The board carries a mix of parts: fine pitch on the top, and a row of through-hole connectors that the harness will plug into. Running the panel over a wave would mean the parts already placed on the underside go through molten solder, so the connector is printed and reflowed instead. Paste goes onto the board over the holes, the pins push through it, and the oven does the rest.
Pin in paste through hole assembly removes a process, and it adds a problem: the paste has to fill a barrel it was not made for, from one side, with no second chance to top it up.
What pin in paste asks of the print
The aperture is not the pad. It is the pad plus an overprint, because the paste has to fill the hole as well as form a fillet. A print that gives a flat, correctly sized deposit on the pad leaves a barrel half empty, and a half filled barrel is the failure mode this process is known for.
The extra paste comes from the aperture area, and the stencil has to hold a gasket against the board around each hole while the squeegee passes. That gasket is why hole position tolerance matters more here than on a normal print: a hole that sits off centre under the aperture leaks paste onto the underside of the board, and the panel is printed on the conveyor before anyone sees it. Stencil condition matters for the same reason, and the cleaning interval is on stencil cleaning frequency.
Paste volume is the whole problem
| What is being set | Typical target | What goes wrong outside it |
|---|---|---|
| Paste volume against hole volume | Roughly 1.5 to 3 times the barrel volume, set by hole and pin diameter | Below it the barrel fills short, above it the paste floods the connector body and the pin |
| Barrel fill after reflow | 75 percent of the barrel height, the level the acceptance table asks for in most classes | A short fill passes a visual check and fails a thermal cycle or a pull test |
| Aperture overprint | Enough area to carry the extra volume without touching the next pad | Too little gives a short fill, too much bridges to the neighbouring hole |
| Pin to hole clearance | Large enough for paste to vent as the pin enters | A tight fit pushes paste out of the barrel and up the pin, and the fill drops |
These are starting points, not settings. The number that matters is the one measured on the first article, cut through a barrel and read at magnification.
Where the joint fails
Four failures cover most of what goes wrong. The barrel fills short, which shows as a fillet on the top and a void under it. Paste is pushed out as the pin enters and lands on the connector body, which fails a cleanliness call and can bridge under the housing. The connector floats on the paste and lifts, so the shoulder sits off the board. And the paste wicks up the pin instead of into the barrel, which leaves a joint that looks formed and is hollow.
All four are volume and vent problems, not temperature problems. Chasing them in the profile wastes the first article.
Which parts belong in the oven, and which do not
Pin in paste works best on connectors with a moderate pin count, pins of a size a stencil can overprint, and a body that tolerates reflow temperature. It works badly on heavy terminals that sink heat out of the joint, on housings that trap the vent path, and on parts with a large thermal mass relative to the paste available. Those parts go to selective soldering or to wave, and the choice is made per part before the build, not after the first article.
Where a connector is a press fit part, it is a different process entirely: no paste, and a hole tolerance that carries the joint. That tolerance is on press fit connector hole tolerance.
Inspection: what the camera shows and what it does not
The top fillet is visible optically, and it is a poor guide to what is inside the barrel. X-ray shows the fill as a density column and is the check that decides pass or fail, with the percentage read against the acceptance class. The first article gets a cross section, because a slice answers the vent question in a way no image can, and it is the record that shows the aperture and the print were set right.
How SUPERB PCBA runs it
The process runs on seven SMT lines with placement to 01005 (0.4 by 0.2 mm) at up to 40,000 CPH, with 3D SPI after print so the deposit is judged before the connector is placed on it. Reflow is a twelve zone nitrogen oven at under 500 ppm oxygen with a SAC305 profile, which gives the soak a barrel needs to fill before the paste wets.
Where a part does not suit pin in paste, the same build goes to selective soldering rather than wave, so the parts on the underside are not exposed. That path is on selective soldering process. After reflow, inspection runs optical for the fillet and X-ray for the fill, and electrical test follows on in circuit or flying probe depending on access.
Questions buyers ask
Can every through-hole part be run pin in paste?
No. Pin size, hole clearance, body mass and vent path decide it, and a few part types are excluded by their own handling notes. The list is checked against the BOM before the build starts.
Is a separate stencil needed?
Usually yes, because the overprint changes the aperture area and the thickness for that region. A step stencil covers some builds, and the choice is made when the tooling is ordered.
How is a short fill corrected?
By changing the aperture area or the pin to hole clearance, not by raising the paste. A raised paste on a hole that cannot vent gives a worse result and a dirtier connector.
Next step
Send the BOM, the Gerber set and the connector drawings, with a note on which parts are through-hole and which side they sit on. The process note comes back with the parts that suit pin in paste, the ones that should go to selective soldering, and the fill level the first article will be cut and measured against.