The main challenges the industrial electronics designer faces in realising an ultra-low power Bluetooth design
Bluetooth Low Energy (LE) is gaining support from an increasing number of component and module suppliers. No wonder: this ultra-low power, short-range RF technology will offer a simple means to connect any slave device to, in future, billions of smartphones, tablets and laptop computers.
The fact is, however, that electronics design engineers are already well served by a wide choice of industry-standard, short-range RF technologies, such as ZigBee and Wi-Fi. So does Bluetooth LE have an advantage over these existing standards?
In fact it does, provided the application requires the transmission of short pulses of data and long battery life. But design engineers will only gain the advantages of using Bluetooth LE if they make the right choice of components for their application and market.
Something new, something old
Despite sharing the ‘Bluetooth’ name with Classic Bluetooth, Bluetooth LE is different in many important ways. Classic Bluetooth is designed to provide an always-on wireless pipe for streaming audio and other streamed data, and supports data rates of up to 3 Mbps. It also features a complex protocol stack, required in part to support the complicated methods for establishing and maintaining a connection between two paired devices.
Bluetooth LE has completely different attributes: it features a simple protocol stack which makes and breaks connections extremely fast, enabling short pulses of data to be transmitted intermittently. It achieves very low power consumption – months or years of operation on a single coin cell – because its mode of operation enables the device to be in a deep sleep mode, with the radio switched off, for most of the time.
In other words, Bluetooth LE is not a low-power version of Classic Bluetooth; it performs a completely different function.
It does share the 2.4 GHz frequency band, an antenna and certain elements of the protocol stack with Classic Bluetooth. As a result, manufacturers of Classic Bluetooth chipsets are able to add Bluetooth LE functionality to their devices at almost no cost. And this means that the smartphones, tablets and laptop computers which in the past universally featured Classic Bluetooth will in future universally feature Classic Bluetooth and Bluetooth LE. (In the consumer-facing marketing term, they will be ‘Bluetooth Smart Ready’. Certified Bluetooth LE devices will be ‘Bluetooth Smart’.)
This is the principal appeal of Bluetooth LE to designers of industrial electronics systems: for the first time, a simple, cheap wireless technology provides an interface between battery-powered slave devices and smartphones, tablets and laptops. Applications which are expected to take advantage of the new technology include ‘appcessories’ such as fitness monitors, personal health devices such as heart-rate monitors and pulse oximeters, proximity sensors, for instance in access control systems, and smart watches.
Component companies have been quick to take note of Bluetooth LE’s appeal: stand-alone chipsets and highly integrated modules supporting Bluetooth LE are already available. Part of the appeal of Bluetooth LE is that it is, according to its proponents on the Bluetooth Special Interest Group (SIG), a simple radio which is easy to implement.
This might suggest to industrial electronics designers that their system architecture could be based on a Bluetooth LE chipset. After all, a chipset offers a marked unit cost advantage over the equivalent integrated module. A chipset-based design therefore appears to offer users a competitive edge compared to designers who base their system on a more expensive but more integrated module.
So just how easy is it to implement a Bluetooth LE design from scratch?
Hardware and software design issues
Certainly, manufacturers of Bluetooth LE chipsets provide excellent, highly optimised products. The problem for many OEMs is that they are optimised for one thing above all: low unit cost. This means that they are not optimised for ease of use, simplicity or integration.
Bluetooth LE chipsets are aimed at high-volume products, for which the production runs are counted in the hundreds of thousands or millions of units. The key design constraint for high-volume OEMs is bill-of-materials (BoM) cost. They can afford to throw design resources at implementing a system design using low-cost discrete components, including RF chipsets.
Any chipset sourced from a reputable supplier will, of course, be functionally sound, and the unit cost will be competitive. But a Bluetooth LE implementation with a chipset will present considerable challenges to most electronics OEMs. These challenges fall into the following categories:
- RF system design;
- protocol software design;
- compliance testing and certification.
RF system design
A Bluetooth LE chipset will include a 2.4 GHz transceiver and a baseband controller. This must be integrated into a system design which can reliably communicate over the required range. Key design tasks will therefore include configuring and placing an antenna, routing of connections to and from the RF sub-system, and designing the board layout. This design must take account of sources of interference, and ensure that the radio’s sensitivity is not compromised. This element of the system design calls for deep RF expertise.
Protocol software design
A Bluetooth LE device must run a Bluetooth LE protocol stack, typically on the system’s main microcontroller. Chipset manufacturers normally provide a protocol stack ‘reference design’ free to users: this is not a complete, ready-to-use stack; it should be viewed as a starting point for the user’s own stack design.
Again, stack development calls for specialised embedded software development skills.
Compliance testing and certification
All new RF products are required to undergo exhaustive tests to verify:
- that their RF emissions are at permitted frequencies and power levels;
- that they do not generate interference outside their permitted frequency bands.
Testing carried out by independent laboratories is expensive and time-consuming. Design teams are always exposed to the risk of cost and time overruns should a design not pass its compliance tests first time.
Should the OEM want its device to carry the ‘Bluetooth Smart’ logo, it will also need to undergo independent Bluetooth validation, to verify that it complies with the specifications. Again, it is expensive to design for compliance and to undergo a complete set of Bluetooth Smart certification tests.
(It is possible to use Bluetooth LE technology without certification by the Bluetooth industry body. This may be appropriate for products which are not intended to be marketed to third parties as Bluetooth Smart devices. All new RF products are required to gain regulatory approvals relating to RF emissions and EMI compliance.)
Designers of many industrial and consumer devices which could benefit from Bluetooth LE produce in volumes of hundreds or thousands, rather than millions. Bluetooth LE is undoubtedly a simpler technology to implement than Classic Bluetooth. But the design tasks described above are far from trivial, and entail an investment in design time and design resources which may be out of reach of many OEMs.
So to what extent does the use of a dedicated Bluetooth LE module mitigate the problems involved in chipset-based designs?
Reducing design cost, time and risk
It is possible, by using an integrated Bluetooth module, to eliminate all of the design tasks associated with the use of chipset. How?
First, no RF system design should be required. All of the RF circuitry, including the antenna, is encapsulated within the module (see Figure 1). There is only one RF design constraint: since the module contains the Bluetooth LE antenna, it must not be shielded, so the device’s case should be made of plastic rather than metal. When integrating the module, the hardware design functions are typically limited to providing a power supply and a USB or UART interface to the system’s host microcontroller.
Figure 1: a module such as the RN4020 from Microchip integrates the entire Bluetooth LE RF circuit including the antenna. The RN4020, housed in an encapsulated 10 mm x 17 mm x 2 mm package, includes a ceramic-on-board antenna, 2.4 GHz Bluetooth LE transceiver, a microcontroller with a complete embedded Bluetooth LE protocol stack, and UART and USB interfaces.
In addition, a module will be supplied with a complete Bluetooth LE protocol stack. A highly integrated module will include an embedded microcontroller, the main function of which is to run the protocol stack.
Clearly the stack provides the means by which the Bluetooth LE radio transmissions are controlled. Again, a module should make the control functions simple to implement. Ideally the stack functions will be abstracted out into a familiar standard instruction set such as the AT commands familiar to modem users. This means that the system designer can treat the protocol stack in the module as a black box; very little knowledge of the stack’s operation is required.
Use of a module also eliminates all of the design risk associated with compliance and certification. A module should be supplied ‘pre-certified’. As a stand-alone device in its own right, it will already have all applicable approvals globally in terms of RF emissions.
Provided the OEM designer follows the module manufacturer’s guidelines in relation to input power, layout and enclosure, the complete end product, with the module embedded in it, is guaranteed to pass all required compliance and certification tests (see Figure 2).
Figure 2: A pre-certified Bluetooth LE module eases the certification process, allowing the host product to carry the Bluetooth Smart logo.
The use of a Bluetooth LE module therefore gives OEM design teams the benefits of reduced design cost, design time and design risk.
Balancing the considerations
Each OEM will make its own calculation of the balance of design costs and unit costs: at a certain production volume, the additional unit cost of a module compared to a chipset outweighs the reduced design cost of integrating it in a host system.
In addition, a module provides a fixed set of features, such as support for certain modes of operation or ‘profiles’, defined in the Bluetooth 4.x specification. A minority of applications might find their requirements are not met by the features provided in commercial off-the-shelf Bluetooth LE modules.
Industrial designers, however, should take care not to under-estimate the scale of the RF design, software design and certification tasks involved in implementing a chipset-based design. While Bluetooth LE is a simpler technology than Classic Bluetooth, it is not simple – unless the design is based on an integrated module.
Acal BFi provides design and supply-chain support to OEMs developing Bluetooth LE designs across Europe.
The author Ton Middelman holds the position of Technical Business Development Manager at Acal BFi.
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
