Hardware Reference
In-Depth Information
Chapter 14. External I/O Interfaces
Introduction to Input/Output Ports
This chapter covers the primary external peripheral input/output ports on modern and legacy PC
systems. This includes the legacy serial and parallel ports that have been standard on PCs since the
beginning, as well as the universal serial bus (USB, which has replaced the older serial and parallel
ports), IEEE 1394 (also called FireWire or i.LINK), and Thunderbolt Technology interfaces. IEEE
stands for the Institute of Electrical and Electronic Engineers. Although eSATA is also considered an
external I/O interface, it is a derivative of Serial ATA (SATA), which is mainly used as an internal
storage device interface. SATA is covered in Chapter 7 , The ATA/IDE Interface .” Small computer
systems interface (SCSI) is another type of internal/external interface; however, desktop PCs today
rarely implement SCSI. If you want to learn more about this architecture, refer to Upgrading and
Repairing Servers (ISBN: 978-0789728159).
We can divide external I/O connections into high-speed and low-speed variants. Currently, the two
most popular high-speed external I/O connections are USB and IEEE 1394; however, Thunderbolt is
replacing IEEE 1394, especially in the high-end video editing arena. Low-speed connections include
standard serial and parallel ports (often called legacy ports ), which have generally been replaced by
USB in newer systems and peripherals.
Serial Versus Parallel
The current trend in high-performance bus design is to use a serial architecture, in which one bit at a
time is sent down a wire. Because parallel architecture uses 8, 16, or more wires to send bits
simultaneously, a parallel bus is faster than a serial bus at the same clock speed . However,
increasing the clock speed of a serial connection is much easier than increasing that of a parallel
connection, so much so that the higher attainable speeds more than completely offset the difference in
the number of bits sent at a time. In the end, serial interfaces can be designed to be faster than parallel
interfaces, and for much less cost and complexity.
Parallel connections in general suffer from several problems, the biggest being signal skew and jitter,
which conspire to limit both clock speeds and signal distances. The problem in a parallel bus is that,
although the multiple bits of data are fired from the transmitter at the same time, by the time they reach
the receiver, propagation delays have conspired to allow some bits to arrive before the others. The
longer the cable, the longer the time variation between the arrival of the first and last bits at the other
end. This signal skew , as it is called, limits both clock speeds and cable lengths. Jitter is the
tendency for the signal to reach its target voltage and float above and below for a short period, which
becomes pronounced at higher speeds and at longer distances.
With a serial bus, the data is sent one bit at a time, one after the other. Because there is no worry
about having multiple bits arrive simultaneously, the clocking rate can be increased significantly.
At high clock rates, parallel signals tend to interfere with each other. Serial again has the advantage
because with only one or two signal wires, crosstalk and interference between the wires in the cable
are negligible.
In general, parallel routing or cabling is far more expensive to implement than serial routing or
cabling. In addition to the many extra traces or wires needed to carry the multiple bits in parallel, the
 
 
 
 
Search WWH ::




Custom Search