Solder Float Test: Thermal Stress and Plated Hole Integrity

The solder float test subjects a plated hole to a sudden thermal shock by floating a sample of the board on molten solder for a fixed time, then examining the barrel for cracks. It is one of the oldest reliability tests in printed circuit fabrication and it remains useful because it is fast, cheap and directly related to the stress a hole sees during assembly. A barrel that survives the float is a barrel that will probably survive reflow and wave soldering.

What the Solder Float Test Does

The sample is placed on the surface of a solder pot heated to a defined temperature, held there for a set time and then removed and cooled. One face of the board is in contact with the molten metal, so the sample experiences a large temperature gradient through its thickness, which is what produces the differential expansion between the copper barrel and the laminate.

After the float the holes are sectioned and examined. What is looked for is cracking in the copper of the barrel, separation between the barrel and the inner layer copper, and delamination of the laminate around the hole. All three are consequences of the same thermal stress, and all three can exist without any visible external sign.

Test Temperature and Dwell

The test temperature is chosen to represent the assembly process. A tin-lead process is represented by a pot at around 260 C, while a lead-free process is represented by 288 C. The dwell is typically ten seconds, and the sample is usually floated once, or several times for a product that sees multiple thermal excursions.

Where the product sees more than one pass, the test can be repeated on the same sample. Repeating the float accumulates the stress in the same way that multiple reflow passes do, and a barrel that passes once but fails on the third float is telling the designer something about the assembly sequence rather than about the material.

Sample Preparation

The sample geometry matters as much as the temperature. A large coupon draws heat away from the hole and reduces the stress; a small one concentrates it. The coupon should therefore be a defined size and should contain holes of the smallest diameter and highest aspect ratio in the product, because those are the ones that fail first.

Holes to be examined should be distributed across the coupon rather than clustered, and the coupon should be taken from a production panel rather than made specially, so that the plating it contains is the plating the customer will receive. The surface finish is normally left in place, since removing it would change the thermal response.

Plated Hole Integrity and Barrel Cracking

Plated hole integrity depends on the copper thickness in the barrel, the ductility of the deposit and the adhesion between the barrel and the inner layer. Cracking initiates where the barrel is thinnest, which is usually the middle of the hole, and propagates around the circumference under continued stress.

Barrel cracking is described by its extent. A crack that passes through the full thickness of the copper creates an open circuit; a partial crack is a latent defect that may grow during assembly or in service. The test result is normally reported as the number of holes examined and the number that showed cracks, with the location and extent of each.

Evaluation by Microsection

The evaluation is done by microsectioning the holes and examining the plating, and the quality of the section decides the value of the test. The sample has to be cast so that the hole is supported, ground in steps and polished without smearing the copper, since a smeared section can hide a crack or create one that is not there.

The standard microsection methods apply, and the examination should be made both as-polished and after a light etch, because the etch reveals the grain structure and the interface with the inner layers. A crack that is invisible as-polished often becomes clear after etching.

Interpreting Failures

A single cracked hole in a coupon may be a local defect, such as a plating nodule or a void, rather than a process that is out of control, and it should not be treated as a process failure until a second coupon confirms it. A pattern of cracks in the smallest holes points to insufficient barrel thickness; cracks in all holes regardless of size points to brittle plating or to a laminate that is expanding more than the copper can follow.

Brittle plating is usually a chemistry problem, most often an organic additive out of balance or an excessive brightener concentration. Where the failure mode is separation between the barrel and the inner layer rather than a through crack, the cause is usually the etchback or the inner layer treatment, and the section will show the gap at the interface.

Comparing With Thermal Cycling

The float test applies a single severe shock; thermal cycling applies many moderate ones. A barrel that survives the float may still fail after a thousand air-to-air cycles, because the failure mechanisms are not identical: the float is dominated by the expansion of the laminate, while cycling also accumulates fatigue in the copper.

For high-reliability products the two are complementary rather than alternatives, and the float is used as a fast process control while the thermal shock or cycling test is used for qualification. The float result should not be presented as evidence of cycling performance.

Common Sources of Error

The most common error is an inconsistent thermal contact. If the sample is not flat, only part of it touches the solder, and the holes in the raised area see less stress than the others. Samples should be flat, clean and free of burrs, and the float should be timed from the moment the sample touches the surface.

The second error is sectioning a hole that has already been damaged by the saw. Cracks in the copper produced during cutting run across the barrel rather than around it, and their orientation is the giveaway. Casting the sample before cutting, and using a slow saw with the correct blade, prevents most of this.

Records and Qualification

The record should carry the test temperature, the dwell, the number of floats, the coupon identification and the number of holes examined, together with a description of any defect found. Those fields make the result reproducible by another operator and comparable between lots.

Where the test is used for process control, the samples should be taken at a defined frequency and the results trended. A shop that floats a coupon only when a customer complains learns about problems late, and it cannot tell whether a failure came from the plating line, the laminate or the press. Recording the hole wall condition at the same time links the two sources of evidence.

FAQ

What temperature should the solder float test use? Around 260 C for a tin-lead assembly process and 288 C for a lead-free one, with a typical dwell of ten seconds. The temperature should represent the process the product actually sees.

How many holes should be examined? Enough to distinguish a local defect from a process trend. A single hole proves very little, while a set covering the smallest and highest aspect ratio holes in the product gives a usable answer.

Does passing the float test guarantee reliability? No. It demonstrates resistance to a single severe shock. Fatigue from repeated moderate cycles is a different mechanism and needs a cycling or thermal shock test.