Power Electronics

Modular Approach Simplifies Power-System Design

A new generation of integrated and intelligent power modules is reducing the time needed to design, assemble and deliver power-conversion and drive systems.

For the PDF version of this article, click here.

The introduction of the first ICs in the early 1960s greatly simplified the design and manufacturability of complex electronic circuits by combining many of the complex functional blocks on a monolithic semiconductor chip. Similarly, a new generation of highly integrated and intelligent power modules can be used as basic building blocks to greatly simplify the design and reduce the time to market of a new power-system design.

If we consider the market for power-electronic modules, it becomes evident that motor drives account for a significant segment (approximately 54% in 2004). A typical industrial motor drive can be functionally dissected into four blocks: inverter, front end/converter, filters, electronics and cooling (Fig. 1).

Generally, the selection of power-electronic devices will impact the design of all other functional blocks of the system. Most modern low- to medium-power industrial drives use insulated gate bipolar transistor (IGBT) or metal-oxide semiconductor field effect transistor (MOSFET) devices — the choice depends on voltage rating requirements for the switching devices — for the inverter stage. These drives also employ a line-commutated diode bridge rectifier front end and a chopper/resistor grid combination to dissipate energy during regeneration as heat.

Besides the wasted energy during braking, there also is the disadvantage of not having any control over the dc link voltage. Also, the drives have to use large capacitor banks to “smooth” the dc link voltage. That's because the ripple frequency due to the rectifier tends to be relatively low and depends on the line frequency and number of diode bridge phases. For example, if the line frequency is three-phase 60 Hz, and if the system uses a three-phase diode bridge rectifier, then the ripple frequency will be 360 Hz. Some systems may use 6-, 12- or 18-diode bridges to improve filter design.

A line-regenerative drive can improve on the diode rectifier front end by employing a 6-pulse SCR/thyristor bridge for rectification and another antiparallel thyristor bridge for regenerative feedback through a transformer. Such a system will allow power to flow back to the line (full four-quadrant operation), eliminating or greatly reducing the need for the brake resistors and improving system efficiency. With the increase in popularity of distributed generation, such front ends are especially useful with wind turbines.

This topology provides several advantages when compared to the diode bridge rectifier front end, including four-quadrant operation and control over the dc link voltage. However, this approach has disadvantages such as the addition of the antiparallel thyristor bridge and the requirement for bulky filter components due to the use of line-commutating devices. Another drawback is the need for expensive compensation equipment to maintain power quality of the regenerative feedback. The large size and weight of the system can pose a challenge in some applications.

Replacing the SCR/thyristor converters with IGBT- or MOSFET-based active converters, also known as active front ends (AFEs), will provide all the advantages of a four-quadrant regenerative drive while virtually eliminating harmonic currents and improving the power quality without the need for expensive compensation equipment. AFEs also improve system efficiency and dynamic behavior of the load, as well as eliminate the need for the antiparallel converter or brake resistors.

The active converters also significantly reduce the size and weight of the system when compared to systems based on line-commutated devices. The ability to switch the devices independent of the line frequency, typically between 15 kHz and 20 kHz, makes the filtering components smaller, lighter and less expensive.

Although the use of AFEs has numerous advantages, it also adds cost and complexity to the system. Integrated power modules can offset these penalties by intelligently integrating many of the hardware components needed for both the inverter and the active converter into a functional “black box” such that the same module can be used for either function.

Integrated Power Modules

Power modules took their familiar plastic package form when the SEMI-PACK was first introduced to the market in 1975. Since then, the packaging technology has evolved by integrating more and more components into the module. The module package initially combined discrete switching devices to form a half bridge, then evolved to include H-bridge circuits, six-packs and, more recently, gate drives and sensors.

Integrated lines of intelligent power modules (IPMs), such as the Semikron Advanced Integration (SKAI) modules, have also recently come on the market (Fig. 2). These modules integrate all the elements of a drive shown in Fig. 3. With this approach, dc-link filter capacitors, current and temperature sensors, gate drivers, heatsink and the digital-signaling-processing (DSP) controller are combined into a single highly optimized module.

Some novel packaging and assembly techniques result in higher reliability within the SKAI modules. For example, pressure contact technology eliminates the large solder interfaces typically seen between the direct-bonded copper (DBC) and the module base plate. The pressure system also reduces the number of wire bonds in the power module by using spring pins to contact the gate and sensor connections from the DBCs to the driver/control board. Greater reliability also is achieved by using an aluminum-nitrite (AlN) substrate rather than the less-expensive alumina (Al2O3).

Although these modules are highly integrated, off-the-shelf products, the underlying technology lends itself to the development of several standard configurations. For example, the modules can be assembled on standard air-cooled or liquid-cooled heatsinks (Fig. 2). Other options include additional output current sensors and a wide selection of switching devices (75/100/150-V MOSFETs or 600/1200-V IGBTs) to populate the module.

Building Blocks

New IPMs allow designers to use a building-block approach when designing power systems, with software and wiring defining the function of each identical block. We can conceptually see how the SKAI can simplify the design of drives with AFE by placing two identical SKAIs back to back (Fig. 4). The module function is configured when the customer downloads the control software for the AFE and the inverter into the respective modules. The addition of filtering components and hook-up hardware such as plumbing and wiring completes the system.

Other uses of the IPM show the versatility of the module for different applications (Fig. 5). The top diagram is an application where the source is a battery and one of the IPMs is being used as an H-bridge to drive the primary side of a transformer. In the bottom application (Fig. 5), two IPMs are configured as upper and lower halves of an H-bridge to drive a three-phase switched reluctance motor (SRM).

In both of these applications, the IPM-based solution is not optimal, as one-third of the switches are not used in the H-bridge configuration and half the switches are not used in the SRM drive application. However, the IPMs offer cost savings due to the commonality of the hardware and short development time for the system.

Additionally, the fact that the module integrates many of the subcomponents means that the user is getting a fully integrated and tested package — a package that is already qualified to meet some of the stringent automotive specifications. This can significantly simplify the mechanical packaging and qualification of the target application.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.