ARM Cortex-M cores are power efficient cores designed especially for microcontroller use. This interview discusses the benefits of using Cortex-M powered microcontrollers in different applications.
Mr. James Wiart, MCU Marketing Head, Greater China and South Asia Pacific region from STMicroelectronics speaks with Rahul Chopra of EFY.
Q. What is the biggest benefit an engineer gets if he uses processor based on the ARM Cortex-M architecture?
A. Different markets have different needs and the ARM Cortex-M family has been able to serve those different needs very well. For the entry-level market, we are using the Cortex-M0 core, which is delivering higher performance than 8 bit and 16 bit cores. Then we have chipsets using Cortex-M3 and Cortex-M4 cores with 90 nm technologies for high end and high performance product lines. These Cortex-M powered product lines continue to be enriched with better performance and more peripherals. So, the benefit is that we are able to have a truly complete system inside one MCU with these ARM-based processors, something that was not possible three years ago.
Q. How important is compatibility in MCUs, and is there a definite benefit to the engineer?
A. Compatibility is very important for customers. When they design a platform, they want to be able to reuse it. Today those platforms where one can just remove a component and put another one in and control it, will provide us a big saving in development time and overall time to market. While it is true that we have many products, we try to keep the usage extremely simple. We are using only three cores and most of the products are compatible. For instance, the main stream product lines F0, F1, F3 – they are fully compatible, so even if someone selects one product they are still free to go out and change it to another one. Thus the nightmare of being strict with only that particular family is gone, now we have pin-to-pin compatible MCUs, giving full freedom to the designers.
Q. How have modern MCUs revitalised inverter functionality?
A. MCU is a key player in inverters; it monitors the power supply conditions and also controls the charging and health of the batteries. Additionally, whenever there is a “powercut”, it is the MCU that decides and takes care of the switchover to battery supply. The traditional inverter applications worked fine with 8-bit controllers where some monitoring and a few computations were the only work required. But now these applications have advanced and emerged with newer consumer facing features. Now you need a bit more processing power where you can say Cortex-M0 based STM 32 F0 family matches very well to those needs.
Q. Is selecting a processor core the single most important thing in product design?
A. No. The core is just like 10% of a product. So, just because the cores are same the products need not be the same. The technology is different, the peripherals are different, and the way they are integrated is also different.
Q. How has reliability improved in latest MCUs?
A. Crystals are very sensitive to shock, so crystal failures usually occur more often. We have designed some hardware around a clock security system such that even if you remove the crystal, the chip is able to safely switch to the internal IC and the system continues to work with the disconnected crystal. It works so smoothly that you may not even be aware that there has been a failure with the crystal.
Q. How robust can clock free MCUs be?
A. Crystals are the most sensitive part in an MCU. If the product falls on the floor, the crystal is the first to break. If you have no crystal in your product, it accounts for more robustness. The latest MCUs offering USB interface are crystal less, because there is a 48 Mhz oscillator required for the USB, integrated inside the chip. So there is no need for an external resonator crystal. This saves your cost and at the same time allows the product to be integrated on a much smaller PCB. Considering current chip sizes, the crystal can be the twice as the size of MCU, so without the crystal you save a lot of space on the PCB. These devices enable product developers to increase system integration, reduce costs, and exceed the traditional price/performance limitations imposed by older, proprietary 8-bit or 16-bit microcontrollers for USB device or USB controller applications.
Q. How is the USB interface in an MCU used in charge detection applications?
A. USB interfaces in the new STM32 F0 family are compliant to battery charger detection (BCD) specification 1.2. It allows you to detect the charging level of the batteries through USB but it requires a specific class and also specific protection on the pins of the device. Most of the companies cannot do this charging detection class on a simple USB, as additional hardware protection on the chip is required to be able to do it.