Product Reliability

A vital element of product design that is too often neglected is the attention to field reliability. This begins in the initial design of a circuit / system with Failure Mode Effect Analysis (FMEA), component stress analysis and MTBF predictions.

Areas of Expertise in Reliability Analysis

  • FMEA analysis of assemblies and systems.
  • Component stress analysis of Printed Circuit Assemblies (PCA’s).
  • MTBF analysis of PCA’s and optoelectronic devices using your Bill of Materials.
  • Design of test fixtures and software for life-testing of devices.
  • Determination of component life models based on physics of the device and augmented with lab test data.
  • System reliability prediction using Monte Carlo analysis

Example Projects

  • Spice analysis of an analog PCBA to determine component stress levels relative to rated values.
  • Life test hardware for a Polygon motor assembly using LabJack I/O boards and custom Java software.
  • Life test hardware for laser diodes using custom analog hardware and software.
  • MTBF predictions of a Large Format Display product encompassing multiple modules each of which includes PCBA’s and opto-electronics.
  • PCBA board test system using National Instruments data acquisition hardware and custom software in C++.

Tools

  • Matlab / Simulink
  • Pspice
  • Mintab
  • Weibull analysis
  • Excel

 

The importance of Reliability in design

 

Reliability should never be an afterthought but instead an integral part of product design. Reliability should begin with specifications during the system design. During design worst case analysis is performed to ensure all components are operated within ratings. Following this the reliability of a subsystem can be computed from a BOM to arrive a “parts count” MTBF.

In the early prototype phases of a project components expected to have low MTBF can be tested under accelerated life conditions to determine the wear-out function. If the assemblies do not meet required reliability extrapolated to normal usage conditions they can then be redesigned before production starts.

Manufacturing yields along with field data on failures can provide feedback to the design process.