Page 86 - NEW Armstrong Book - 2

Silicon Carbide Adoption
Enters Next Phase
By Orlando Esparza
D
emand continues to grow for silicon carbide tech- nology, which can maximize the efficiency of today’s power systems while simultaneously reducing their
and development tools. Developers looking to future-proof their designs will also need to explore the latest capabilities, such as digital programmable gate drivers that solve earlier implementation issues while enabling system performance “tuning” with a keystroke.
First step: three key tests
A trio of tests provides the data to evaluate SiC device reliability: avalanche capability, the ability to withstand short- circuits, and the reliability of the SiC MOSFET body diode.
Adequate avalanche capability is critical: Even a minor mal- function by a passive may cause transient voltage spikes that exceed rated breakdown voltage, ultimately resulting in failure of the device or, possibly, the entire system. SiC MOSFETs with adequate avalanche capability reduce the need for snubber circuits and extend application lifetimes. The best-rated options demonstrate high unclamped induc- tive switching (UIS) capability of up to 25 joules per square centimeter (J/cm2). These devices show little parametric degradation even after 100,000 cycles of repetitive UIS (RUIS) testing.
The second key test is short-circuit withstand time (SCWT), or the maximum time before device failure under a rail-to- rail short condition. The result should be close to that of IGBTs used in power-conversion applications, most of which have a 5- to 10-μs SCWT. Ensuring sufficient SCWT allows systems the opportunity to service fault conditions without system damage.
A third key metric is forward-voltage stability of the SiC MOSFET’s intrinsic body diode. This can vary substantially from one supplier to another. Without proper device design, processing, and materials, the conductivity of this diode may degrade during operation, leading to an increase in on-state drain-source resistance (RDS(on)). Figure 1 sheds some light on the differences that exist. In a study undertaken by Ohio State University, MOSFETs from three suppliers were evaluated. On one end of the results, all devices from Supplier B showed degradation in forward current, while on the other, no degra-
size, weight, and cost. But SiC solutions are not drop-in replacements for silicon, and they are not all created alike. To realize the promise of SiC technology, developers must care- fully evaluate product and supplier options based on quality, supply, and support, and they must understand how to opti- mize the integration of these disruptive SiC power compo- nents into their end systems.
Growing adoption
SiC technology is on a steep upward adoption curve. Product availability has increased along with a breadth of choices from multiple component suppliers. The market has doubled over the past three years and is projected to grow twentyfold to more than $10 billion in value within the next 10 years. Adoption is extending beyond on-board hybrid and electric vehicle (H/EV) applications to non-automotive power and motor control systems within trains, heavy-duty vehicles, industrial equipment, and the EV charging infrastructure. Aerospace and defense suppliers are also pushing SiC qual- ity and reliability to meet those sectors’ notoriously stringent demands on component ruggedness.
Developers must understand how to optimize the integration of disruptive SiC power components into their end systems.
A key part of a SiC development program is validating SiC device reliability and ruggedness, as these differ greatly among suppliers. With the increasing trend to a total system focus, designers also must evaluate the scope of the sup- plier’s product offering. It is important for designers to work with suppliers that offer flexible solutions such as die, dis- crete, and module options that are backed by global distribu- tion and support, as well as comprehensive design simulation
72
ASPENCORE GUIDE TO SILICON CARBIDE