You should know hot-air removal (8.4), how to protect the neighbors (8.6), iron-and-wick SMD removal (6.6), the package types (6.1), and the low-melting-point alloy and its alloy-compatibility caveat (glossary).
  • Flux (Chapter 3) — essential for every method; it helps the joints reflow and flow
  • Fresh solder and solder wickto add solder for iron methods and to clean pads after (6.6)
  • A low-melting-point (tin-bismuth) alloy removal kitfor the low-melt-alloy method on leaded ICs
  • A scrap board with a range of packages — chips, an SOIC, a QFP, a QFN, and if possible a practice BGA — to practice each method
  • Isopropyl alcohol and a brush — to clean flux and, crucially, remove all low-melt alloy residue (6.7)
  • Eye protection and ventilation; a vacuum pen for liftinghot parts, blown parts, and flux fume (Chapter 3)
  • A temperature-controlled iron and a hot air rework station (Chapter 4; 8.1)
  • Fine tweezers and a vacuum pen — to lift freed parts without touching them (8.4)
  • A range of nozzles including a box nozzle for QFP and BGA (8.3)
  • A preheater if you have one — to ease removal and cut thermal shock (8.2)
  • A magnifier, good light, a board holder, and an ESD-safe surface (Volume 2, Chapter 9)

Real-World Applications

Every chip removal starts by identifying the package and choosing the method, and the choice shapes the whole job. A technician pulling a failed SOIC with no hot air on the bench runs a bead of low-melting-point alloy along both rows of leads, keeps them molten with the iron, lifts the chip, then wicks every trace of the alloy off the pads before fitting the new part. A repairer removing a QFP with hot air fits a box nozzle, reflows all four sides, and lifts it straight up with a vacuum pen (8.4). Someone without hot air removing a small QFP uses the iron flood-and-drag — a fat solder blob dragged along each side of leads — to free the part a side at a time. A repairer taking off a QFN reaches for hot air, because the leads and the big center pad are underneath where no iron tip can go, and heats until even the center pad releases. And anyone reworking a BGA knows it is hot air only — even heat, a straight-up vacuum-pen lift once the hidden balls are molten, and reballing if the part is to be reused (8.5). The failures these methods prevent: pads torn by prying a half-stuck part, a QFN or BGA damaged by forcing, and a weak new joint from low-melt alloy left on the padsall avoided by matching the method to the package and following its rules.

Common Challenges

  • A multi-lead IC won't come free with the iron. You can't keep all the leads molten at once by handuse hot air, a flood-and-drag blob, or a low-melt alloy so every lead stays liquid together (8.4).
  • A QFN or BGA won't release even when the outer joints look done. The center pad or inner balls hold heat and release lastheat evenly and longer, and never pry (5.4).
  • The new joint is weak after a low-melt removal. Bismuth alloy was left on the padswick and clean off every trace before re-soldering (alloy compatibility).

Safety Notes

Risk Level: Medium. These are hot procedures, and the low-melt-alloy method adds a bismuth-and-lead residue you must handle and fully remove.

Professional Tips Before Starting

  • Identify the package, then pick the method. Reachable gull-wing leads give you hot air, flood-and-drag, or low-melt alloy; hidden joints under a QFN or BGA mean hot airchoose before you heat (6.1).
  • Flux everything, and add solder for iron methods. Flux helps every joint reflow, and extra solder on the leads gives the iron a molten bridge to work with (Chapter 3).
  • With low-melt alloy, plan the cleanup. The method is only safe if you wick and clean every trace of the bismuth alloy off the pads before re-solderingnever solder a new part onto contaminated pads.

Matching the Removal Method to the Package

The Method Depends on the Package

There is no universal way to remove a chip — the right technique is dictated by the package, and picking it is the first real decision of the removal. Three things about the package decide the method: the lead type (leaded and reachable, or no-lead and hidden), the pitch (how fine the leads are), and whether the joints can be reached by an iron tip at all (6.1). A part with leads splayed out along its sides can be reached and heated in ways a part with its connections underneath cannot. But whatever the package, the underlying goal never changes: you must get every one of the part's joints molten at the same time — or otherwise release them — so that the part lifts straight up freely, without any joint still solid to tear its pad (5.4). Every method in this section is just a different means to that one end. An iron can melt a chip's two joints together easily, but it cannot hold a QFP's forty-eight leads molten at once by touching them one at a time — so for that part you need hot air, a dragged blob of solder, or an alloy that keeps them all liquid. Understand the goal — all joints molten together, then a clean straight-up lift — and the method follows from the package: identify what you are removing, and the technique chooses itself.

Small Two-Lead and SOT Parts

The easiest removals are the small two- and three-terminal parts, which come off in seconds with just an iron. A chip component — a resistor or capacitor with a metallized end at each side — has only two joints, and an iron melts them together trivially (6.6). The quickest method is to add a little fresh solder across both ends so the iron bridges them, then heat and gently rock or slide the part off its pads while both joints are molten; a blob of solder that touches both ends lets a single tip keep them both liquid at once. An SOT or other small three- to five-lead transistor package comes off much the same way, or with a quick pass of hot air. These parts carry little risk: they are small, their few joints are easy to melt together, and there is no hidden connection to fight. The main cautions are the ordinary onesdon't overheat the little pads, and don't drag the hot part into a neighbor (8.6). For these small parts, an iron and a dab of solder is faster than setting up hot air, so the method is simply: add solder, melt both ends together, and lift. Save the more involved techniques for the multi-lead packages that actually need them.

Gull-Wing Leaded ICs — Hot Air, Flood-and-Drag, and Low-Melt Alloy

The leaded ICs — SOICs, SOPs, and QFPs, with their gull-wing leads splayed out along the sides — are the packages that give you a real choice of three methods. The first is hot air: reflow all the leads at once with a nozzle sized to the part and lift it straight up (8.4) — clean and fast if you have a station. The second is the iron flood-and-drag: you flux the leads, lay a fat blob of solder across a whole row so it bridges every lead in that row, and drag the iron back and forth along the blob to keep all those leads molten at once, gently lifting that side of the part as it freesthen do the same on the opposite row (or work around a QFP's four sides). The molten bridge is the trick: it lets one iron tip hold a whole row liquid together, which touching leads individually could never do. The third is a low-melting-point alloy (the tin-bismuth, ChipQuik-type material): you add it along the leads so it mixes into the joints and lowers their melting point, and then an ordinary iron can keep every lead molten at the same time long enough to lift the part off. All three achieve the same thing — every lead molten together — by different means: hot air when you have it, flood-and-drag or low-melt alloy when you want an iron-only method with little heat to the neighbors (8.6). For a gull-wing IC, pick whichever of the three suits your tools and the board.

No-Lead Packages — QFN and DFN

The no-lead packages hide their connections underneath the body, which takes the iron methods off the table and makes hot air the tool. A QFN (and its two-sided sibling, the DFN) has small lead pads along the underside edges and, usually, a large exposed center pad that carries ground and heat (6.1) — none of which an iron tip can reach, because they are beneath the part. So you remove a QFN with hot air: flux it, heat the whole package evenly with a suitable nozzle until all the underside joints reflow, and lift it straight up (8.4). A low-melt alloy can sometimes be wicked in along the edge to help, but hot air is the reliable method. The feature to respect is the center thermal pad: it is a large mass of solder tied to copper, so it holds a lot of heat and is the last joint to meltthe edge leads may look done while the center is still solid. If you lift then, you fight the center pad and risk tearing it or the part. So you heat evenly and a little longer than the edges alone suggest, and you lift only when the whole part floats free with no resistance (5.4). Never pry a QFN: the center pad releasing last means patience, not force, is what gets it off cleanly.

Ball-Grid Arrays

The ball-grid array is the hardest package to remove, and it is strictly a hot-air job. A ball-grid array has its connections as a grid of solder balls entirely underneath the package, completely inaccessible to any iron (6.1) — so hot air is the only way to melt them, and you cannot even see them reflow. Removal is the careful hot-air routine (8.4): flux, preheat the board, and heat the whole package evenly with a box nozzle until every hidden ball is molten, relying on even heating and the right time and temperature rather than sight. Then lift the part straight up with a vacuum pen — gently, only when it floats completely free, because forcing a BGA whose inner balls have not all melted tears pads under the package where you cannot even see the damage. The inner balls, like a QFN's center pad, hold heat and release last, so even heating and patience are everything. If the BGA is to be reused rather than replaced, its balls are left uneven by removal and it must be re-balled — the reballing step — before it can go back on (8.5). A BGA is the package that most demands hot air, even heat, a straight-up lift, and zero force: respect it, and it comes off; rush it, and you damage a board you cannot easily inspect.

The Low-Melt-Alloy Method and Choosing

The low-melting-point alloy method is worth understanding in its own right, both for how it works and for the one rule that makes it safe. The low-melting-point alloy — a tin-bismuth blend sold in removal kits — works by alloying with the existing solder on the joints: when you add it and melt it into the leads, it forms a mixture that melts at a much lower temperature and, crucially, stays liquid far longer than ordinary solder (tin-bismuth solder). That extended liquid time is what lets a single iron keep every lead of an IC molten at once, so you can lift the whole part with no hot air at alla real advantage on a bench without a station, and gentle on the neighbors because the temperatures are low. But it comes with an absolute rule: the bismuth alloy contaminates the joint. A joint made with bismuth-contaminated solder is weak and brittle, so after you remove the part you must wick and clean every trace of the low-melt alloy off the pads before you solder the new componentthis is the alloy compatibility caution, and skipping it gives you a fragile repair that can fail later. Choosing among all the methods comes down to a few factors: the package (hidden joints force hot air), your tools (an iron-only bench leans on flood-and-drag and low-melt alloy), the risk to neighbors (low-melt alloy and a box nozzle are gentle), and whether the part is being reused (a BGA needs reballing). Match the method to the package, your tools, and the situation — and if you use low-melt alloy, clean it off completely before the new part goes on.

Common Mistakes

  • Using one method for every package. An iron can't reach a QFN or BGA's hidden jointsmatch the method to the package (6.1).
  • Prying a multi-lead part before all joints are molten. It tears pads (5.4) — keep every lead liquid with hot air, flood-and-drag, or low-melt alloy, and lift with no force.
  • Lifting a QFN or BGA before the center pad or inner balls release. They hold heat and melt lastheat evenly and longer, and never force (8.4).
  • Leaving low-melt alloy on the pads. Bismuth makes the new joint weak and brittlewick and clean off every trace before re-soldering (alloy compatibility).
  • Ignoring the neighbors. Removal heat spillsshield, heat-sink, and use a sized nozzle (8.6).

Troubleshooting Guidance

Removal problems trace to the wrong method for the package, or forcing. If a multi-lead IC won't lift with an iron: you can't hold all the leads molten at onceuse hot air, flood-and-drag, or low-melt alloy (8.4). If a QFN or BGA won't release: the center pad or inner balls aren't molten yetheat evenly and longer, never pry (5.4). If you tore a pad: you lifted before all joints meltednext time keep every joint liquid and lift with no resistance. If the new joint is weak after a low-melt removal: bismuth residue on the padswick and clean it all off before re-soldering (alloy compatibility). If the low-melt alloy won't keep the leads molten: not enough was mixed in, or the iron is too cooladd more alloy and flux and keep the tip on the blob. If a neighbor reflowed: heat spilledshield and heat-sink (8.6). If a BGA looks fine but fails after reinstall: hidden joints were damaged by forcing, or it wasn't reballed properlyinspect and re-ball (8.5). The throughline: match the method to the package, keep all joints molten, never force, and clean off any low-melt alloy.

Verification & Testing Methods

Use this as a removal-by-package check:

  • [ ] I choose the removal method from the package — iron for small two-lead parts; hot air, flood-and-drag, or low-melt alloy for gull-wing ICs; hot air for QFNs and BGAs (6.1).
  • [ ] I remove a small chip or SOT by adding solder to bridge the ends and lifting while both joints are molten (6.6).
  • [ ] I can remove a gull-wing IC three ways — hot air, an iron flood-and-drag, and a low-melting-point alloy — keeping every lead molten at once.
  • [ ] I remove a QFN or BGA with hot air, heating evenly and waiting for the center pad or inner balls to release, and lift only with no resistance (5.4).
  • [ ] I never force any part before its joints are all molten or freed, and I protect the neighbors from spilled heat (8.6).
  • [ ] After a low-melt-alloy removal, I wick and clean every trace of the bismuth alloy off the pads before re-soldering (alloy compatibility).

Then try the practice exercises below — removal practice; scenarios differ from the quiz.

Practice Exercises

  1. Remove a small chip and an SOT (6 minutes, applied). On scrap, add solder to bridge the ends of a chip passive and an SOT, melt both joints together, and lift each — with no torn pads.
  2. Flood-and-drag an SOIC or QFP (10 minutes, applied). Flux a leaded IC, lay a solder blob across a row of leads, drag the iron to keep them molten, and free the part a side at a time. Then wick the pads clean (6.6).
  3. Use a low-melt alloy (10 minutes, applied). Remove a leaded IC with a low-melting-point alloy kit, then wick and clean every trace of it off the pads and inspect that the pads are bright and residue-free (alloy compatibility; 6.7).
  4. Reason about a hidden-joint package (4 minutes, reasoning). Explain why a QFN and a BGA need hot air and how their center pad or inner balls behave, and why you must never pry them (5.4).

These core ideas — the method depends on the package, small-part iron removal, the three gull-wing methods, hot air for QFNs and BGAs, and the low-melt-alloy method with its cleanup rule — are tested in the Chapter Quiz at the end of this chapter, where a score of 80% is required to continue.

Key Takeaways

  • There is no single removal method — it is chosen by the package, and the constant goal is to get every joint molten at once (or freed) so the part lifts straight up with no force (5.4).
  • Small two-lead and SOT parts come off with an iron: add solder to bridge the ends, melt both joints together, and lift (6.6).
  • A gull-wing leaded IC (SOIC, SOP, QFP) can be removed three ways — hot air (8.4), an iron flood-and-drag (a solder blob dragged along a row of leads to keep them molten), or a low-melting-point alloy that lowers the melt point so an iron holds every lead liquid at once.
  • No-lead QFNs and DFNs, and ball-grid arrays, hide their joints underneath, so hot air is the method; the QFN center pad and the BGA's inner balls hold heat and release last, so heat evenly, lift with a vacuum pen only when free, and never pry — a reused BGA needs reballing (8.5).
  • The low-melt-alloy method lets an iron remove a multi-lead IC, but the bismuth alloy contaminates the pads and must be completely cleaned off before re-soldering, or the new joint is weak (alloy compatibility); choose the method by package, tools, neighbor risk, and reuse.

Skills Learned

  • You can now choose a removal method that suits the package.
  • You can now remove a small two-lead or SOT part with an iron.
  • You can now remove a gull-wing IC by hot air, flood-and-drag, or low-melt alloy.
  • You can now remove a no-lead QFN or a BGA with hot air, heating the hidden joints evenly.
  • You can now use a low-melt alloy and clean it off completely before re-soldering.

Glossary Additions

  • flood-and-drag — an iron-based method for removing a multi-lead surface-mount IC, in which a large blob of solder is flowed across a whole row of leads (flooding them so the molten solder bridges every lead) and the iron is then dragged back and forth along that blob to keep all of the leads molten at the same time, so the part can be lifted a side at a time. Flood-and-drag lets a single soldering iron hold a whole row of gull-wing leads liquid at once — something touching the leads individually cannot do — and is a common way to remove SOICs and QFPs without hot air; it differs from flood-and-wick, which floods joints in order to wick the solder away rather than to lift a part.

Suggested Next Sections

Must read next:

  • Pad Condition Assessment After Removal — with the chip off, the next section is how to inspect the site it left behind — checking the pads for lifting, damage, bridged or leftover solder, and contamination — so you know whether the board is ready for a new part or needs repair first.

Recommended: