How to Develop a Sellable Bluetooth Low-Energy (BLE) Product

Have an idea you’d like to bring to market? In this series, John Teel walks through the process of scaling up from prototype to production. Follow each installment for a closer look at how to incorporate individual components.

If you plan to develop a new electronic product that needs to wirelessly transmit and receive small amounts of data then there is no better solution than Bluetooth Low-Energy (BLE). Although not generally fast enough for streaming audio or video, or transferring large amounts of data, BLE is perfect for applications needing moderate levels of data transmission.

Don’t be fooled by the term “low-energy”, because it’s a great wireless solution even when battery life isn’t critical. Bluetooth Low-Energy (also called Bluetooth Smart) is the easiest and lowest cost wireless functionality to implement in a new product.

Unlike Bluetooth Classic, Bluetooth Low-Energy is a single chip solution. BLE solutions use what is known as a System-on-a-Chip (SoC). A BLE chip solution includes both the RF transceiver and the microcontroller running the Bluetooth stack (firmware) all in a single chip. Bluetooth Classic on the other hand is a more complex two-chip solution: a transceiver chip and a separate microcontroller chip.

BLE is not only a less complex wireless protocol to implement, but it’s also considerably cheaper. One of the biggest development costs for Bluetooth Classic is the cost of the software stack which must be purchased separately. This stack cost is usually at least $10,000 USD upfront, plus a per unit licensing fee for every unit sold.

On the other hand, with BLE the stack is already built in to the SoC so your only cost is the chip itself. A BLE chip only costs about $1 to $2 USD (in volumes of at least 1k units). Plan on another $1 to $2 for all the support components (crystal, capacitors, antenna, etc.) required by the BLE chip.

Module versus Chip Solution

For most wireless technologies it’s best to begin with using a module solution instead of a custom discrete chip solution. A Bluetooth LE module saves you the large upfront costs required for FCC certification.

However, the cost savings of using a module for BLE aren’t quite as drastic as for other wireless protocols. This is because for Bluetooth LE the stack is already built into the chip so there is no extra stack cost. Nonetheless bypassing FCC certification (or at least the most expensive part called intentional radiator) is still a significant advantage for using a module solution for BLE. Modules also eliminate the need for antenna tuning saving you a couple thousand dollars.

For Bluetooth Classic the cost of a module solution is usually significantly more than a discrete chip solution. However, most developers are initially willing to absorb an inflated unit cost in order to reduce their upfront costs.

One additional advantage of BLE is that a module costs only a few dollars more than a discrete chip solution. In fact, for most products you’ll need volumes of at least 500k units before it becomes more economical to use a chip over a module.

Once you reach volumes that high and decide to migrate to a chip solution it’s a fairly straightforward design change especially if you use the same chip used on your module.

The table below shows several of the most popular BLE chip solutions.

Antenna Design

If you’re not using a module with a built-in antenna, one of the most critical aspects of your design will be the antenna. For product such as this there are generally two types of antennas available: a ceramic antenna and a trace antenna.

The advantages of a ceramic antenna are the smaller size and simplified tuning. All wireless devices need to have the antenna tuned for maximum performance (maximum range). Tuning is a fairly complicated process that requires very expensive equipment so tuning is generally outsourced to companies specializing in RF tuning.

An antenna circuit will usually use a pi-network for matching (i.e tuning) the antenna. The value of the capacitors and inductors used in the pi-network is tweaked to optimize antenna tuning. If range isn’t super critical for the application, then a chip antenna doesn’t necessarily require tuning for BLE, at least for early testing.

A trace antenna is designed into the Printed Circuit Board (PCB) itself. The primary advantage of a PCB trace antenna is reduced cost. In fact, since the antenna is simply a PCB trace the antenna is basically free.

However, chip antennas are quite cheap, so I generally recommend chip antennas at least initially. Once the product reaches really high manufacturing volumes then you may consider replacing the chip antenna with a PCB trace antenna to lower your costs and increase your profit margins.

PCB antennas are more complicated to tune and may require multiple PCB revisions to optimize tuning. Changes to the PCB can also have a more significant impact on a trace antenna’s performance compared to a chip antenna.

Summary

Bluetooth Low-Energy is a fantastic technology. This is mostly because of it’s simplicity, and of course it’s super low power usage. So for your product be sure to consider BLE before any other more complex wireless technologies.

I’ve created a detailed spreadsheet comparing eight Bluetooth Low-Energy chips currently on the market. It also includes the modules available based on each chip and estimated prices. You can download it for free.