BGA Inspection and Rework Practice
A ball grid array hides its joints by design, which makes inspection a matter of indirect evidence and rework a matter of thermal control. Both activities are more demanding than their equivalents on a leaded package, and both depend on the same thing: knowing what the joints look like when they are good.
The package also concentrates heat in a small area, so rework requires a profile that melts the whole array without overheating the laminate or the neighbouring parts. That profile is developed before the first rework rather than improvised on a live board.
Why Optical Inspection Stops at the Edge
An optical system can confirm that the package is present, level and correctly oriented, and it can examine the outer row of balls from an angle. Everything inside that ring is invisible, so an optical pass says nothing about the majority of the joints.
The outer row is also the least likely to be defective, because it is the row that reflows with the most exposure to the oven atmosphere. Judging the array by its perimeter is therefore doubly misleading.
X-ray and Computed Tomography
Transmission X-ray shows the shape and the position of every ball as a projection, which reveals missing balls, bridges and gross displacement. Computed tomography adds depth, allowing a slice through a single row and a measurement of the joint geometry.
Both methods struggle with a thin joint that is present but not connected, because the projection cannot show the interface. The limit is a recognised characteristic of the technique, and it is why the process controls matter as much as the inspection.
Daisy Chain and Electrical Evidence
A test vehicle with a daisy chained package measures the resistance of a chain of joints, which detects an open or a high resistance joint that the image could not resolve. Chain continuity through thermal cycling is the standard way to qualify a package and a process together.
The technique requires a dedicated coupon, so it is a development tool rather than a production check. Where a product is critical, the coupon data is carried forward as evidence that the process window produces sound joints.
Rework Equipment
Rework of an area array requires a machine with top and bottom heating, a controlled profile and a placement capability accurate enough to align the part. Hot air from above alone cannot melt an array under a package on a heavy board.
The bottom side heater brings the board to a temperature below the alloy melting point, and the top side nozzle supplies the final increment locally. A preheat that is too low forces the nozzle to work harder and risks the laminate, while one that is too high can reflow nearby joints.
Removal and Site Preparation
Removal requires the whole array to be molten at the same moment, so the profile has a dwell rather than a peak. A twist or a lift applied while any ball is still solid pulls pads and damages the laminate.
Site preparation is the step that decides whether the new part will be reliable. Residual alloy is removed with a vacuum and a wick, the pads are levelled, and the residue is cleaned before the new paste is applied.
Reballing
Reballing places new spheres on the package, either with a preformed array in a stencil or with a reballing fixture. The spheres must be the same alloy and the same diameter as the original, since a different size changes the standoff and the collapse behaviour during reflow.
The stencil method is faster for a package that will be reused many times, while the fixture method is used for a single package. Either way the result should be inspected before the part is placed, because a missing sphere is much harder to correct afterwards.
Placing the Replacement
The new package is aligned optically, often with a split vision system that shows the balls and the pads together. Flux is applied thinly, since excess flux creates voids and residue that the original process did not have.
Where paste is used rather than flux alone, the volume has to be controlled, because the collapsed ball height depends on it. A reworked joint that is taller or shorter than its neighbours changes the mechanical load on the package.
Verification After Rework
The reworked joint is verified with the same methods used for the original, which for an array means X-ray plus an electrical test. A functional test that passes should not be taken as proof of a sound joint, since a ball can be attached at one end only.
Where the product is critical, a cross section on a sample from the rework station shows whether the process is producing good joints rather than merely working ones. The practice of recording the profile for every rework belongs to the same discipline described in manufacturing processes.
Reducing the Need for Rework
The best rework is the one that is not required, and the measures that avoid it are the same ones that improve first pass yield: a verified profile, a controlled paste volume, a flat board and a package that has been baked and handled correctly.
Where a design allows, choosing a package with a larger ball pitch or with a visible joint feature such as a solder ball with an exposed collar makes both inspection and rework easier. That choice is made at design time and costs nothing later.
Additional Considerations for This Build
Practical attention to X-ray inspection pays for itself here, because it decides whether the finished board behaves as the drawing intended. Where the requirement is not stated on the fabrication drawing or in the assembly notes, the shop has to assume a default, and that default is rarely the value the design was simulated with. Stating X-ray inspection explicitly, together with the tolerance that applies, removes the assumption and keeps the result predictable from batch to batch.
Process Control and Verification
On a design of this kind, X-ray inspection is the item that decides how the rest of the board is arranged. Where the requirement is not written down, the shop supplies its own default, and the default is chosen for the process rather than for the design. The measurements that matter are the repeatable ones: conductor width and spacing, annular ring, finished hole size, plating thickness and surface finish are all verifiable on a coupon that travels with the panel.
Running a first article through the same checks as the production panel confirms that the two agree, and that comparison is the cheapest form of process control available at prototype stage.
Related reading: our fabrication notes, board quality and design release notes cover the same ground.
A first article check confirms that the process and the drawing agree on the points listed above, and that the coupon data supports the values used in the design.
Where a measurement falls outside the expected window, the sample is retained so that the cause can be established before the balance of the batch is released.
A first article check confirms that the process and the drawing agree on the points listed above, and that the coupon data supports the values used in the design.
Where a measurement falls outside the expected window, the sample is retained so that the cause can be established before the balance of the batch is released.
FAQ
Can a BGA be reworked more than once? It can, but each cycle consumes laminate and pad life, so a second rework on the same site should be a deliberate decision rather than a routine one.
Is X-ray enough to release a reworked board? It shows the geometry, and it cannot prove the interface, so an electrical test on the relevant net should accompany it.
Does rework change reliability? It can, because the joint has an extra thermal history; where the product is critical the reworked site should be recorded and monitored.
Should a reworked package be reballed or replaced? Reballing is economical for an expensive package, while replacement is faster and gives a known history for a low cost part.
