Implantable Medical Device PCB: Manufacturing and Design Guide

An implantable device is the most demanding application for a printed circuit board, because it cannot be repaired, it cannot be replaced without surgery, and it operates in an environment that is warm, wet and chemically aggressive. An implantable medical device PCB is therefore built to standards that have more in common with aerospace than with consumer electronics, and almost every decision is made in favour of proven reliability rather than performance or cost.

What Makes the Environment Hostile

The body is a saline environment at roughly thirty seven degrees, which is warm enough to accelerate every chemical process and conductive enough to short anything that is exposed. Water vapour permeates polymers over time, so a coating that appears to seal a board will allow moisture to reach it after a few years. The immune system reacts to foreign materials, so anything in contact with tissue has to be biocompatible and stable, and the device has to tolerate the mechanical loads of movement and, in some cases, impact.

The consequence is that the electronics are enclosed in a hermetic package, usually titanium, and the board inside it is protected by that package rather than by a coating. That single fact shapes the whole design, because the package has to be sealed, tested and connected to the outside world without compromising its integrity.

The Hermetic Package and Feedthroughs

A titanium can is welded shut in an inert atmosphere after the electronics are inside, and the weld is one of the critical processes in the whole product. Every conductor that must reach the outside, whether it is an electrode, a sensor lead or an antenna, passes through a feedthrough: a ceramic or glass insulator brazed into the titanium wall with a conductor through its centre. The feedthrough is a hermetic seal in its own right, and each one is a potential leak path.

Inside the can, the atmosphere is controlled. Helium is often used because it conducts heat better than air and because it allows a leak test to be performed by detecting helium escaping from a suspect unit. That means the board has to be compatible with the fill gas and with the vacuum bake that precedes sealing, and no material in the assembly may outgas in a way that would contaminate the internal atmosphere. Our component tolerance and reliability notes describe how those material choices are assessed.

Materials and Board Construction

The laminate must be stable, low outgassing and proven in long term implantation. Polyimide and certain ceramic substrates are used, and every adhesive, solder mask and coating in the assembly is selected from a qualified list rather than chosen for convenience. Where solder is used, the alloy is chosen for fatigue resistance and for the absence of elements that are unacceptable in an implanted device.

The board is usually small, with a high interconnect density, because the package volume is the limiting factor in how small the device can be. That drives fine lines, small vias and often a high layer count in a very small area. Our blind and buried via article describes the fabrication techniques that make that density possible.

Power, Telemetry and Longevity

Power comes from a battery that has to last years, and in some devices from an energy harvesting source. The battery chemistry is chosen for its stability and for its predictable end of life, and the electronics are designed around an energy budget that leaves little room for waste. That usually means aggressive power management, a low duty cycle and a processor that spends almost all of its time asleep.

Telemetry, where it exists, is usually a short range radio link or an inductive link. Both impose constraints on the board: the antenna has to work inside a titanium can, which is conductive and therefore a barrier to radio, and the inductive link needs a coil whose geometry is set by the package rather than by the electronics. The layout of the radio front end therefore follows the mechanical design rather than the other way round.

Reliability Engineering and Qualification

Reliability is demonstrated rather than assumed. Accelerated life testing exposes devices to elevated temperature and humidity, thermal cycling and, in some cases, accelerated battery discharge, and the results are used to predict behaviour over the intended service life. The prediction carries uncertainty, which is why the design margins are large and why the qualification programme is long.

Every unit is also tested individually. Because a device cannot be recalled from inside a patient, the manufacturing test is exhaustive: functional test at temperature, hermeticity test, and often a burn-in period during which the device operates under load. The board design has to make those tests possible, which means test points, accessible connections and a firmware mode that exercises the hardware without requiring the intended application to be running. Our design release checklist places those considerations in the sequence.

Regulation and Process Control

The regulatory framework is part of the design process rather than a step at the end. Design controls, verification and validation, traceability of every component and documented process changes are all requirements, and they extend to the fabricator and the assembler as much as to the designer. A change of laminate supplier is a change that requires re-verification rather than a substitution.

Process control is the practical expression of that. Soldering profiles, cleaning processes, coating thicknesses and the welding parameters are all specified, monitored and recorded, and any deviation is investigated rather than accepted. In a product where the cost of a failure is measured in a patient’s health, the discipline that this imposes is not bureaucracy; it is the design.

Miniaturisation and Yield

The package volume sets the size of the board, and the board is small enough that the component density is high by the standards of any other application. Fine lines, small vias and thin laminates are all required, and each of them reduces the fabrication yield. When a board costs a great deal and a device cannot be recalled, the yield matters less than the confidence that the boards which pass are good, which is why the process is qualified and the acceptance criteria are tight rather than economical.

The same reasoning applies to assembly. A single unit is often assembled and tested on its own, with each joint inspected and the finished assembly photographed for the record. That is slow and expensive, and it is the correct approach when the alternative is a device that fails after implantation.

FAQ

Why is a coating not enough to protect an implanted board? Because water vapour permeates polymers over the years a device is implanted. The protection comes from a hermetic metal package rather than from a coating on the board.

How does a radio signal get out of a titanium can? It generally does not, unless the can is used as part of the antenna or a feedthrough carries the signal to an external element. This constrains the telemetry design more than the electronics.

What limits the life of an implanted device? The battery, in most cases, followed by the hermetic seal and the feedthroughs. The electronics themselves are usually the most reliable part of the system.