Complex electronic products present design engineers with an increased need for cost-effective thermal management.
Today, many designs continually push limits for heat transfer, resulting in high device operating temperatures. PCB materials that increase thermal conductivity to reduce peak operating temperatures can improve component life. For over 30 years, high temperature applications have used polyimide laminate and prepreg systems to provide device reliability in operating environments exceeding 200°C. Newer, lead-free compatible epoxy systems offer a cost-effective approach with requirements that are less critical.
How materials react to cyclic thermal exposure remains a critical contributing factor in determining device reliability. Designing PCBs on materials that reduce or control thermal expansion will improve PTH reliability and reduce stress and fatigue on solder joints to SMT components. Low-loss materials for microwave and high frequency applications minimize heat generated by transmission line loss.
Device reliability is a complex function of the heat generated by the operation of an electronic device, the tools used to dissipate or manage the heat, the thermal stability of the materials used and the environment in which the device is required to operate. Because of diversity of applications and the increasing demand for electronics, diverse thermal management tools have evolved to help mitigate reliability issues. These tools include gap fillers, active cooling systems, heat pipes and heat sinks. While many of these tools overlap in terms of potential benefits, the selection of which tools to apply depends on the ultimate constraints of the device in terms of cost, power requirements, weight, size and reliability.
The breakdown of a typical electronic device demonstrates the various tools used to facilitate heat dissipation when designing for reliability. FIGURE 1 shows a generic device with an active component mounted on a typical circuit board. Complementing these active components may be heat sinks, thermal interface materials, thermal vias and active cooling systems. Many of these approaches are implemented to compensate for the fact that most traditional electronic components and dielectric materials are thermal insulators that necessitate secondary cooling systems, such as heat sinks and cooling fans. Thermal interface materials are used to minimize gaps or variation between materials that can occur in assembly, which can retard heat transfer.
Selection and optimization of thermal management tools are often based on a combination of experience, knowledge and device testing to understand failure mechanisms. Device failure is a function of the reliability of the components, materials, time and operating environment (humidity, temperature, thermal cycling, etc.). In most cases, failures can be grouped into one of several categories, such as component failure, connector/solder joint failure or board failure, to facilitate further root cause analysis. Ultimate causes of these failures may include chemical or electrical degradation of base materials, connection failures caused by thermal expansion mismatches, air gaps causing a reduction in heat transfer and oxidation caused by high temperature or mechanical failures. One cause of temperature related failures of boards or components relates to change or degradation at the molecular level. This type of failure is best modeled as a first order kinetic reaction, typically described as an Arrhenius Equation, which is proportional to the inverse log of the temperature. A simplified Arrhenius equation and a resulting reliability plot against operating temperature are displayed in FIGURE 2. Since failure rates, often described as a mean-time-to-failure (MTTF), increase exponentially with temperature, a 10° C-increase in temperature can double the failure rate. In an operating device where reliability is critical to success, even 1° C can matter. The key to improving reliability is to reduce device temperature by increasing the rate at which heat is removed from the device and from the working area of the PCB immediately adjacent to the device. Understanding heat transfer then becomes the next step.
Heat is generated every time an active device is in operation. Device operating temperature is a result of the balance between heat generation and heat dissipation. Heat itself does not become a problem until there is enough heat to result in an increase in temperature above a critical point, in many cases about 105° F to 120° F. Since many designs are set based on function, the heat generation side of the equation is already determined by the time it comes to managing heat dissipation. As such, it is important to understand the basics of heat transfer to determine possible strategies for reducing device temperature. In simple terms, the whole business of managing heat in a PCB assembly is about preventing the junction temperature from getting high enough to “fry” the active devices.
Heat is moved from a “hot body” to a “cooler body” by one of three basic modes: conduction, convection or radiation. In a PCB assembly, all three are in play to one degree or another. Conduction can be the most effective for heat transfer, where the cooler body is in direct (and preferably intimate) contact with the warmer one, and the heat moves from hot to cold materials in an attempt to reach equilibrium. The rate at which heat is carried from one to the other depends on the thermal (temperature) gradient, the coefficient of heat transfer of materials involved (thermal conductivity), the amount of material involved in the thermal path (thickness), the quality of the interface and to a lesser extent, the heat capacity of the cool body that is absorbing the heat. The combined effects of thermal conductivity, the material thickness in the thermal path and the interface effects on heat flow are often characterized in terms of the thermal impedance.
Convection is the transfer of heat from a hot body to a cooler fluid that carries it away through molecular motion. This can happen naturally in a fluid based on resulting density gradients caused by temperature variation. Convection may be aided by forcing the cooling fluid to flow past the warm body, thus carrying away the heat faster. Conversely, convection heat transfer can be significantly impeded by device enclosures that restrict air flow, resulting in higher device temperatures.
Radiation is the removal of heat from a body by the emission of energy in the form of electromagnetic radiation, which may be in the infrared (heat) or even visible (light) parts of the spectrum depending on the temperature of the radiating body. RF signals, such as those generated by an antenna, are also a type of radiation that dissipates energy.
Thermal Management Design Options
The number of approaches and possible solutions to reduce temperature through heat removal from active devices is almost endless. This remains an active area for development across all application areas, and the tools available to designers will continue to evolve.
Tools for increasing convection include: active cooling such as forced air or conditioned air; water; and vapor cooled. While active systems, such as cooling fans, may be extremely effective in heat removal, they add additional design and power requirements, increase costs and add to device size and weight. They also add to potential reliability concerns, as failures of these systems typically result in device failures.
Tools for increasing conduction include heat sinks, heat spreaders, heat risers and heavy metal backplates. Additional conduction methods involve thermal vias (a PTH used to leverage thermal conductivity of copper), thermal coins (inserts of metal conductors in PCB cut-outs under active components to improve heat transfer), thermal interface materials, thermally conductive adhesives, gap fillers and other ways to create a thermally conductive printed circuit board using conductive materials.
Each of these tools brings their own advantages and disadvantages. Heat sinks and backplates add cost and weight to a system. Design and optimization of these systems requires a consideration of the metallurgy (usually copper, brass or aluminum) to balance heat transfer requirements with material thermal conductivity, heat capacity, density, machinability, processing requirements and costs. Thermal vias can be extremely cost effective, but there are practical limits in the area covered. Heat transfer in thermal vias is limited to thru-plane heat transfer and can cause reliability concerns if the PTH fails due to stresses from thermal expansion. Thermal coins are potentially more cost-effective than a large heat sink and potentially more reliable than thermal vias, but they add to assembly complexity and costs.
Thermal interface materials can be either electrically conductive or electrically insulating and cover a wide variety of materials, such as thermal pastes, greases, phase change materials, tapes, bonding plies and prepregs. Their primary advantage is that they can displace air gaps and reduce impedance to heat flow at interfaces. Some types of materials can also offer thermal-mechanical decoupling to minimize stresses from thermal expansion and to reduce vibration related failures or joint failures and improve reliability through thermal cycling.
Circuit board thermal conductivity is often neglected as a potential area to improve heat transfer since this has historically not been a controllable design option. The primary purpose of these materials is to provide electrical isolation of the various components and traces. This has yielded materials that are also good thermal insulators that trap heat within the active components of the board. By modifying the base properties of these materials, it is possible to create materials that provide not only the electrical insulation required for typical circuit boards, but also improve the heat transfer rates relative to traditional materials. This benefit is demonstrated by comparing a traditional FR-4 material with laminates of increasing thermal conductivity in the same circuit design with a 0.5-Watt heat source.
Design optimization usually begins with thermal modeling tools to identify hot spots or other issues. There are several good software packages on the market today that can help designers understand the affects and trade-offs of the various thermal management tools. This can be time consuming and requires a good understanding of the model limitations to fully translate the results into practice. Unfortunately most modeling software assumes isotropy which may result in over or underestimation of the heat removal. In most PCB materials, the heat transfer coefficient in the perpendicular direction (down through the PCB) is different from that in the plane of the board. The models also rarely account for potential issues, complications or variation in materials, fabrication or assembly. Interfaces interfere with the heat transfer mechanism, through imperfections in conduction. This often results in reduced efficiency in heat transfer which results in excessive temperatures. Addressing these potential issues is generally where thermal testing of subassemblies and devices can help. Such testing can vary from power cycling or single point thermocouple measurements to thermal imaging equipment. This level of testing can help designers further optimize their designs by identifying potential problems or lower cost substitutions, such as trade-offs in materials choice (aluminum vs. copper, cast vs. machined heat sinks, etc.).
The first step towards ensuring long-term survivability of a device in a thermally stressing environment is to select materials that can meet the thermal excursions in the PCB manufacturing and assembly process. IPC specifications for Lead-Free and High Reliability materials include such factors as Tg (glass transition temperature), Td (thermal decomposition temperature), total thermal expansion perpendicular to the plane of the board (Z-axis) and short term stress tests such as T260, T288 and T300. The key to success is to avoid latent defects such as hidden cracks in PTH copper that can later propagate and cause resistance change or can open when the board is at operating temperature.
Arguably, Tg is usually the first consideration in material selection as it relates to the survivability of PTH’s during thermal cycling. A high Tg material will exhibit lower overall Z-expansion from room temperature to the operating temperature and will normally retain its adhesion to copper up to and even somewhat beyond the Tg. Z-Axis thermal expansion from 60° C to 260° C is also a key metric in comparing materials for potential PTH reliability.
Decomposition Temperature (Td) is a determinant in assessing long term survivability of materials at elevated temperatures, and while the temperature of a PCB rarely reaches anything approaching its Td, it is reasonable to say that a material with a high Td is likely more survivable at any long-term temperature exposure than a material with a lower Td. T260, T280 and T300 tests are good measures of the likelihood of a PCB substrate surviving short-term exposure at the extreme thermal limits. These tests are measured by ramping samples to the specified temperature and holding them isothermally until irreversible change occurs, usually in the form of delamination.
Beyond the temperatures observed during PCB fabrication and assembly, the working environment of the finished PCB, the thermal environment in which it will actually live and work, will throw new challenges at the devices and substrates employed in making the board. Any active device on a PCB will generate its own heat as the junctions on its silicon chip flip on and off billions of times per second. When a piece of electronic gear is turned on and off over time, the temperature inside the enclosure will cycle up and down. Beyond these temperatures, electronics today often encounter elevated service times and temperatures in certain difficult environments such as down-hole drilling, under-hood automotive electronics and some space applications. The combination of the environment with the heat generated from the device results in a widely varying thermal operating zone that requires special design consideration to insure device reliability.
High temperature electronics are devices that are designed to operate in severe conditions, such as engine mounted sensors, jet-engine control systems and chemical process or oil-drilling instrumentation. Operating temperatures for these devices can exceed 250° C. These temperatures can start to chemically degrade polymers and lead to device failure. Selecting a thermally robust system that is designed to handle high temperatures for long periods is critical for success. PCD&F
Ed. Note. The complete article including a Chinese translation will be included in the March 2009 edition.
Helena Li is with Arlon-Med and can be reached at This email address is being protected from spambots. You need JavaScript enabled to view it..
