Graping Defect: 5 Causes in Solder Paste Reflow

Graping is a reflow defect in which the paste on a pad turns into a cluster of small balls instead of one smooth fillet. The balls sit on the pad like a bunch of grapes, each one individually rounded, and the pad surface between them is often tinned but not covered. It is a coalescence failure rather than a wetting failure, which is why the joint may still conduct while looking wrong.

The mechanism is oxide on the surface of the solder powder. Each particle is coated with oxide as it melts, and if the flux cannot break that coating the particles stay separate and pull into spheres under surface tension. Everything that increases the oxide, reduces the flux available, or shortens the time the flux has to act makes graping more likely.

What Graping Looks Like and Where It Appears

The defect is easiest to see on a large pad with a thin deposit, such as a thermal pad, a pad under a large component or a test point. The balls are small, numerous and evenly distributed, and they remain separate even when the paste around them has melted. Under magnification the surface of each sphere is bright and smooth, not rough.

It differs from solder beading, where balls appear beside the pad because paste was squeezed under the stencil, and from a cold joint, where the alloy has not fully melted. The distinction decides the response: beading is a printing problem, graping is a paste and profile problem, and they are often seen on the same board for different reasons.

Solder Powder Oxidation and Powder Type

Solder powder oxidation is the root of the defect. Finer powder has more surface area for the same mass, so a type 5 or type 6 powder carries more oxide than a type 4 for the same deposit volume. The paste maker compensates with a more active flux, but the margin is thin, and the balance moves with storage and age.

Powder type selection is therefore a coalescence decision as much as a printing decision. Where the pad is large and the deposit is thin, a coarser powder reduces the solder powder oxidation load and improves coalescence, provided the aperture can still be filled. Our notes on powder grade cover the trade.

Flux Activity and the Volume of Flux Available

Flux activity has to reduce the oxide on every particle and then keep the surface clean until the alloy coalesces. Where the metal load is high and the deposit is thin, there is very little flux present per unit of powder surface, and the flux is exhausted before the job is finished. That is why graping appears on the thin deposits rather than on the tall ones.

Flux activity is the other half. A paste with a mild no-clean chemistry relies on a clean powder and a well-controlled profile; the same paste on aged stock or with a low metal load will grap. Changing the paste to a more active type is a valid fix, but the residue and cleaning implications have to be accepted at the same time.

Soak Time and the Reflow Profile

The soak time is the part of the profile that gives the flux time to work. During the soak, the paste is hot enough for the flux to activate and reduce oxide, but not hot enough for the alloy to melt. If the soak time is too short, the alloy melts before the oxide has been dealt with, so the particles stay separate.

A reflow profile that ramps straight to the peak, often chosen to protect a heat-sensitive part, is a common cause of the defect for that reason. Lengthening the soak time and the time above liquidus, and confirming the peak is high enough to fully melt the alloy, usually reduces the defect more than any other single change. Our notes on the soak profile describe the measurement.

Deposit Volume and Pad Geometry

Deposit volume is the geometric half of the problem. A large pad printed with a thin deposit has a high surface-to-volume ratio, which means more exposed paste and less flux per unit of powder. The same paste that flows perfectly on a fine-pitch pad can grap on a thermal pad beside it.

The remedy is more paste, not more heat. A thicker stencil region, a larger aperture or a step stencil that delivers more volume onto the large pad gives the joint enough material to coalesce, and it also improves the joint the pad was designed for. Our notes on paste rheology explain how the deposit holds together before reflow.

Storage, Age and Handling Effects

Paste that has been stored warm, opened repeatedly or left on the stencil for hours carries more oxide and less effective flux than fresh material. The defect it produces appears gradually, which is why a defect that starts on one shift is often blamed on the profile rather than on the jar.

Storage control is therefore part of the cure. Keeping the jar sealed and refrigerated, warming it fully before use and limiting the time on the stencil all preserve the balance the paste maker designed. A paste that has been opened for a whole shift should be tested on a standard coupon before it is used on a critical pad.

Nitrogen and Atmosphere Effects

A nitrogen atmosphere reduces the oxide that forms during the ramp, so it reduces graping. It does not eliminate it, because the oxide already present on the powder is not removed by the gas. Where graping is severe, nitrogen alone will not fix it, and the paste or the profile has to change as well.

The atmosphere does shift the balance enough to matter on marginal products. A board that graps in air and passes in nitrogen is a board whose margin has been bought with gas, and the same result could usually be obtained with more deposit volume at lower running cost. The trade should be measured per product rather than applied across a factory.

Detection and Acceptance Criteria

Detection is visual and optical. Automated inspection can be programmed to flag an uncovered pad area, but the classification is difficult because the defect differs in severity from a light scatter of balls to a pad that is barely tinned. The practical measure is the proportion of the pad covered, checked against a boundary sample.

Where the balls smear into a single fillet on a second reflow, the original defect was a coalescence failure and not a wetting failure, which is diagnostic information worth recording. Acceptance criteria should say whether any uncovered area is permitted, because a joint with visible spheres is not the same as a joint with a full fillet. The IPC workmanship criteria give the framework.

Fixing Graping in the Right Order

Start with the deposit, because it is the cheapest item to change and the one that most often explains the defect. Then check the paste condition: age, storage, powder type and time on the stencil. Only then change the reflow profile, and change the soak time and the time above liquidus rather than the peak first.

Where those steps run out, the paste itself is the remaining variable. A different flux activity or a different powder grade changes the whole balance, and the change should be qualified on a production board with a follow-up inspection plan, as the sequence in our joint defect notes describes.

FAQ

Is graping the same as a cold joint? No. In this defect the alloy has melted and formed spheres, but the spheres have not merged. In a cold joint the alloy has not reached its liquidus properly, so the surface is rough and the joint is weak. The two can appear together when the reflow profile is badly short.

Can graping be reworked by reflowing again? A second reflow often merges the spheres, which confirms the diagnosis, but it is not a production fix for a product that is assembled once. The cause has to be removed from the print or the profile rather than repaired afterwards.

Does graping always mean the joint is unreliable? A pad with a full coverage of merged solder around the spheres can still form an acceptable joint, and small amounts are common on large pads. The concern is the uncovered area and any sphere that is not bonded, since that is where a crack can start.