I have always found it interesting that how you approach a new market has more to do with people than with technology. Take the example of deploying microcontrollers in power-supply applications - “the last bastion of purely analog design.” Thanks to the availability of economical, low pin count microcontrollers (MCUs), digital assist and digital control are now a viable option for various power-supply functions.
However, this wasn’t the case a few short years ago. I can still hear one power-supply designer saying, “Now wait a minute, we are concerned about reliability, cost and size." So, when we walked in asking the power-supply designers to consider using MCUs as a control function in their designs, the designers had strong reservations that using an MCU would violate most, if not all of these three requirements. Many started with, “I don’t know how to calculate the reliability with these MCUs, let alone the firmware. What happens if they have a static hit, EMI, or RFI; what happens to the micro?” Then came conversations regarding the size and cost considerations of the MCU.
The task of persuading power-supply designers to consider MCUs was even harder in patriarchal companies staffed with older, more experienced designers who had been building power supplies for many years. Sales pitches that started with how MCUs in power supplies were the wave of the future, how they would save the engineer cost, board space, and seriously expand the design’s capabilities, held no sway with these designers.
Having designed power supplies from the ground up with MCUs gave us the confidence to try a systems-design approach, which emphasized the natural fit of MCUs in power-supply designs. Further, our experience over the last couple of years has taught us to handle the change more as an exercise in evolution, rather than revolution. Typically, we would start with the proposal of replacing existing expensive power-supply functions with the cheaper, programmable, single-chip functions found in an MCU. As a result, the size and cost of MCUs started making designers’ objections work for the MCU, rather than against it.
To get the MCU foot in the analog door, designers were tasked to revisit the part that they were using for dithering the clock. The ASIC solution they were using was only available in an 8-pin SOIC, but with a 6-pin MCU in a SOT-23 the dithering function could be done in half the board space.
Soon, we broke the barrier and MCUs started to gain acceptance. This allowed us to bring in other circuit functions. For example, using the MCU’s comparator, designers could put in an undervoltage lockout. Then, using a simple software-based PWM, they could implement a soft-start function to the power supply. And so on it went throughout the design.
Once designers had tried an MCU in their power supplies and discovered its advantages, they realized that a few lines of code on a tiny 8-bit MCU enabled them to replace the circuitry they were originally doing with diodes, transistors and jury-rigged logic.
From there, it was only a short step for these astute designers to discover that the MCU can monitor temperature, making it easy to implement a temperature-compensation function. Using software-based timers in place of RC delays, they could do the power sequencing as well. Just like that, MCUs started picking-up little bits and pieces in the design. Suddenly, power-supply designers found that MCUs in power supplies were reliable. By using watchdog timers and brownout resets, they were able to build extra safeguards into their power-supply designs.
In replacing existing hardware functions, designers realized that the MCU-based power-supply design was actually smaller and less expensive! Going about it in this way, MCUs slowly grew on designers—gradually absorbing other parts and implementing the functions needed in a more cost-effective and compact form. Somewhere along the way, completely unnoticed, the hype and mistrust evaporated and designers added MCU design and programming skills to their power-supply expertise.
Keith Curtis is Principal Applications Engineer with Microchip Technology’s Security, Microcontroller and Technology Development Division, where he is responsible for developing training and reference designs for incorporating microcontrollers into intelligent power-supply designs. Curtis has a BSEE from Montana State University and over twenty years of experience in electronic design engineering and management. He sits on the PMBus development committee and is chair of the PMBus development tools subcommittee.