Universal Serial Bus (USB) is the most successful peripheral interface in the
history of the personal computer. USB is a standard connection interface between computers and digital devices.
It enables the easy transfer of data through a direct connection or cable. USB was created with
simplicity of use in mind, and thus “plug-and-play” became the expectation as drivers loaded themselves and
printers, external hard drives, and other peripherals “just worked.”
USB has other features that make it attractive for product designs. It provides a small current available to the
peripheral, and so USB became a standard for recharging and data transfer in one cable. USB 3.0
ups that current to 900 mA from 500 mA (in 2.0).
The USB 3.0 bus and the USB 2.0 bus operate separately in the same USB 3.0 cable and thus USB
3.0 is backwards compatible with USB 2.0 products by inclusion, with the right cable. SuperSpeed traffic travels
in both directions simultaneously, referred to as “full duplex.” The SuperSpeed portion of USB 3.0 hubs do not
poll device endpoints to check for ready data; devices operating at SuperSpeed initiate communication with the
USB 3.0 hub only when they need it. USB 3.0 also provides for more aggressive power savings modes.
| Speed |
Low |
Full |
High |
SuperSpeed |
| Bits per Second |
1.5M |
12M |
480M |
4.8G |
SuperSpeed now enables streaming video and mass storage at a new level of performance with much faster data
transmission rates. Much of USB 3.0 is the same as USB 2.0 from a software standpoint in terms
of device classes, data transfer types, and descriptors. Hardware is affected by differential signaling at
significantly higher frequencies, however, which can present new challenges to designers in terms of increased
signal loss or degraded signal integrity.
Note that USB 3.0 systems cannot utilize both USB 2.0 and 3.0 busses simultaneously; they are mutually exclusive.
This holds true with device power consumption as well. Therefore, don’t expect to draw 900mA with a low-, full-,
or high-speed USB device. With the largest installed base for interfaces (of USB 2.0) combined with backwards
compatibility (of USB 3.0), the new USB is sure to be a winner.
The USB host controller, often simply referred to as “the USB host,” typically
resides in a computer or head-end system. The host controller acts as an interface to move data
between the computer and USB bus-resident devices by processing data structures, initiating
transactions, and sending the data as packets over the USB bus. It can control several USB ports.
Information flow towards the host is called “upstream” traffic and the flow of information away from
the host is referred to as “downstream” traffic. The USB specification requires the host controller
to include a “root hub,” which provides data flow control for the bus.
Hubs are the supporting backbone of communication in
the USB bus network. A hub in the host controller, called the root hub, provides data flow control
at the highest protocol level of the USB network, which is physically nearest to the computer, or
head end of the system.
A USB transceiver is a physical layer device that
prepares data for transmission and then sends to, and receives data from, another transceiver. The
transceiver detects connection and provides the low level USB protocol and signaling. The term
“transceiver” indicates an implementation of both the transmit and receive functions. It transmits
and receives, encodes and decodes data, provides error indication, implements buffers to stage data
until it can be managed, and adjusts for the clock rate from the serial stream on the USB SuperSpeed
bus to match that of the “link layer” higher up on the communication stack.
This device is used to
provide USB 3.0 connectivity to storage devices such as hard disk drives, optical drives and other
SATA-dependent devices. It can be used to implement both storage and human interface device class
devices for USB. It supports both generation 1 & 2 SATA speeds (1.5 Gbps & 3.0 Gbps, respectively)
and includes a Cortex M3 microcontroller. Serial peripheral interface (SPI) and pulse width
modulation output are included as well as provision for some general purpose I/O.
USB 3.0 will
be targeted initially at the PC market and in devices requiring high rates and volumes of data
transfer, such as external storage, consumer electronics, and communications devices with increasing
amounts of storage. High frequency signals are subject to problems with signal quality. Signal
conditioners can help by filtering out anything above the expected ceiling frequency, and can
“re-drive” the signal by retransmitting at increased signal power, effectively “re-driving” it
through the signal path. USB is a consumer-friendly technology. Cable lengths for USB devices are
not necessarily going to remain fixed at the length of the cable that came in the box. Therefore,
the re-driver should continuously monitor the signal and dynamically equalize (or adaptively
balance) itself in order to optimize the signal.
USB plugs and
receptacles are meant to reduce human error by their unique shape; they fit together in only one
way. USB plugs and receptacles come in Type A (typically connecting to hosts or hubs) or Type B
(typically connecting to devices) and 3 sizes: standard, mini, and micro. Type A plugs always face
upstream, Type B faces downstream.
All USB devices (or peripherals) have an upstream connection to the host and all hosts have a
downstream connection to the device. USB 2.0 plugs will fit into USB 3.0 receptacles but do not
contain the extra wires needed to convey SuperSpeed communications. Note that USB 3.0 Type B plugs
do not fit into USB 2.0 receptacles. A USB 3.0 device going to a USB 2.0 host, for example, will
only work with a USB 2.0 cable assembly.
There is no set
standard on cables for USB, but rather for performance of the cable or system. USB 3.0 operates at
data rates of up to 5 GHz by employing data-doubling techniques on a 2.5 GHz signal. Higher
frequency signals suffer notably more signal losses than lower frequencies, so signal quality can be
greatly affected with lower quality cables. To avoid signal degradation, higher-quality
thicker-gauge cabling should be used with USB 3.0 devices.
If you have ever been zapped by a socks-wearing kid who has just discovered static
charge build up, you have experienced ESD first hand. ESD is like a miniature, localized lightning
bolt caused by an electrical discharge. ESD can have seriously damaging effects on an integrated
chip or system, or can cause poor performance or failure later on by merely weakening the circuits.
ESD protection should be used where ESD exposure can occur, and is recommended for data cables
exposed via their connection points such as USB cable assemblies.
ESD can also be a periodic weak ESD that eventually breaks down the circuit. Devices may escape
detection in the factory test harness, but are going to suffer a shorter product life. ESD may occur
on any pin that is exposed to its environment. Common mode chokes can also be helpful in reducing,
but not eliminating, ESD caused by transmission lines.
Common Mode Chokes are often placed at the base of a USB
connector to reduce noise coming from the cable. Common mode choke coils consist of two or more
magnetically combined coils that act to isolate unwanted AC frequency currents from a main circuit
via two methods: common mode noise rejection and by creating a signal boundary at a defined cut-off
frequency. Therefore, chokes can be seen as signal conditioners of a kind. Note that chokes are not
full ESD (electrostatic discharge) protection devices, but can help reduce aspects of ESD system
upset in addition to their main function of general noise reduction.
The chokes suppress only the noise common to currents travelling in parallel and in the same
direction (common mode noise) while maintaining good differential (equal but opposite) signal
quality. In USB, which employs signals going in opposite directions, noise typically flows in
the same direction (in a common mode.)
For USB 3.0, a choke should align with the USB 3.0 standard, which requires 90 ohm differential
impedance. If the impedance of components inserted along a transmission line do not match the
impedance of the transmission line itself, then problems like signal reflection can occur.
USB 3.0 Switches provide smooth transitions. By maintaining
low resistance and impedance, the signal switch can keep insertion losses down and preserve the
signal so that no glitches are reflected back to the transmitter, thus helping to maintain a good
quality signal.
USB 3.0 requires a Clock. Clock technology is more expensive
and more complicated at higher frequencies. A lower frequency clock can be used and then multiplied
to achieve the desired clock rate; however any jitter (error) is also multiplied. (Jitter is
shakiness in the signal, caused by electro-magnetic radiation or the influence of nearby signals.)
Circuitry supporting clock function should be selected for low noise.
Cypress Semiconductor EZ-USB® FX3
SuperSpeed USB Controllers
EZ-USB® FX3 SuperSpeed USB Controller enables developers to add USB 3.0 functionality to any system.
EZ-USB FX3 has a fully configurable, parallel, General Programmable Interface (GPIF II™) which can
connect to any processor, ASIC, or FPGA. Integrating USB 3.0 and USB 2.0 physical layers with a 32-bit
ARM926EJ-S processor, the FX3 provides exceptional processing power, great flexibility, and data transfers
as high as 320 MBps from GPIF II™ to USB.
Contributed by TI
Ever wonder why the standards body did what they did? What's that bump on the B-connector for?
Efficiency, standards for charging devices, the truth about backward compatibility, and the mysteries of old
and new connectors are discussed here.
Contributed by Littelfuse
SuperSpeed introduces faster transfer rates, which means higher frequencies. With higher frequencies
comes greater complexity in selecting the right protection. The introduction of additional differential data
pairs requires more data lines to be protected against ESD than USB 2.0.
