It’s well understood that the number of connected devices will grow exponentially over the coming years.
The most disruptive innovation will see connectivity introduced to traditional non-wireless products, or elements of products, that inherently change how they interact with users and the world around them.
We're delivering high-performance wireless communication across diverse markets
In such cases, a detailed picture of how a device’s connectivity is impacted by external influences is now a critical component of the development process, enabling product designers to mitigate risk and increase commercial viability.
One of the most successful gaming products over recent years has been the Nintendo Switch, which went on to become the fastest selling home video game system of all time in the U.S., selling over 4.8 million units in 10 months during 2017.
The detachable wireless controllers were one of several innovations the Switch brought to the handheld gaming market, however it had to overcome early connectivity challenges to achieve success.
Soon after launch, some users reported outages of wireless connectivity from the Switch’s left-hand Joy-Con controller. Despite being similar in appearance, the left and right Joy-Cons are mechanically different. In the left-hand Joy-Con the joystick is below the buttons, whereas is it above the buttons on the right-hand controller. As the joystick is a significant component it had an impact on the electrical design, but how did this affect performance and was it behind the reported connectivity outages?
We took an original Nintendo Switch into Cambridge Consultants’ Satimo chamber with the aim of understanding why this problem occurred and what could be learnt when developing similar technology in the future. Furthermore, we contrasted results with the latest Switch model to illustrate the extent of improvements made by Nintendo.
Nintendo Switch – Satimo analysis
Satimo is Cambridge Consultants’ anechoic radio and antenna test chamber - one of only a handful of similar facilities in Europe. Within this six-metre cube, we can carry out automated three-dimensional measurement and visualisation of wireless device behaviour and test body-worn technology using body phantoms or real live test subjects. This helps us understand and optimise real-word RF performance of products such as smart implants (including implanted telemetry), connected drug delivery devices, and wearables.
In order to analyse Joy-Con connectivity in normal use, we placed them in a hand phantom model within the chamber. As Satimo allows us to test standard wireless protocols without dismantling the devices, or placing them into a specific test mode, we were able to test the Joy-Cons with a standard wireless connection to a Switch console (whilst playing Zelda).
Figure 1 shows the Equivalent Isotropic Radiated Power (EIRP) for a left-hand Joy-Con from an original Nintendo Switch, whilst Figure 2 shows the same test with the most recent product. The performance improvements made by Nintendo are evident, with the new Joy-Con performing around 3dB higher when averaged across the 2.4GHz Bluetooth channels.
Interestingly, the performance of the original left-hand Joy-Con is noticeably worse when measured in the hand phantom than when measured in free space. This illustrates the importance of testing RF devices in the environment in which they will be used – human tissue is very good at blocking radio signals and detuning antennas. Just because a device works well on the lab bench does not guarantee it will in its intended environment.
The shape of the transmit power may also have affected the performance of the device. The outage feedback from the original Joy-Con only came from a subset of users – for many users they worked fine. Figure 3 illustrates the left controller in relation to the original Switch console. In this position, much of the transmitted power is directed towards the floor, or the player, rather than towards the console. In this case the connection will be dependent on multipath reflections and may perform quite differently depending on the person or room in which it is used. As the controllers can be used in different orientations, wireless performance may also depend on what game is being played.
The Switch and the Joy-Cons are built around standard off-the-shelf Bluetooth components (the Broadcom BCM20734 dual-mode Bluetooth chipset) commonly used in Human Interface Devices (HID) like keyboards and joysticks. Bluetooth HID devices can suffer from unreliability, particularly when connecting or re-connecting, or when their host powers off or goes into low power mode. By contrast, the Nintendo Switch performs excellently in this regard, offering seamless and reliable connectivity when changed from wired to wireless.
However, this performance may come at a cost due to the fact the Joy-Cons rely on a high bandwidth and low latency wireless link. A typical HID device will send small amounts of asynchronous data only when it is used. The Joy-Cons send a constant synchronous stream of data even when they are sitting idle on a bench. This will have an impact on the device battery life and may be the reason the Switch does not support a direct connection to Bluetooth wireless headsets. It is likely that synchronous connection to two Joy-Cons, combined with a synchronous voice connection to a headset, would stretch the Bluetooth capacity beyond its practical limits.
In conclusion, the Satimo analysis has proved the significant impact the human body, local environment and the way in which a device is held can have on connectivity performance. As innovation increases and the way we interact with devices becomes more engaged, early testing of this level will become more and more critical to the product design process.