Most discussions of BGA (Ball Grid Array) solder voids stop at a single number: IPC-A-610 calls a void a defect once it exceeds 25% of the joint area. In real engineering, where a void sits and what type it is often decide whether the joint fails, more than how much area it covers.
The causes (flux outgassing, moisture) and the basic fixes (paste, reflow profiling, vacuum reflow) are well documented, covered in this guide on solder voiding causes and prevention.
This article focuses on three overlooked dimensions that drive BGA reliability judgment: void location, void type, and the separate problem of thermal pads.

Void Location Matters More Than Total Area
Two solder balls with the same 20% void level can differ sharply in reliability, and the difference is where the void sits.
Stress concentrates at the edges and corners of a solder ball, where deformation under thermal cycling and vibration is greatest. The same void size carries different risk by location.
This means checking only whether the void rate is under 25% is not enough. A ball with 24% voiding at the center may be fully usable, while one with 15% at the edge is the real risk. High-reliability inspection must read location, not just a percentage.

Void Types Under IPC-7095
IPC-A-610 gives a general acceptance line, while IPC-7095, the standard specific to BGA design and assembly, classifies voids by cause and form. It is the more professional basis for reading BGA joints, yet rarely explained.
The via-in-pad void points to something often missed: many BGA voids originate in design, not on the line. An unfilled via inside the pad becomes a gas trap that paste and profile changes cannot fix.
A supplier that flags this kind of manufacturability risk before the prototype build saves you repeated rework later.
Head-in-pillow, by contrast, is not a true void but a non-coalescence defect that looks connected while remaining an intermittent open, so it must be caught by X-ray reading rather than area judgment.
Thermal Pad Voiding Is a Separate Problem
Voids under signal balls affect connection, while voids under the large thermal pad of a power device or QFN affect heat dissipation. These are different problems often treated as one.
A thermal pad is large, so gas under the molten solder has a long escape path and easily leaves wide voids. Those voids raise the device junction temperature and shorten the life of power parts, so the fixes differ from signal balls.
A window-pane stencil splits one large aperture into several smaller ones, controlling paste volume while giving gas multiple escape routes.
For large pads, Venture Electronics relies on nitrogen vacuum reflow soldering, which evacuates chamber pressure while the solder is molten and forces out the trapped gas that ordinary reflow leaves behind.

Reading X-ray, Not Just Running It
The industry knows BGA voids need X-ray, but the point is not whether X-ray exists, it is whether the results are read correctly.
2D X-ray shows a void’s presence and rough area. Judging its position within the ball, or catching HiP, needs 3D CT X-ray with multi-angle reconstruction.
More importantly, the inspector has to read against IPC-7095 type and location criteria, not just report an area percentage. Venture Electronics covers hidden joints on BGAs and QFNs through its PCBA testing service, reading void location and type rather than area alone.

Plan Your BGA Assembly Project
Judging BGA voids is not about the 25% number alone. Location, type, and the separate risk on thermal pads are often what decide whether a joint is usable.
For power-dense or high-reliability BGA boards, align these reading criteria with your supplier at the prototype stage. If your product involves BGAs or has defined void limits, you can review Venture Electronics’ BGA assembly capabilities to match an approach to your reliability requirements.
FAQs About BGA Solder Voids
Q1: Is void location more important than void size in BGA joints? Yes. A void at the ball edge or corner, where stress concentrates, is more dangerous than one of the same size at the center. Venture Electronics reads void location against IPC-7095, not area alone.
Q2: What does IPC-7095 cover that IPC-A-610 does not?
IPC-A-610 gives a general acceptance line, while IPC-7095 is BGA-specific and classifies voids into macrovoids, planar microvoids, shrinkage voids, and via-in-pad voids. It is the more precise basis for judging BGA joints.
Q3: Why are thermal pad voids treated differently from signal ball voids? Thermal pad voids affect heat dissipation, not signal connection, raising junction temperature and shortening power-device life. Venture Electronics controls them with window-pane stencils and vacuum reflow.
Q4: Can solder voids come from PCB design rather than assembly? Yes. An unfilled via inside a pad traps gas and cannot be fixed by paste or profile changes, only by via fill at design stage. Venture Electronics flags this manufacturability risk before the prototype build.
Q5: What is head-in-pillow and why does it matter? HiP is a joint where the ball and paste never coalesce, looking connected but potentially open. It evades area-based judgment and needs X-ray reading experience to catch.


