Smart Card Manufacturing Notes
Published: November 16, 2026 | ZOWINDA Application Engineering
A dual-interface card carries two ways in and one chip behind them. The contact plate sits on the surface, the antenna is buried in the body, and a small tail of lead frame called the bridge connects the two. Almost everything difficult about dual-interface production happens in the few millimetres around that bridge, and almost every quality argument about dual-interface cards is really an argument about an assembly tolerance nobody wrote down.
On a contact-only card the module is a lead frame with six or eight contact pads on top and a chip bonded underneath, and the only external connection is the reader. On a contactless card the chip connects to an antenna through two pads. A dual-interface module combines both: the contact plate keeps its ISO 7816 pad layout, the chip behind it drives a radio front end, and two pads on the back of the module, usually designated LA and LB, connect down to the antenna ends buried in the card body. The bridge is that connection, and it is made at implanting time, not at module manufacturing time.
This is why the process sequence dominates the failure population. The cavity is milled to a depth that will leave the module flush, an adhesive or anisotropic film is placed, the module is inserted and cured, and only then is the bridge electrically continuous. If the antenna pads in the body are a tenth of a millimetre off, or the module tail lands on the pad edge rather than in its centre, the joint is marginal in a way that no visual inspection at the implant station will see. The card passes a functional test because the contact is good, ships, and comes back four months later with a contactless failure.
Electrically the bridge is also where two very different domains meet. The contact path is a direct galvanic connection, exposed to whatever a reader slot can deliver, including electrostatic discharge from a human hand. The contactless path is a tuned resonant circuit at 13.56 megahertz, where a few picofarads of stray capacitance at the bridge changes the resonance of the whole antenna. That is why the same card can pass contact testing and fail radio frequency testing: nothing is broken, the joint simply added capacitance. Fixture capacitance during test makes this worse, because a tester that grips the contact plate loads the circuit differently from an antenna in free space.
The mechanical side is equally specific. The bridge sits in the deepest part of the cavity, at the end of a tail that has to bend from the module plane down to the antenna plane, and it is surrounded by cured adhesive that is stiffer than the card body. When the card is bent, that transition is a stress riser: the adhesive holds the module rigid while the body flexes, and the tail takes the difference. A joint that measures a few milliohms on the line can lift after a thousand bend cycles, and the failure mode is an intermittent card rather than a dead one, which is the most expensive kind to diagnose in the field.

Implanting sets the bridge: cavity depth, adhesive volume and module position all move the joint. Equipment: Automatic Milling and Implanting Machine for Smart Card Modules.
It survives or fails on flexure long before it fails on current. Qualify it with bending and torsion cycles followed by resistance measurement, not with a single functional pass.
A card that reads on contact proves the chip works. It says nothing about the antenna, the bridge or the resonant circuit. Dual-interface quality needs contact, radio frequency and resonance measured on the same card.
Where a ferrite sheet stiffens the area around the antenna pads, the bridge sits on a brittle substrate. A cracked ferrite raises losses and detunes the antenna without breaking any electrical connection, so the card tests alive and reads short.
The contact plate is the entry point for discharge during handling and personalisation. Add the discharge protection at the chip, and treat every contact-side test station as an ESD-controlled area, because a damaged input can present as a radio frequency fault.
Typical reference bands for a standard 0.76 millimetre dual-interface card with an etched or printed antenna. Values move with antenna design, ferrite construction, module supplier and card body material, so confirm them on first articles before setting process limits.
The same joint appears wherever a surface contact has to reach a buried conductor inside a laminated body:
Practically no. Removing a cured module from a milled cavity destroys the cavity wall and usually the antenna pads with it, and the body has already been through lamination. Rework of dual-interface cards at module level costs more than the card and adds a hidden population of previously heated bodies. Scrap and record it.
Three common causes, in order of frequency: the bridge never made a reliable joint, the antenna was detuned by the implant or by the body, or the ferrite under the antenna cracked during implanting. Contact testing exercises none of these, which is why both interfaces have to be measured on every unit.
There is no universal requirement, but the design must decide deliberately. Some designs tie the plate to a chip ground for ESD reasons, which introduces a path between the two domains and can shift resonance. Whatever the choice, it belongs in the module specification, not in a supplier default discovered during qualification.
More equipment in this range is listed on the ZOWINDA CardMachine product page, and related application notes appear in our Industry News section.
Related reading:
Contact us for a tailored solution and quotation on inlay production, card testing and personalisation equipment.
Email: info@zowinda.com
WhatsApp: +86 186 2085 0485
We are ready to answer your questions.