Contact Us
  • Home
  • BLOG
  • What Causes Blowholes in a Through-Hole Joint?

What Causes Blowholes in a Through-Hole Joint?

A blowhole is gas that was already inside the joint before the solder arrived. It is rarely the iron and rarely the flux on its own. Solder fills the barrel from the source side, the barrel has to vent to the other side, and anything holding water or plating chemistry in the wall releases vapour as the alloy freezes. The hole on the top face is the last bubble to reach the surface.

Two sources account for most of them. The laminate holds moisture, typically 0.1 to 0.5 percent by resin weight after a week in a workshop at 30 C and 60 percent relative humidity, and that water turns to steam at wave temperature. The copper plating holds organics from brighteners and carriers, invisible at goods in, released when the barrel passes 250 C. A board can pass bare board electrical test and still blow holes on every third joint.

What the standard actually measures

IPC-A-610 Table 7-4 sets vertical fill at 75 percent for Class 2 and Class 3, and the unfilled height includes the depressions on both the source and the destination side. A blowhole therefore spends the same budget as an unfilled top: a joint that looks acceptable from above can be a defect once the crater is counted. There is a conditional allowance down to 50 percent or 1.2 mm for leads tied into internal thermal planes, and it comes with wetting conditions on the source side. It is not a general Class 2 relaxation, which is the most misquoted line in the standard. Class 3 does not take the allowance at all.

Why the barrel holds the gas

A plated through hole is a closed tube with one open end while the solder rises. Venting depends on the clearance between pin and hole wall, on the drill wall, and on how well the plating has sealed the resin. A pin that fills the hole too tightly leaves no path for the air. A hole with rough drill walls and smear traps the chemistry that later comes out as gas. Thin plating makes it worse, because a barrel at 20 um has more chance of a pore than one at 30 um. The same applies to back drilled holes, where the unused stub becomes a pocket, and to holes that have already been through one soldering pass.

Aspect ratio pushes in the same direction. A 0.8 mm hole in a 1.6 mm board is 2 to 1, the plating throws evenly and the rinse water exchanges out of the barrel. A 0.3 mm hole in a 3.2 mm board is close to 11 to 1, and the middle of that barrel is where rinse water and brightener stay longest. It is also the part of the barrel that no X-ray at goods in will look at.

Baking, the number most people get wrong

IPC-1601 gives the bake before assembly as 120 plus or minus 5 C, with time set by how long the board has been out of the bag: one hour for 5 to 60 days, two hours for 2 to 6 months, four hours for 6 to 12 months. The bake belongs just before soldering rather than at goods in, because every bake oxidises the finish and grows intermetallic on a thin tin or HASL layer. A sealed bag with a humidity indicator card is good for a year. Once the card has moved, the bag is no longer protection and the board is a moisture source at the wave.

If you want to know whether a panel is wet rather than guess, weigh a coupon of the same thickness on a balance, bake it two hours at 100 C, let it cool and weigh again. The difference is the water, and it gives a number to argue with instead of an opinion about the oven.

Where builds go wrong

Baking hot and long to be safe. Twelve hours at 125 C removes the water and takes the solderability with it. The board then wets poorly, the operator adds dwell time, and the extra heat produces the blowholes the bake was meant to prevent.

Chasing the profile while the plating is the source. Raising preheat and slowing the wave helps water and does nothing for carbon included in the copper. If the holes that blow are a subset of the board, or one panel out of a lot, the pattern points at the bath and the drill, not at the oven.

Judging by the top face. A crater can look shallow and still break the 75 percent fill once the source side depression is added. Class 3 builds need X-ray, and a cross-section is the only way to see whether the void sits at the plating interface or in the middle of the joint.

How we hold the joint

SUPERB PCBA specifies 25 um minimum copper in the barrel and holds the plating bath on a carbon treatment schedule rather than waiting for a defect. Boards are baked to IPC-1601 with the time set by date code and storage, sealed bags are opened only against a build, and the preheat is set so the top side is dry before the wave reaches the joint. Through hole joints are checked by X-ray after wave and by cross-section on the first article, and the finish on the pins and the barrels is recorded per lot.

The wave side of the line is described on the wave soldering page and the acceptance criteria on quality standards. The barrel itself is covered in plated hole copper, and the soldering requirements in IPC J-STD-001.

What to send with the question

The board thickness and layer count, the storage history since the panels were made, the pin to hole clearance, the finish on the pins, the soldering method, and a photo of a failed joint with its location on the panel. Six items, and the answer separates a moisture problem from a plating problem before anyone changes a profile.

Need a quote or a review of your build against these limits? pcba@superb-tech.com · WhatsApp +86 133 9608 1443