Competing with Customers in Power Management

Vendors competing with customers isn't a new concept to the power management market. The first PWM controllers were introduced in an environment rich with discrete implementation, that of the self-oscillating switcher. Early discrete designs were hard-won, high-maintenance “file to fit and paint to match” solutions with performance tuned to the gain of the power switch, characteristics of the power transformer and filter capacitors.

The first PWM controllers introduced in the early 1980s emphasized fixed-frequency operation and classic voltage-mode control, stability and phase compensation. Not surprisingly, early adopters designed for high-rel applications where worst-case analysis and unconditional stability were prime considerations. Consequently, IC manufacturers weren't considered competition for the mainstream power supply manufacturer, who maintained the black art of self-oscillator design.

In the late 1980s, the stage changed dramatically with the advent of current-mode controllers and the influx of ICs from consumer applications — particularly color TV. At that time, a semiconductor manufacturer introduced a basic point of load regulator that rapidly became the benchmark for ease of use. It was supported by software design-automation tools and an array of inductors from passive component manufacturers keen to cash in on any standardization in a notoriously fragmented market.

Even today, more than 60% of the designers using these online tools have little or no formal PSU design experience, underscoring the successful strategy of simplifying the design of switching regulators. This product hasn't antagonized the PSU design community, as it's considered to be at a low level of abstraction. Conversely, two areas of marketing strategy have presented a direct threat to the power supply manufacturer: embedded and direct competition.

Let's discuss the less provocative one first. In the late 1990s, one semiconductor company introduced its “embedded solution” directly to telecom and datacom customers, encouraging them to circumvent module or brick manufacturers who were deemed too expensive. This seed fell on particularly fertile ground with datacom manufacturers, who were conditioned to expect the aggressive dollar-per-Watt benchmark set by the silver box in PCs and servers. To be absolutely fair, this company probably felt that PSU manufacturers were not receptive to their embedded systems expertise “brand,” and the company was not overly concerned about the loss of its business.

This competition reached the ultimate level of intensity when another semiconductor company purchased a modular power supply company in the late 1990s, followed by another smaller semiconductor company that had established enviable technical credentials in the PSU design community.

“What's the downside of this supply chain competition?” you may ask. The straight answer is that it depends on the new product definition strategy of the semiconductor manufacturer. If engagement with PSU designers is a key element in new product definition and evaluation, and competitive threat is an element of their vendor selection criteria, a manufacturer may justifiably exclude a semiconductor vendor from all but tactical commodity discussions.

Semiconductor new product strategy ranges from the technology-driven “Field of Dreams” — if we build it, they will come — to the full custom environment where the customer writes the whole specification, knowledgeable of the IC design and process constraints. Each extreme has its merits, but the best case is arguably in the middle where products are defined in partnership with key “lighthouse” customers who have everything to gain from integration — the elusive win-win scenario.

The timing couldn't be more appropriate. PSU manufacturers are ready to partner with semiconductor suppliers and take advantage of the time-to-market, power density, reliability and competitive barrier that integration presents, particularly in modular dc-dc converters driven by the inexorable proliferation of the information infrastructure.

The challenge is crystal clear: The opportunity to set new standards beckons. Now is not the time to have an identity crisis.