DfR Solutions Reliability Designed and Delivered

4 Steps to Demonstrating Product Lifetime and Achieving Long-Term Product Reliability

Posted by Craig Hillman on Apr 13, 2017 9:23:00 AM

Long-Term-Product-Reliability.jpgAs a new engineer, you have just been assigned as the lead designer on a revolutionary product for your company. It’s an exciting, career-defining opportunity for advancement. One problem: The customer wants the product to last 20 years. How are you going to meet this requirement? Use DfR Solutions' exclusive four step process to rise to the challenge:

1. Clearly define your product reliability goals

A 20-year product lifetime is actually an incomplete specification from your customer, because the acceptable failure rate after 20 years is not known. Is it 1%? 10%? 50%? Mean time to failure (MTTF) can be a common definition of failure, but a MTTF of 20 years means that 63% of the product will have failed after 20 years — an unacceptable outcome for customers who expect a 20-year lifetime.

2. Establish a realistic worst-case use environment

Realistic worst-case specifically dissuades the design team from choosing either an average or an extreme environment as the appropriate use case. For most market leaders, realistic worst-case typically correlates to the 95th or 98th percentile of the population. For example, in the automotive world, realistic worst-case use is typically represented by a stay-at-home mom who lives in Phoenix, the hottest large city in North America. The key idea is that this use case is not based in an average North American city, but it is also not based on daily driving in the hottest place on Earth, Death Valley.

3. Develop a checklist of what technology can degrade in electronics over a 20-year time period

To determine if these technologies exist within your product design, start with the list DfR Solutions compiled, based on almost two decades of experience helping companies design for long-term performance:

  • Interconnects (wire bonds, die attach, solder joints, vias)
  • Electrolytic Capacitors (electrolyte evaporation, dielectric dissolution)
  • Integrated Circuits (EM, TDDB, HCI, NBTI)
  • Ceramic Capacitors (oxygen vacancy)
  • Film Capacitors
  • Memory Devices (limited write cycles, read times)
  • Light Emitting Diodes (LEDs), Laser Diodes, and Optocouplers
  • Resistors (if improperly derated)
  • Silver-Based Platings (if exposed to corrosive environments)
  • Relays and other Electromechanical Components
  • Connectors (stress relaxation, fretting corrosion)
  • Tin Plating (tin whiskers)

4. Acquaint yourself with the concept of Physics of Failure (PoF)

PoF, a methodology used by the electronics industry since the 1960’s, is the process of using information about a product’s environment, design and materials to predict degradation and eventual failure. PoF does not necessarily require a specific educational background and can be used by all personnel involved in product development including electrical engineers, mechanical engineers, component engineers, test engineers and reliability engineers. And, it can be used throughout the design and validation process, including as early as component selection.

Component Selection - Interconnects

PoF is critical for the initial component selection process. Take interconnects, for example. Due to differences in thermal expansion between electronic components and printed circuit boards (PCB), solder joints are placed under stress every time there is a change in ambient temperature or power dissipation. This stress, over time, will fatigue and eventually crack the solder joints. To assess this risk, the product team should first identify critical component packaging during the component selection process. Primarily, this will be ball grid array (BGA), chip scale package (CSP), quad flatpack no-lead (QFN), land grid array (LGA) and large leadless ceramic devices (crystal oscillators, chip resistors).

When a critical component is identified, the design or component engineering team should request board level reliability test (BLRT) data. Test to failure is preferred, but even test to pass (typically 1000 cycles) is of some value. Once the information is received, a PoF-based lifetime prediction is performed and compared to the supplier test data. One example of a PoF equation for chip component solder joints is shown below:


It may take a few iterations and discussions with the supplier before your prediction matches their test results. But, once everything is correlated, the next step is to re-run the PoF analysis for your specific design constraints and the realistic use-case.

Demonstrating 20-year Lifetime – electrolytic Capacitors

Compared to using PoF for interconnects, the PoF approach for demonstrating reliability of electrolytic capacitors takes a slightly different approach. The reliability of electrolytic capacitors is typically evaluated by doubling the rated lifetime for every 10℃ below the rated temperature. However, there is nothing typical about how some electrolytic capacitor manufacturers complicate this basic calculation by embedding ripple current, temperature rise and voltage into the range of life equations. When different exponents, different constants and even different behaviors (exponential, power law, linear, etc.) are added in, it makes you wonder if they are all building the same technology.

The bigger issue is the meaning of lifetime. Based on extensive testing at DfR Solutions, lifetime can mean anywhere from 1% to 15% failure, depending on the manufacturer. The variability suggests that a more robust PoF-based approach is preferred.

If electrolytic capacitor performance is critical for a 20-year product lifetime, there are a few things to keep in mind. Be aware that some electrolytic capacitor manufacturers will not guarantee lifetimes beyond 15 years. This is, in some cases, driven by aging and hardening of the rubber bung that is used to seal in the liquid electrolyte. There are some ways around this, such as potting the capacitors, and the evidence of this 15-year life limitation is circumstantial.

Any thermal exposure should also be uniform. Hot components adjacent to electrolytics can create temperature and pressure differentials that can accelerate failure and induce catastrophic explosions due to a lack of release through the pressure vent.

If these two design rules are met, then a robust lifetime prediction can be performed using an iterative methodology developed two decades ago by Ford (Gasperi, 1997). The process involves calculating equivalent series resistance (ESR) as a function of temperature, then using this information in combination with ripple current and heat transfer equations to calculate core temperature rise. Once the core temperature rise is known, the vapor pressure can be determined using the following equation:


Vapor pressure will drive the leak rate, which will contribute to loss of electrolyte. Since ESR demonstrates a squared dependence on volume of electrolyte, the rate of ESR increase can be determined, with greater accuracy the smaller the time step.

Demonstrating a 20-year product lifetime can be a daunting task, but not an insurmountable one if you carefully plan and leverage the advantages of a PoF approach and powerful testing tools like Sherlock Automated Design Analysis™ software. To learn more, download Test Plan Development: How to Do It. Click the button below for your free copy.

Test Plan Development: How To Do It

Topics: Test Plan Development