The search for a simpler, more cost-effective way to implement USB 3.0 functionality

October 09, 2015

The search for a simpler, more cost-effective way to implement USB 3.0 functionality

The USB 3.0 SuperSpeed standard presents an order of magnitude increase in data rates compared with USB 2.0, thereby giving engineers the ability they...

The USB 3.0 SuperSpeed standard presents an order of magnitude increase in data rates compared with USB 2.0, thereby giving engineers the ability they need to address far more data intensive application scenarios. However, a number of flashpoints are arising that could potentially slow the adoption of this next-generation interface by the engineering community.

Figures compiled and published earlier this year by market research firm Global Industry Analysts suggest that worldwide sales of USB 3.0-enabled devices will hit the 3 billion unit mark by 2020. It’s likely that the Asia-Pacific region will dominate demand over this period (with a CAGR of 36 percent expected there).

One of the best opportunities to materialize so far for USB 3.0 technology has been in the bulk transfer of data to external storage devices. In this application, the added speed that can be delivered is of huge benefit, allowing increasingly common data logjams to be avoided. It’s important to note that USB 3.0 won’t merely be a consumer-led phenomenon. The projected market growth and rising interest from engineers in this standard will bring several additional dynamics into play, each of which will have significant influence. Among the other key drivers that will cause this market sector to mature will be higher performance security equipment and industrial data acquisition hardware, as well as increasing prevalence of medical imaging and high definition (HD) display systems in digital signage, point of sales units, etc.

USB 3.0 functionality and uses

The future for USB 3.0 looks to be a prosperous one, offering the same plug-and-play connectivity that made USB interfaces so popular originally, and helped it become the world’s most ubiquitous interface technology (with over 3 billion ports shipped every year), but with some added advantages. It can transfer data at rates up to 5 Gbps through a full duplex I/O structure, as opposed to the 480 Mbps that was attainable with USB 2.0. Furthermore, it ramps the power delivery from 2.0′s 10 W to 100 W. Hence, there’s far greater scope for its deployment in contemporary system designs.

USB 3.0′s backward compatibility with version 2.0 is another plus, facilitating upgrades to existing apparatus. Compared to other interconnect technologies, it outperforms the 800 Mbps of the latest FireWire incarnation and the 2.08 Gbps of CameraLink, while being more versatile than HDMI. Limited operating-system support may have put restrictions on it for a while, but now that USB 3.0 can work with Linux, it holds greater promise for all aspects of the embedded industry.

A key prospect for USB 3.0 outside of the consumer space is in industrial data logging, for example monitoring the liquid flow through a pipeline, so that leaks and blockages can be safeguarded against, or environmental sensing systems, which can ensure that the levels of airborne chemical compounds or particulates don’t exceed a safe threshold. Another example is the transfer of multimedia content. Supporting 1920 pixel by 1080 pixel video at 120 Hz, USB 3.0 has plenty of bandwidth available. HD image data can be displayed at frame rates that will permit smooth video playback without any lag or distortion being seen.

As there’s no need for multiple cables, the expense of deployment is lowered and valuable space is saved too. It could be employed in machine vision, enabling higher resolution image sensors, so that the items on a factory production line can be examined in fine detail at greater speed. Quality control benchmarks can thereby be upheld while simultaneously elevating output levels.

Surveillance cameras could profit from this technology, with greater image clarity for the observation of traffic (so that license plates can be identified when laws are infringed upon). Another place where USB 3.0 is certain to get attention is in 3D imaging, where banks of cameras can take pictures from a multitude of angles at the same time, for use in immersive gaming or body imaging for the fitting of prosthetic limbs.

ICs supporting USB 3.0

It’s clear that there’s a wealth of possibilities on the horizon for USB 3.0, but it must be noted that, to date, the semiconductor solutions available that support this standard have their limitations. A critical factor contributing to this is that these ICs have been predominantly microcontroller (MCU) based. Questions are now being raised about their suitability in a large proportion of applications.

As it’s necessary to create, compile, and debug reams of code for operating the constituent MCU element, the first wave of USB 3.0 interface devices have shown themselves to be complex to implement. They require experienced engineering teams and a lot of development time. The MCU cores featured within these ICs also make unit costs relatively high, and take up considerable board area too. In addition, there’s the increased power consumption to be taken into account. There’s a need for an alternative to the MCU approach to USB 3.0 system design, one that can address the bill-of-materials concerns, board-space utilization, and power budget.

FTDI Chip’s FT600/1 USB 3.0 to FIFO bridge devices can deliver 3.2 Gbps data bursting rates. The FT600 has a 16-bit wide FIFO bus interface, while FT601 has a 32-bit one. Acting as their foundation, they have a proprietary hardwired 32-bit, 100 MHz processor core that controls the internal sub-units of each IC. This reduces the amount of code that needs to be written.

The devices in the FT600 series have two interfacing modes, a standard synchronous 245 FIFO mode that’s which is optimized for simple straight-forward implementations, and a multi-channel FIFO mode that supports up to four logical FIFO channels and data structures optimized for higher throughputs, as well as providing more system design flexibility

Based in Singapore, Chee Ee Lee heads FTDI Chip‘s product development. He has over 20 years of industry experience in various high-level engineering roles.

Based in Glasgow, Scotland, Gordon Lunn is FTDI Chip’s Global Customer Engineering Support Manager. He graduated from Heriot Watt University with a BEng (Hons) in Electrical & Electronic Engineering.

Gordon Lunn, FTDI Chip