QFN Soldering: Thermal Pad & Stencil Design Guide

QFN soldering succeeds or fails on one feature: the central exposed thermal pad. Print 100% solder paste on it and the molten solder floats the component, tilts it, and lifts the perimeter joints off their pads — the classic QFN failure. The fix is to under-print the thermal pad to 50–70% coverage with a windowpane stencil, plug the thermal vias so they don’t wick solder away, and control reflow so flux outgasses before the joint solidifies. This guide covers the thermal pad and stencil design that makes QFN soldering reliable: aperture patterns, paste coverage, stencil thickness, via plugging, and how to keep voiding under the IPC-7093 limit. Get it right and you get a flat part, a solid ground connection, and void-free heat transfer.

Key Takeaways

• QFN soldering relies on the exposed thermal pad for heat and ground; a full-coverage paste print floats the part and lifts perimeter joints.
• Use a windowpane (grid) stencil aperture at 50–70% paste coverage on the thermal pad — the gaps vent flux and cut voiding.
• Stencil thickness: 0.125 mm (5 mil) for 0.5 mm pitch, 0.10 mm (4 mil) for 0.4 mm fine pitch; perimeter apertures slightly smaller than the copper pad.
• Plug or tent thermal vias with solder mask; open vias wick solder out of the joint and create voids and worse thermal transfer.
• IPC-7093 caps void coverage under the thermal pad at 50%; verify with X-ray, since QFN joints are hidden and can’t be inspected visually.

What Is a QFN and Its Exposed Thermal Pad?

A QFN (Quad Flat No-lead) is a leadless surface-mount package: instead of gull-wing leads, it has metallized lands around the perimeter of its underside and, in the center, a large exposed thermal pad — also called the e-pad, EPAD, or die-attach pad (DAP). That pad does double duty, carrying heat from the die into the board and usually serving as the ground connection. QFN belongs to the bottom-termination component (BTC) family, alongside DFN, SON, and LGA packages, and all of them share the same assembly challenge: the joints sit under the body where you can’t see them.

Engineers reach for QFN because it’s small and electrically quiet. A QFN is roughly 60% smaller than an equivalent QFP, its parasitic lead inductance is under 1 nH (versus 3–5 nH for a leaded package), and the exposed pad lets it shed 1–5 W of heat through the board. That combination is why more than 60% of wireless chipsets now ship in QFN. Packages run from 2×2 mm up to 12×12 mm with 8 to 100 pins. The catch is that none of those advantages show up unless the thermal pad is soldered correctly.

QFN vs QFP: Why Leadless Packages Solder Differently

The cleanest way to understand QFN soldering is to compare it to the package it replaced. A QFP vs QFN comparison comes down to the leads: a QFP has gull-wing leads that splay out from all four sides, forming visible solder fillets, while a QFN has nothing but flat pads underneath. Both are JEDEC MO-220 packages, but they behave very differently on the line.

Attribute	QFN	QFP
Leads	Leadless (bottom pads)	Gull-wing, all four sides
Size (same pin count)	~60% smaller (e.g. 7×7 mm)	Larger (e.g. 10×10 mm+)
Parasitic inductance	< 1 nH	3–5 nH
Thermal path	Exposed pad into the board	Mainly through the leads
Solder joints	Hidden under the body	Visible fillets
Inspection	X-ray required	Visual / optical
Hand-solder & rework	Difficult	Straightforward
Self-centering	Minimal	Some (surface tension)

Here’s the counterintuitive part that catches people. A QFP’s visible leads let surface tension during reflow pull the part into alignment — the package self-centers. A QFN has no perimeter leads to grab, so it effectively rides on the pool of solder under the thermal pad. That’s exactly why too much paste in the center is catastrophic: there’s nothing to counteract it, so the part floats up and tilts, and the tiny perimeter joints lift away. The whole game is balancing center paste against perimeter paste.

QFN Thermal Pad Stencil Design: Windowpane Apertures & Paste Coverage

This is the section that decides your yield. The stencil aperture for the thermal pad must never be a single full-size opening. A 100% deposit traps flux gas under a large molten pool — which voids — and floods the center with solder, which floats the part. Instead, divide the thermal-pad aperture into a windowpane (grid or dot-matrix) pattern that lays down 50–70% paste coverage. The bridges between openings give flux a path to escape during reflow, and the reduced volume keeps the part flat.

The numbers matter here, so design to them rather than eyeballing it. The parameters below are the working envelope for reliable QFN soldering.

Parameter	Recommendation	Notes
Thermal-pad paste coverage	50–70% (up to 80–90% on ENIG)	Windowpane / grid; never 100%
Aperture pattern	Windowpane, grid, or dot-matrix	Gaps between openings vent flux
Grid spacing	1.0–1.5 mm	E.g. 3 mm pad → four 1.2 mm openings, 0.3 mm bridges
Stencil thickness (0.5 mm pitch)	0.125 mm (5 mil)	Laser-cut for clean release
Stencil thickness (0.4 mm pitch)	0.10 mm (4 mil)	Step stencil if mixed with larger parts
Perimeter aperture	Slightly smaller than copper pad	Prevents bridging, allows a fillet
Area ratio	≥ 0.5–0.66	Minimum for clean paste release
Solder paste	Type 4 or finer	Needed for fine-pitch printability

How to Size the Windowpane Grid

Scale the grid to the pad. For a 3 mm × 3 mm thermal pad, a clean starting point is a 2×2 array of 1.2 mm squares separated by 0.3 mm bridges, which lands around 64% coverage. For a 5 mm pad, step up to a 3×3 or 4×4 grid. And here’s the second non-obvious point: surface finish changes how aggressive you can be. ENIG is flat, so paste releases cleanly and you can window up to 80–90% coverage. On HASL, the scalloped, uneven surface causes inconsistent release across the windowpanes, giving you patchy deposits and — counter to intuition — more voids, not fewer. If you’re running QFN on HASL and fighting voids, the finish may be the culprit, not the aperture.

Thermal Vias and Solder Mask for QFN

The exposed pad usually needs thermal vias to carry heat into inner copper layers and the far side of the board. Place them in an array across the pad on a 1.0–1.2 mm grid with a 0.25–0.30 mm drill. But an open via is a problem during reflow: molten solder wicks straight down the barrel, pulling solder out of the joint and leaving a void behind while starving the thermal connection. So plug or tent the vias — fill them or cap them with solder mask — or use via-in-pad with a filled-and-plated process. Thermal-relief spokes on the via lands also slow heat and solder loss.

Solder mask design matters too. For QFN, use non-solder-mask-defined (NSMD) pads with roughly 0.05–0.075 mm of mask pullback from each copper edge; NSMD produces a more consistent fillet and is the IPC-recommended approach for leadless parts. Some designs also add a small venting channel through the mask at the thermal-pad perimeter to give outgassing flux another exit. On the perimeter joints, IPC-7093 calls for about 100–125 µm (4–5 mil) of paste to yield a finished standoff near 50–75 µm — the small gap that lets the joints survive thermal cycling.

QFN Reflow and Inspection: Avoiding Voids and Floating

QFN reflow uses a standard SMT profile, but two adjustments fight the package’s failure modes. A slightly extended soak above 150 °C gives flux time to outgas before the solder paste reaches liquidus, which cuts voiding, and a prebake on moisture-sensitive boards does the same. Keep ramp, soak, and peak controlled — an aggressive spike can flash off volatiles and float the part. QFN essentially requires reflow or hot-air; a plain iron can’t reach the hidden thermal pad, and an ENIG or OSP finish keeps the lands flat for even wetting.

A reliable QFN reflow sequence looks like this:

1. Preheat and soak. Ramp to a 150–180 °C soak and hold to activate flux and drive off volatiles before melting.
2. Reflow to peak. Cross liquidus into the peak window for the alloy (about 245 °C for SAC305), keeping time-above-liquidus in spec.
3. Controlled cooldown. Cool at a steady rate to refine grain structure without thermally shocking the part or the board.
4. X-ray the result. Transmission X-ray reveals thermal-pad voiding and open perimeter joints that no visual check can see.

Inspection is where QFN bites teams used to leaded packages. Because the joints hide under the body, visual and optical inspection only catch edge wetting — everything else needs X-ray. IPC-7093 limits void coverage under the thermal pad to 50% for reliable heat transfer, and for RF or high-speed parts, voids in that ground pad also lengthen the return-current path and raise impedance. Note that a QFN is a BTC with only a bottom fillet required, not a toe fillet, per IPC-A-610 — so don’t reject parts for missing side fillets they were never designed to form.

QFN Soldering Best Practices & Common Mistakes

An RF client once shipped a QFN-based module that passed every bench test, then watched it overheat and drift in the field. X-ray traced it to more than 50% voiding under the thermal pad: the board used a single full-coverage paste aperture on a HASL finish with unplugged vias, so flux gas had nowhere to go and solder wicked down the barrels. Switching to a 60% windowpane aperture, plugged vias, and an ENIG finish dropped voiding under the IPC-7093 limit and the thermal and RF-ground problems disappeared. Almost every QFN defect traces back to one of these habits.

1. Window-pane the thermal pad to 50–70%. Never a single full aperture. The gaps vent flux and the reduced volume keeps the part flat.
2. Plug or tent the thermal vias. Open vias wick solder out of the joint, creating voids and degrading heat transfer. Fill them or cap them with mask.
3. Shrink the perimeter apertures. Make them slightly smaller than the copper pad to prevent bridging on fine pitch while still forming a fillet.
4. Match stencil thickness to pitch. Use 0.125 mm at 0.5 mm pitch and 0.10 mm at 0.4 mm pitch; add a step stencil when a fine-pitch QFN shares a board with larger parts.
5. Spec ENIG or OSP and Type 4 paste. Flat finishes release paste cleanly and enable more aggressive windowing; fine paste prints the small apertures without starvation.
6. Extend the soak and prebake. Give flux time to outgas above 150 °C before liquidus, and prebake moisture-sensitive boards to suppress voiding.
7. X-ray every new QFN design. Hidden joints can’t be inspected visually. Confirm voids stay under 50% and perimeter joints are closed before you release the design.

For quick diagnosis, the common QFN defects and their fixes:

Defect	Cause	Prevention
Thermal-pad voiding	Flux gas trapped under a large paste deposit	Windowpane aperture, extended soak, plug vias
Floating / tilting	Too much thermal-pad paste lifts the part	Reduce to 50–70% coverage
Open perimeter joints	Part floats on center solder, lifting edges	Balance center vs perimeter paste volume
Solder bridging	Excess paste on fine-pitch perimeter	Aperture slightly smaller than the pad
Solder wicking	Open thermal vias pull solder from the joint	Plug or tent vias with solder mask

Frequently Asked Questions About QFN Soldering

What is the exposed pad on a QFN?

The exposed pad is the large metal pad in the center of a QFN’s underside, also called the thermal pad, e-pad, or die-attach pad. It carries heat from the die into the PCB and usually serves as the ground connection, so it must be soldered to a matching board pad with thermal vias.

Why does my QFN float during reflow?

Floating happens when there’s too much solder paste on the thermal pad. The molten pool lifts the part and tilts it, pulling the perimeter joints off their pads. Reduce thermal-pad coverage to 50–70% with a windowpane stencil aperture instead of a single full opening.

What paste coverage should a QFN thermal pad have?

Target 50–70% coverage using a windowpane or grid aperture pattern, never a full 100% deposit. On a flat ENIG finish you can push to 80–90%. The gaps between openings let flux outgas during reflow, which keeps voiding under control and the part flat.

Do you need to plug vias under a QFN?

Yes, in most cases. Open thermal vias wick molten solder down the barrel during reflow, pulling solder out of the joint and creating voids while weakening the thermal path. Plug or tent the vias with solder mask, or use a filled-and-plated via-in-pad process.

Can you hand-solder a QFN?

It’s difficult because the joints are hidden under the package and the thermal pad can’t be reached with an iron tip. Hot-air reflow is the reliable method; for rework, you can drag-solder the perimeter and feed the thermal pad through a via from the back, but results vary.

How do you inspect QFN solder joints?

With X-ray. Because QFN joints sit under the body, visual and optical inspection only catch perimeter edge wetting. Transmission X-ray is the only way to see thermal-pad voiding and open perimeter joints, which is why IPC-7093 ties acceptance to measured void coverage.

What stencil thickness should I use for QFN?

Use a 0.125 mm (5 mil) stencil for standard 0.5 mm pitch QFN and step down to 0.10 mm (4 mil) for 0.4 mm fine pitch. When a fine-pitch QFN shares a board with larger components, a step stencil gives each the right paste volume.

How much voiding is acceptable under a QFN?

IPC-7093 recommends keeping void coverage under the thermal pad at 50% maximum for reliable heat transfer. For RF and high-speed designs, aim lower, since voids in the ground pad lengthen the return path and raise impedance. Verify the actual percentage with X-ray.

Get QFN Soldering Right the First Time

Reliable QFN soldering is mostly upfront design: window-pane the thermal pad to 50–70%, plug the vias, shrink the perimeter apertures, pick a flat finish, and X-ray against the IPC-7093 void limit. Nail those and the package delivers the size, low inductance, and thermal performance you chose it for. If you’d like a second set of eyes, send your Gerber and BOM and we’ll review the QFN land pattern and stencil as part of a DFM review.