Thermal Pad Voiding: Causes and Control

A thermal pad is soldered like any other joint and judged differently. Instead of a fillet that can be inspected from the side, the connection is a large area under a package, and the quality of the joint is expressed as the proportion of that area which is not filled with solder.

Thermal pad voiding is the reason that measurement exists. A void under a power device raises the thermal resistance of the path from the die to the board, and the same void reduces the mechanical strength of the attachment. Both effects grow with the size of the voided area and with where it sits.

Why Voids Form Under a Thermal Pad

A void appears when something other than solder occupies the space between the pad and the component. The usual causes are flux volatiles that are released during reflow and cannot escape, air that is trapped when the paste is printed or when the part is placed, and the shrinkage that occurs as the alloy solidifies.

The geometry makes all three worse. A large area with a thin gap offers an easy path for air to be trapped and a long path for volatiles to travel before they reach an edge. That is why a thermal pad is more prone to voiding than a row of small joints printed with the same paste.

Paste Volume and Aperture Splitting

Too little paste leaves gaps that the solder cannot fill, and too much paste traps more flux under the component than can escape. The window between the two narrows as the pad area grows, and it is different for every combination of pad size, paste and profile.

Splitting the aperture into a grid of smaller openings rather than a single full pad opening is the standard remedy. It gives the volatiles a network of channels to travel along and out, and it produces a more even distribution of solder across the pad rather than a pool in the middle. The split also changes the total printed area, so the volume has to be recalculated rather than assumed.

The Reflow Profile and Outgassing

The profile determines how much of the flux vehicle has evaporated before the alloy melts. A slow ramp with a proper soak lets the volatiles leave while the paste is still permeable, and a profile that reaches the liquidus quickly traps them under a skin of molten solder.

The soak also has to be long enough for the whole pad to reach temperature evenly. A large thermal pad on a thick board heats from the edges inwards, and if the alloy melts in the centre while the outside is still cold, the vapour that is generated has nowhere to go except into the joint.

Flux Chemistry and Its Contribution

The flux is what makes the joint possible and what causes most of the voiding. A more active flux with a higher solvent content will wet better and produce more vapour, and the balance between the two is a property of the paste rather than of the machine.

Where voiding is the limiting defect, a paste with a lower volatile content or a different activation system is worth evaluating, even though it may need a slightly hotter profile. The change should be qualified on the real assembly rather than on a test coupon, because the escape path depends on the component and on the layout around it.

Surface Finish and Wetting

The finish on the pad affects how the solder spreads. A finish that wets quickly allows the alloy to flow across the pad and close the gaps early, which can trap the volatiles that are still being generated underneath. A finish with slower wetting gives them more time to escape.

Oxidation works in the other direction. A pad that has oxidised during storage wets poorly, and the solder sits in islands rather than covering the area. Storage conditions and the shelf life of the board belong in the same conversation as the paste and the profile, since all three determine whether the pad is wetted at the moment the alloy melts.

Measuring Voiding with X-ray

X-ray inspection is the only practical way to see underneath the device. The image shows the solder as a bright area and the voids as darker regions, and software can calculate the voided proportion of the pad and the position of the largest void.

Solder voiding measurement has limits. It cannot see a void that is not in the path of the beam, and the contrast depends on the thickness of the solder and on the settings of the machine. A single image of a single unit is evidence about that unit; a picture of the distribution across many units is evidence about the process.

What Void Percentage Actually Matters

The acceptable void proportion follows from the thermal requirement. Where the pad is the main path for heat leaving the device, a large void in the centre raises the junction temperature, and the limit is set by the thermal design rather than by a general rule.

The position matters as much as the area. Several small voids spread across the pad are less harmful than one large void in the middle, because the heat has more than one path to follow. Specifications that mention only the total percentage can therefore be met by a joint that performs badly.

Design Choices That Reduce Voiding

The layout and the footprint influence how bad the voiding will be. A pad with a defined escape route for the vapour, a solder mask opening that does not restrict the flow, and a thermal via arrangement that does not provide a path for the paste to drain all change the result.

Where a via is placed in the pad, it has to be filled and capped, otherwise the paste drains into the barrel during reflow and leaves a cavity that shows up as a void. The requirements are the ones described for a via plugging process, and they belong in the same package as the manufacturing tolerances.

Process Control and First Article

The first article establishes what the process produces on that product; the routine checks confirm that it has not changed. The X-ray measurement should be made on a defined sample per build, at a defined magnification, and compared with the limits agreed for the product.

Where a change in the voiding is found, the trail runs back through the paste batch, the stencil condition, the placement and the profile. Keeping those records per build is what makes the investigation short, and it is the same habit that supports any yield and quality control programme on a board with power devices.

FAQ

Is zero voiding achievable? Not in production. The aim is to keep the voided area small and dispersed, within limits that are set by the thermal requirement of the device.

Does a large void always cause a failure? Not immediately. It raises the thermal resistance, so the device runs hotter, and the effect appears as reduced margin rather than as an instant defect.

Can voiding be measured without X-ray? Only indirectly. Thermal measurements on the finished product can show the effect, and they are the reference that the X-ray limits are set against.