Selecting a 32-bit Microcontroller
Selecting a 32-bit Microcontroller That Makes the Developer’s Job Easier
Traditionally, the key factor in selecting a 32-bit microcontroller (MCU) has been the choice of central
processing units (i.e., the core CPU). Until recently, 32-bit MCUs were based on a variety of cores
including, in some cases, proprietary architectures; as a result, embedded designers either had to stick
with one core or significantly increase their development time to learn new hardware intricacies and port
existing software code. Over the last few years, the emergence of ARM Cortex cores in microcontroller
products has changed the embedded landscape. Developers are migrating from proprietary 32-bit cores
to MCUs based on ARM Cortex processors, so they are no longer locked into a single provider for their
MCUs. A large and growing ecosystem has developed around ARM processor-based MCUs including
third-party tools for compilers, real-time operating systems, software stacks, LCD graphics and more.
Most major microcontroller suppliers now have an ARM processor-based offering, making the ARM
Cortex core the de-facto 32-bit microcontroller standard.
Selecting a 32-bit MCU based on a standard core opens the door to more options than ever before.
Therefore, choosing the right MCU for a given application can be a difficult task with many variables to
consider. First, designers must narrow the potential number of microcontroller choices based on a variety
of “checkbox” items, such as memory size, number of input and output pins, and communication
interfaces. It is very likely that multiple suppliers of ARM processor-based MCUs will meet the basic
requirements list. Therefore, developers will need to refine their selection by focusing on other important
factors, such as mixed-signal integration, configurability, power consumption and ease of development.
Selecting a 32-bit microcontroller that integrates commonly used components will help the developer
reduce overall system cost, design complexity and development time. For example, Silicon Labs’
™
Precision32 mixed-signal MCUs are engineered to provide a number of integrated features not typically
found in other MCUs, such as a USB oscillator, 5 V regulator, six programmable high-drive pins that can
provide up to 300 mA of current and 16 capacitive sensing input channels for touch buttons or sliders.
This high level of integration helps to eliminate the need for multiple discrete components and provides a
more flexible power domain, thereby reducing the bill of materials (BOM) and simplifying the development
process.
To illustrate the benefits of using a highly-integrated mixed-signal microcontroller, let’s examine a typical
barcode scanner application. To read a barcode, the scanner shines a laser at an oscillating mirror
powered by a motor (see Figure 1). This light illuminates the barcode, and the image is then captured by
a charge-coupled device (CCD) sensor. The CCD sensor is a camera that captures one line of pixels at a
time, e.g., 1 x 1024 pixels. The analog light levels are shifted out and captured by an analog-to-digital
converter (ADC). Having a microcontroller that can supply high current eliminates the need for power
transistors used to drive the laser and motor. Choosing a microcontroller that is designed to provide an
interface with the CCD sensor for clock synchronization can make a designer’s job easier, as well.
Silicon Laboratories, Inc.
Rev 1.0
1
Figure 1. Barcode Scanner Design Benefits from Highly Integrated MCUs
For the optimal results, the microcontroller’s ADC should be able to keep pace with the fast rate of the
CCD camera (typically >1 MSPS). For 5 V CCD sensors, a power management IC is also necessary in
most designs to provide the input voltage for the sensor, and a 3.3 V input voltage is needed for the MCU
and other components.
In this barcode example, the Precision32 SiM3U1xx USB MCUs can drive a synchronized clock to the
sensor, easily keep up with the fast CCD sampling rate and provide a 3.3 to 5 V dc-dc boost controller to
power the sensor, further reducing the number of components in the system. Furthermore, in USB
powered scanners, the Precision32 MCUs have a built-in voltage regulator to run the microcontroller
directly off of USB power and an internal 48 MHz oscillator with an innovative clock recovery circuit that
locks to the USB signal, providing better than 0.25 percent accuracy and allowing for crystal-less USB
operation. Other integration in barcode scanners can include directly driving a buzzer to alert the user
when a successful scan has been made, capacitive touch buttons to replace mechanical buttons and HW
encryption for wireless scanners that need to send data securely.
Another important design consideration is the flexibility to accommodate change quickly and easily
without driving up development cost. To speed development, designers often start with a previous project
and modify the setting to fit the new requirements. However, to effectively repurpose the design, it is
important to be able to choose and modify the microcontroller peripherals used and their placement. Most
MCUs provide a preset location for peripherals with a fixed alternate selection. The preset pinout often
leads to pin conflicts, forcing developers to change their design or move to a larger, more expensive
package. An alternative dual-crossbar microcontroller architecture patented by Silicon Labs (shown in
Figure 2) gives developers greater flexibility by enabling them to first select the peripherals needed and
then choose the pin placement for them.
Silicon Laboratories, Inc.
Rev 1.0
2
Figure 2: Flexible Dual-Crossbar Architecture Used in Precision32 MCUs
Having the ability to “cherry-pick” only the required peripherals often enables smaller, more cost-effective
packaging options. For example, in a communications hub that required four UARTs with flow control (16
pins) and two SPIs (6 pins), a developer should be able to select a microcontroller with slightly more than
22 I/Os. However, using the standard fixed-architecture, four UARTs and three SPIs would necessitate a
64-pin or even a 100-pin package to achieve the precise peripheral mix. Using a flexible, configurable
crossbar technology, the developer could easily fit this peripheral mix into a 40-pin package with several
I/Os to spare. In addition, by optimizing the peripheral location, developers can place peripherals close to
their connecting circuits, enabling shorter routes and potentially reducing the number of PCB layers
needed for the design. Best of all, last-minute design changes can easily be accommodated in software.
For example, if the communication hub required another IC with an SPI interface, that’s no problem. A
third SPI port can easily be added in the same footprint simply by modifying the software.
Given the benefits of a flexible crossbar architecture, are there any downsides to using a microcontroller
with a highly-configurable crossbar? Some developers are concerned that a crossbar architecture may be
complex to program. Silicon Labs has addressed this concern by creating AppBuilder, a free software
development tool designed to simplify initialization and configuration. The GUI-based AppBuilder tool
enables developers to quickly and graphically choose their peripheral mix, select peripheral properties,
set up clocking modes and customize pinouts, all without having to read the data sheet. AppBuilder even
generates source code that can be used with popular compilers, such as Keil, IAR and GCC.
A final important consideration in selecting a 32-bit microcontroller is power efficiency. In fact, ultra-low
power has become a critical concern in a wide variety of embedded applications. With today’s emphasis
on “going green” and minimizing energy consumption, designers must pay close attention to their total
power budget. There are various methods of significantly reducing energy. The most effective method
depends on the end application. For example, with a blood glucose monitor, the patient only uses the
device a handful of times throughout the day. The vast majority of the time, the device is in a deep sleep
mode. Therefore, in this application, it is most important to minimize the power consumption in sleep
modes.
Silicon Laboratories, Inc.
Rev 1.0
3
A sensor node device, on the other hand, must constantly be monitoring for an event to happen. If it is
constantly monitoring for an event, it must always be in active mode, right? Not really! The sensor node
can be in sleep mode, wake up quickly, determine if an event, such as detecting smoke, is happening,
and then go back to sleep. In this type of system, it is important to have a low-power sleep mode that
includes a real-time clock (RTC) to wake up at a regulator internal, such as every 100 microseconds. It is
also important to have a fast wake-up time and a processor to quickly run the fixed number of commands
to determine if an event is happening.
Some applications, such as equipment on a factory line, never go to sleep. In these applications, having a
microcontroller with low-power active current is very important. In addition, there are other tricks that can
be employed to save power, such as reducing the frequency rate on-the-fly to use only the processing
speed needed for a given task.
It can be challenging to find a 32-bit microcontroller that excels in ultra-low power in sleep mode, active
mode, wake-up time and on-the-fly frequency changes. The Precision32 32-bit microcontroller family
addresses these challenges by providing several low-power options, as shown in Figure 3. The MCUs
can operate below 100 nA and include a brown out detector and 4 kB of RAM retention. The real-time
clock can be enabled for an additional 250 nA of current. There is an analog comparator option for an
additional 400 nA and even a low-power timer and pulse counter. The MCUs can wake up from lowpower sleep mode in microseconds. In addition, the Precision32 MCUs have a very low active mode of
275 µA/MHz, and they feature a sophisticated PLL that can lock to any frequency from 1 - 80 MHz,
enabling the developer to optimize power for the task at hand.
Figure 3: Precision32 MCUs Engineered for Ultra-Low Power in All Modes
At a time when many major microcontroller suppliers have families of 32-bit devices using the same core
with similar memory sizes and large numbers of I/Os and serial peripherals, it is easy for a designer to
think that the choice of MCUs in an embedded design doesn’t really matter. However, by selecting the
right microcontroller for a given design, developers can significantly reduce development time, power
consumption and total system cost while gaining the flexibility to make last-minute changes without a
major system redesign. In short, it makes sense to choose a 32-bit microcontroller with a flexible
architecture engineered from the ground up to make the developer’s job easier.
# # #
Silicon Labs invests in research and development to help our customers differentiate in the market with innovative low-power, small
size, analog-intensive, mixed-signal solutions. Silicon Labs' extensive patent portfolio is a testament to our unique approach and
world-class engineering team. Patent: www.silabs.com/patent-notice
© 2012, Silicon Laboratories Inc. ClockBuilder, DSPLL, Ember, EZMac, EZRadio, EZRadioPRO, EZLink, ISOmodem, Precision32,
ProSLIC, QuickSense, Silicon Laboratories and the Silicon Labs logo are trademarks or registered trademarks of Silicon
Laboratories Inc. ARM and Cortex-M3 are trademarks or registered trademarks of ARM Holdings. ZigBee is a registered trademark
of ZigBee Alliance, Inc. All other product or service names are the property of their respective owners.
Silicon Laboratories, Inc.
Rev 1.0
4
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising