How to Improve PCB Reliability

Estimated reading time: 2 minutes

If you research electronics reliability for the automotive industry, for example, you will find they all address only the electronic components mounted on a PCB. One of the more referenced reliability studies for the industry was conducted by Telcordia Technologies, titled, "Reliability, Prediction Procedure for Electronic Equipment (SR-332)." Like most other electronics reliability studies, it excludes the PCB.

Automotive manufacturers need to have stringent reliability requirements, and thus they have set goals for reducing component defect rates. For example, a defect rate of less than 10 ppm for an engine control unit (ECU) or less than 1 ppm for a component within the ECU. The end game is "zero defect, zero failure."

The automotive industry has quantified the reliability rate for all the various component packages today and report defect rates as follows:

• ASIC—0.2 ppm
• Microprocessors—0.5 ppm
• Inductors—0.2 ppm
• Resistors—0.0 ppm

Component failure rates have steadily declined over the years to the point where non-component failure sources have become the dominant cause of failures for a PCB.

The problem with automotive-electronics reliability studies is they do not consider the PCB. If the reliability of the components is becoming a non-issue, then the only way to improve automotive electronics further is to look at the non-component aspects of the electronics, and a significant one is the PCB.

Each PCB design is a custom component, with its own unique, complex recipe:

• Material types and stackup
• Copper features
• Mechanical processes
• Chemical processes
• Electronic components
• Component mounting methodologies
• Impedance and capacitance concerns

All of the above form dependencies and constraints to every aspect of the PCB. How do you assess reliability for a “component” you have never used before? You measure the elements that make up the component, that is, the structure and features of the assembled PCB, including the solder-paste connection to the components.

First, measure and count the critical features of the PCB to assess the reliability of a custom component:

• Type of PCB—rigid, flex, flex-rigid, packaging substrate
• Size of PCB
• Number of layers
• Number of vias
• Size of vias
• Minimum annular ring
• Microvia stackup
• Number of component packages
• Double-sided SMT
• Range of package sizes
• Copper weight
• Copper distribution
• Test-point density
• Singulation method
• Zero-offset devices
• Gold fingers
• Vias in pad
• Embedded devices
• Aspect ratio

Second, list the electronic component and padstack details to be used to assess the reliability of a PCB assembly:

• Accurate body dimensions, preferably with tolerances
• Component standoff from PCB
• Lead type
• Pitch
• Smaller form packages are more susceptible to design and manufacturing issues
• Land pattern optimized for manufacturing
• TH pin diameter and length
• Solder mask-defined pads

Third, take into account the manufacturing processes used to produce the PCB:

• Conventional or sequential lamination
• Mechanical drill, laser drill or backdrill
• Number of panels stacked for drill
• Solder mask-defined pads
• Route or vscore singulation
• Flying prober or ICT
• Reflow, wave, or selective soldering
• Flow solder direction
• Conveyed edge for assembly
• SMT, auto-insertion, pressfit, or manual placement
• Stepped stencil
• Rework candidate

Manufacturing is a process, and processes also have tolerances. A reliability analysis needs to consider these fabrication tolerances.

To read this entire article, which appeared in the November 2016 issue of SMT Magazine, click here.