QFP coplanarity is the vertical deviation of the highest lead tip from the seating plane formed by the three lowest lead tips. Fine-pitch parts are usually held to 0.08 mm; 0.65 mm pitch and coarser to 0.10 mm. Exceed the limit and the failure mode is an open joint on a corner lead, not a short. What follows is where that number comes from, how it is measured, what drives leads out of spec, and how to design the margin back in.

[IMAGE 1: Side view of a 0.5 mm pitch LQFP on a land pattern, with the seating plane drawn through the three lowest lead tips and the deviation of one high lead dimensioned | alt: “QFP coplanarity measured as lead tip deviation from the seating plane”]

Key Takeaways

• The seating plane is set by three lead tips, and the triangle they form must enclose the projected center of gravity of the package (JESD22-B108A, clause 3.1).
• 0.08 mm max is the value printed on most JEDEC MS-026 LQFP and TQFP outline drawings. A 0.5 mm pitch part measuring 0.10 mm is a reject.
• QFP coplanarity is a room-temperature number. Warpage at reflow temperature is a separate test, JESD22-B112, and the two do not correlate.
• Stencil foil thickness is the main design lever, because printed paste height is what absorbs lead height variation.
• After reflow the workmanship limit is far looser: NASA-STD-8739.2 permits 0.26 mm (0.010 in) between any part of the lead foot and the pad, provided a heel fillet is present.

What QFP Coplanarity Actually Means

QFP coplanarity is the largest vertical distance between any lead contact point and the package seating plane, the plane defined by the three lowest lead tips whose triangle encloses the package center of gravity. It is a single worst-case number per device, measured with no external force applied, at room temperature.

Two consequences follow from that definition, and both get missed.

First, one lead condemns the part. JESD22-B108A states that any device with one or more terminals exceeding the specified deviation constitutes a failure. There is no averaging and no percentage allowance.

Second, the number is relative. It describes lead-to-lead variation, not how the body sits. A device can pass at 0.07 mm and still land tilted, because tilt is shared by the whole lead field and partly cancels in the seating-plane calculation.

Scale matters here. On the Analog Devices ST-48 LQFP outline (JEDEC MS-026-BBC, drawing rev 01-17-2018-A), pitch is 0.50 mm BSC, lead thickness runs 0.09 to 0.20 mm, foot length 0.45 to 0.75 mm, and the standoff is 0.05 mm minimum, 0.10 mm nominal, 0.15 mm maximum. The coplanarity note on that same drawing reads 0.08 MAX. A worst-case in-spec lead can therefore sit nearly as high as the nominal standoff of the package.

The QFP Coplanarity Limit That Applies to Your Part

The limit tightens with pitch and with pin count, because both reduce the compliance available per lead. The values below are taken from published supplier specifications and outline drawings rather than from rules of thumb.

Package class	Max coplanarity	Basis
QFP, lead pitch 0.65 mm or greater	0.10 mm (0.004 in)	Altera Device Package Information, ver. 9.1, Table 15
QFP, lead pitch 0.5 mm	0.08 mm (0.003 in)	Altera Device Package Information, ver. 9.1, Table 15
QFP, 208 pins or more	0.08 mm (0.003 in)	Altera Device Package Information, ver. 9.1, Table 15
LQFP / TQFP to JEDEC MS-026	0.08 mm max	Coplanarity note on the outline drawing (e.g. ADI ST-48, MS-026-BBC)
PLCC, J-lead	0.10 mm (0.004 in)	Altera Device Package Information, ver. 9.1, Table 15
JLCC, ceramic J-lead	0.15 mm (0.006 in)	Altera Device Package Information, ver. 9.1, Table 15
BGA and FineLine BGA, for comparison	0.20 mm (0.008 in)	Altera Device Package Information, ver. 9.1, Table 15
Post-solder, lead foot to pad	0.26 mm (0.010 in)	NASA-STD-8739.2, clauses 7.1 and 12.9.2.b.3

Table 1. Maximum lead coplanarity by package class

One point deserves emphasis because it causes real supplier disputes: JESD22-B108A does not set a limit. Clause 7 puts the number of parts, the acceptance number, the maximum deviation from coplanarity, and the preferred measurement method into the applicable procurement document. The standard defines the method. Your purchase specification, or the supplier package spec, defines the pass value. If a distributor cannot produce that document for a part you are second-sourcing, you have no agreed limit to argue from.

How QFP Coplanarity Is Measured

JESD22-B108A permits two methods. Both are usually programmed into the same laser or optical measurement system, and both are reported to a resolution the equipment must support: it has to be accurate to within ±10% of the specified deviation. For a 0.08 mm limit, that is ±8 µm of instrument error.

Seating Plane Method

The device is measured with no external force applied, preferably in the dead-bug (leads-up) position. Every terminal apex is located, then the three apexes furthest from the package substrate are taken as candidates for the seating plane.

The triangle those three form must enclose the projection of the center of gravity. If it does not, the next-highest terminal is substituted and the test repeats. Where several valid seating planes exist, the standard requires the one giving the worst-case measurement. Deviation is then the largest distance from that plane to any terminal apex.

Regression Plane Method

A least-squares best-fit plane is computed through all terminal apexes, then offset parallel to the apex furthest from the substrate. Deviation is the distance from that offset plane to the terminal furthest from it. JESD22-B108A notes the regression plane may be used to emulate package coplanarity during reflow soldering, which is why suppliers of large, thin, warpage-prone bodies often prefer it.

Traditionally the seating plane method is preferred, and the regression plane method is acceptable only where its results correlate to seating plane results. If your incoming inspection and your supplier report different methods, the numbers are not directly comparable.

Where the Measurement Misleads You

The seating plane algorithm has a known weakness, documented in the prior art behind US 6,713,311: when the three plane-defining leads sit close to one another, or when the package body is warped, a device can pass the bench measurement and still fail on the board. The center-of-gravity rule in B108A reduces that failure mode but does not remove it.

The larger trap is temperature. Coplanarity is measured at room temperature. Package deformation at 245 °C to 260 °C is characterised under JESD22-B112 instead. A thin 20 mm or 24 mm body can be well inside 0.08 mm on the bench and still lift its corners during ramp-up.

What Drives QFP Leads Out of Spec

• Trim and form. Leads are singulated and formed after test. Die and tooling wear shows up first as a systematic offset on one side of the package, not as random scatter.
• Test and burn-in contact. Socket pins bear on the lead feet. Repeated insertion is a common source of coplanarity drift on parts that have been retested.
• Handling media. Trays hold the body and leave leads clear, which is why they dominate for large and fine-pitch QFPs. Carrier tape can transmit mechanical stress into gull-wing leads, and tape-and-reel carrier designs include a pedestal specifically to keep leads off the pocket floor.
• Shock and stacking. Dropped dry packs and over-stacked trays deform leads on the outermost devices, so a lot can look mixed rather than uniformly bad.
• Recirculated rejects. Parts that a placement machine rejected for coplanarity and an operator returned to the tray are the single most common cause of a recurring, low-rate defect.

The Defects You Actually See

Out-of-spec leads produce a characteristic signature: opens clustered at corners and along one edge, with the rest of the device wetting normally. Corner leads sit furthest from the seating-plane triangle, so they carry the largest deviation.

Expect the following, roughly in order of frequency:

• Missing or shrunken heel fillet on individual leads. NASA-STD-8739.2 makes the heel fillet mandatory with a positive contour, so this is a reject even where continuity passes.
• Open joints that pass in-circuit test through an adjacent bridge or through a parallel path, then fail in the field on thermal cycling.
• Reduced end joint width. The workmanship floor is 75% of lead width; a lead resting on the very top of a collapsed deposit rarely reaches it.
• Placement machine rejects, which cost throughput rather than yield and are the cheapest place to catch the problem.

Worth knowing before you blame the component: of the three most common QFP defects illustrated in NXP application note AN4388 rev. 2.0, only one is a lead problem. The other two are a solder-mask blister lifting a lead off its pad, and foreign matter under the package body lifting the corner leads. Both look exactly like a coplanarity failure on an AOI image. Check the bare board and the pick nozzle before you quarantine the reel.

[IMAGE 2: AOI image of a fine-pitch QFP edge showing three leads with no heel fillet alongside correctly wetted neighbours | alt: “Open QFP leads with missing heel fillets caused by out-of-spec coplanarity”]

Designing the Margin Into the Board

Printed solder paste is the compliance in the system. The lead field does not flatten; the paste deforms around it. Both Toshiba and NXP state this explicitly in their mounting guidance: for packages without an exposed pad, metal mask thickness is determined by lead pitch and coplanarity, and for exposed-pad parts the package standoff enters the calculation as well. NXP recommends 0.13 to 0.20 mm stencil thickness depending on pitch, at a nominal 0.1 mm standoff.

Lead pitch	PCB pad width	Typical foil thickness	Applicable coplanarity limit
0.40 mm	0.26 mm	0.12 to 0.13 mm	0.08 mm
0.50 mm	0.30 mm	0.12 to 0.13 mm	0.08 mm
0.65 mm	0.38 mm	0.15 to 0.20 mm	0.10 mm
0.80 mm	0.50 mm	0.15 to 0.20 mm	0.10 mm

Table 2. Land and stencil parameters by QFP lead pitch

Pad widths above are NXP recommendations from AN4388 and work out to roughly 60% of pitch. Pad length comes from the drawing, not from the pitch: take the nominal foot length and add 0.5 mm at the heel and 0.5 mm at the toe. For a 14 × 14 mm QFP at 0.65 mm pitch with a nominal 0.80 mm foot, that gives a 1.80 mm pad. Keep pads out from under the body.

Worked Example: 0.5 mm Pitch LQFP

Take a 144-lead 0.5 mm pitch LQFP with a 0.08 mm coplanarity limit, printed through a 0.13 mm laser-cut foil.

• Wet deposit height at 88% transfer efficiency is roughly 0.115 mm.
• Placement pushes the low leads into the deposit. Assume the paste compresses to about 0.07 mm under the seated leads.
• A worst-case in-spec lead sits 0.08 mm above the seating plane, so it barely reaches the compressed paste surface. That is not an accident; the limit is set close to the available paste height.
• Solder paste is roughly half flux by volume, so on reflow the deposit collapses to on the order of half its printed height. Final solder thickness under a seated lead lands near 0.05 to 0.06 mm.

The conclusion is uncomfortable but useful: at 0.5 mm pitch there is essentially no margin above the spec limit. A lead at 0.12 mm is not marginal, it is open. Dropping to a 0.10 mm foil to fix bridging on an adjacent 0.4 mm pitch part removes roughly 20 µm of the compliance the QFP was relying on, which is why bridging fixes so often produce opens two weeks later. If both pitches share the board, step the foil rather than thinning it globally.

Screening, Inspection, and Rework

Incoming QFP coplanarity screening to JESD22-B108A is worth specifying only when you have a history with the part or the supplier. Laser triangulation systems map every lead apex and compute both planes; the throughput cost is real, and a sample plan with a stated acceptance number belongs in the purchase spec, not in an email.

On the line, the placement machine is the practical gate. Do not widen its coplanarity window to clear a jam. Reject to a bin that operators cannot feed back into the tray, since recirculated rejects defeat every upstream control.

After reflow, NXP recommends optical and X-ray inspection to verify opens and shorts. Judge against workmanship criteria rather than against the component spec: a heel fillet must be present with a positive contour, and non-planarity between any part of the lead foot and the pad must not exceed 0.26 mm per NASA-STD-8739.2. That is more than three times the incoming component limit, which is why parts that were marginal on the bench often still produce acceptable joints.

For rework, use a mini stencil with the same foil thickness and aperture pattern as the production stencil. Bake before removal at 125 °C for 16 to 24 hours, or 95 °C for 16 to 24 hours where temperature-sensitive parts are on the board. Keep site redressing below 245 °C to avoid lifting pads. Do not reuse the removed device: supplier qualification typically covers three reflow passes total, one of which is the rework pass you just spent.

Straightening bent leads restores geometry but work-hardens the material at the bend. It is defensible for prototype and low-volume repair with a forming template and magnification. It is not a production control.

Frequently Asked Questions

What is the standard coplanarity for a QFP?

Most JEDEC MS-026 LQFP and TQFP outline drawings carry a 0.08 mm maximum coplanarity note. Suppliers commonly apply 0.10 mm to QFPs at 0.65 mm pitch and coarser, and 0.08 mm at 0.5 mm pitch or 208 pins and above. There is no single universal value; the applicable limit comes from the package drawing or the procurement specification.

How do you measure coplanarity of QFP leads?

Per JESD22-B108A, using either the seating plane or the regression plane method, with no external force on the device and preferably leads-up. Laser triangulation systems locate every lead apex and compute the deviation automatically. Equipment must be accurate to within ±10% of the specified deviation.

What causes coplanarity issues with QFP packages?

Lead forming and tooling wear, test and burn-in socket contact, shock in shipping, over-stacked trays, and carrier tape stress on gull-wing leads. In practice the most common cause of a recurring low-rate defect is machine-rejected parts being returned to the tray and placed on a later board.

Is coplanarity the same as package warpage?

No. Coplanarity is lead tip deviation measured at room temperature under JESD22-B108. Warpage is body deformation measured at elevated temperature under JESD22-B112. A package can pass coplanarity on the bench and still lift its corner leads during reflow, which is why thin, large-body QFPs need the warpage data as well.

Can you fix a bent QFP lead?

Dimensionally yes, with a forming template, fine tooling, and magnification, working from coarse to fine adjustment. The bend work-hardens, so fatigue life drops. Treat it as a prototype or field-repair technique. For production volume, scrap the device rather than carry an unquantified reliability risk into the field.

What to Do Next

If you are specifying a part, pull the outline drawing and read the coplanarity note before you commit the footprint, then put the limit, the sample plan, and the measurement method into the purchase specification. If you are debugging opens on corner leads, measure the stencil foil and the printed deposit height first, because the paste is where the margin lives, and check the bare board for mask blisters before blaming the reel.

And if the part is a large, thin body at 0.5 mm pitch or finer and the opens track the reflow profile rather than the lot code, stop measuring QFP coplanarity. Ask the supplier for JESD22-B112 warpage data instead.

Sources

• JEDEC JESD22-B108A, Coplanarity Test for Surface-Mount Semiconductor Devices, January 2003.
• JEDEC JESD22-B112, Package Warpage Measurement of Surface-Mount Integrated Circuits at Elevated Temperature.
• Altera Device Package Information, data sheet A-DS-PKG-09.1, May 2001, Table 15.
• Analog Devices package outline ST-48, 48-lead LQFP, JEDEC MS-026-BBC, drawing rev 01-17-2018-A.
• NXP (Freescale) AN4388, Quad Flat Package (QFP), rev. 2.0, February 2014.
• Toshiba Package Mounting Guide SOP/QFP, rev. 1.0, 17 March 2016.
• NASA-STD-8739.2, via NASA Workmanship Standards Pictorial Reference, section 7.07, Gull-Wing / L-Leaded Packages.
• US Patent 6,713,311, Method for screening semiconductor devices for contact coplanarity.
• IPC-7525, Stencil Design Guidelines, and IPC-7351B, Generic Requirements for Surface Mount Design and Land Pattern Standards.