AGP (Accelerated Graphics Port) – mature PC graphics IO standard

Analog signals – use multiple voltage levels to represent information (i.e. gas gauge receives 0 to 12 volts from tank level sensor to represent empty to full)

ASI (Advanced Switching Interconnect) – highend server serial IO standard

ATCA (Advance TeleCom Architecture) – PICMG’s latest embedded telephony/industrial computer and backplane/chassis standard

ATM (Asynchronous Transfer Mode) – highend telephony IO standard

Attenuation – refers to any reduction in the strength of a signal (loss). It is usually measured in decibels (dB) which is a logarithmic term that ratios the output voltage or power as a function of input voltage or power. As an example, 3dB loss is equivalent to 30% loss of voltage and 6dB loss is equivalent to 50% loss of voltage. In the case of power, 1.5dB loss is equivalent to 30% loss of power and 3dB loss is equivalent to 50% loss of power. Typically, cables are rated in dB loss per 100 feet.

AP = 10log (POUT /PIN ), where AP = power attenuation

AV = 20log (VOUT /VIN ), where AV = voltage attenuation

Aurora – new open protocol and physical layer spec from Xilinx
Backward Crosstalk – see near-end crosstalk

Bandwidth – The bandwidth of a signal is a measure of the range of frequencies that comprise a signal. Typically, adigital signal is composed of its fundamental frequency (the one we always talk about) and several higher order frequencies that give it its shape. The term is also used in reference to the frequency-response characteristics of electronic components such as drivers, receivers, connectors, and backplanes. All transmitted signals, whether analogor digital , have a certain bandwidth. Typically a connector requires a bandwidth greater than the frequency of the signal transmitted to minimize the distortion of the signal. Bandwidth figures are meaningless unless they are defined as to how they were arrived at. For example, the bandwidth of a connector might be defined as 2 GHz, where 2 GHz is the frequency at which a sine wave is attenuated by 3 dB.

BER – the bit error rate (BER) is the percentage of bits that have errors relative to the total number of bits received in a transmission, usually expressed as ten to a negative power. For example, a transmission might have a BER of 1 x 10-6, meaning that, out of 1,000,000 bits transmitted, one bit was in error. The BER is an indication of how often a packet or other data unit has to be retransmitted because of an error. A high BER may indicate that a slower data rate would actually improve overall transmission time for a given amount of transmitted data since the BER might be reduced, lowering the number of packets that had to be resent. A BERT (bit error rate test or tester) is a procedure or device that measures the BER for a given transmission.

Bit rate – number of bits that pass through a given point in a given amount of time, usually expressed as Kbps (thousands of bits per second), Mbps (millions of bits per second) or Gbps (billions of bits per second)

Bits – digital data is transmitted and stored as strings of 1’s and 0’s, each one of these 1’s and 0’s is a bit. See digital signals .

Broadside-coupled – refers to differential PCB traces that are physically above and below each other

Bytes – a group of bits make up a byte (word) (typically 8,16 or 32 bits to a byte)

Capacitance – refers to the ability of a device to store energy in the form of an electrostatic field. In its simplest form, a capacitor is a pair of parallel plates spaced apart with a dielectric material between them. Coax cables have a certain capacitance per foot. The dielectric materials in a cable affect the capacitance of the cable. If two cables are identical in geometry, the cable with the higher dielectric constant insulator will have the higher capacitance per foot and, thus, the lower impedance. The standard unit of capacitance is the farad, abbreviated F. This is a large unit; more common units are the microfarad, abbreviated µF (1 µF = 10-6 F) and the picofarad, abbreviated pF (1 pF = 10-12 F). The reactance of a capacitor decreases as frequency increases.

XC = 1/(2πfC), where XC = capacitive reactance

f=frequency and C=capacitance

Clock Speed – in a computer, clock speed refers to the number of pulses per second generated by an oscillator that sets the tempo for the processor. Clock speed is usually measured in MHz (megahertz, or millions of pulses per second). A typical computer clock runs at several hundred megahertz. The clock speed is determined by a quartz-crystal circuit, similar to those used in radio communications equipment.

CML (Compact Multi-Lane) – SFF-8088, mini-SAS IO connector

Controlled Impedance – a specific impedance that is maintained along the length of the transmission line. Seeimpedance.

cPCI (Compact PCI)  – embedded computer backplane/chassis standard

Crosstalk – the effect whereby a signal on one conductor will be transposed to an adjacent conductor due to mutual capacitance and mutual inductance. It is usually measured as a percentage of the applied voltage; for example, if a 1 volt peak to peak signal is transmitted on one line, a nearby line may show a 0.1 volt peak to peak signal; this corresponds to 10% crosstalk. The frequency or risetime of the signal must be specified. In general, crosstalk will increase with higher frequencies or faster risetimes. Measurements are usually made with one or more active lines and one quiet or victim line. The active lines are the conductors that are driven by a signal; the quiet or victim line is the conductor being affected by the driven lines.

CSI (Coherent Scalable Interface) – like SCI?, new Intel IO standard

CX4 – Coax or twinax version of 10GEthernet 4 lanes x 3.125G

DDWG (Digital Display Working Group) – Business Consortium

Dielectric Constant – the property of a dielectric (insulator) which determines the electrostatic energy that can be stored. The higher the constant the higher the energy stored within the capacitor. The dielectric constant affects the properties of transmission lines. The dielectric constant can change with frequency; however, teflon-based dielectrics are quite constant over a wide frequency range. Common dielectric constants: air = 1; foamed FEP = 1.53; solid PTFE = 2.03; solid polypropylene = 2.18; FR-4 = 4.5

Dielectric Materials – are insulators that are used to provide separation between conductors. Examples are air, Teflon and FR-4. Their inherent dielectric constant affects how the conductors behave electrically.

Differential Impedance – refers to the impedance of a pair of conductors when driven in a differential mode, that is, when the conductors are driven by signals that have opposite polarity edges.

Differential Mode Transmission – uses two conductors to carry the signal, and one shield to provide the return path. The signal conductors are driven by opposite polarity signals. Typical examples: twinax, shielded twisted pair, ribbon cable

Digital Signals – use two states (on/off, high/low, 1/0) to represent data

Dispersion – the phenomenon in cables and PCB’s whereby higher frequency signals propagate faster along the length of the conductor than do lower frequency signals. This causes signal distortion because most digital signals are made up of many frequencies. This can be due to dielectric constant changes with frequency or to changes in reactance of the cable with frequency.

DisplayPort – new IO standard –

Distortion – refers to the changing of any of the attributes of a signal because of the medium through which the signal flows (i.e., a connector or cable) These attributes may be amplitude, risetime, pulse width, etc.

DVI (Digital Visual Interface) –, display video IO standard

Edge-coupled – refers to differential PCB traces that are side-by-side

Electrical length – refers to the total time it takes for a signal to traverse a conductor and may be expressed in nanoseconds. See also propagation delay.

Equalizer – a reactance connected in series with a transmission line to alter the frequency response characteristics of the line. Typically this is used as a high-pass filter which attenuates low frequencies to compensate for the cable’sattenuation at high frequencies. It will improve the “eye opening” for a cable. See also eye-pattern.

Even mode coupling  – the coupling between the signal paths and ground within a differential pair.

Even mode impedance – the impedance of a single line in a coupled-line pair when a common-mode signal drives both conductors within the pair.

ExpressCard –, latest PCI Express packaging standard.

Eye-pattern – the name given to the “picture” or pattern produced when multiple traces are recorded on an oscilloscope by sending multiple 1’s and 0’s down a cable assembly and “looking” at the far end of the cable. The 1’s and 0’s are randomly generated so that many different sequences of highs and lows are examined by the eye pattern.

This test is an excellent way to look at how the cable assembly will perform in the “real world”. It encompassesattenuation and risetime degradation (reduction in amplitude and risetime as a function of frequency components of the pulse and bit pattern sequence), dispersion (jitter due to differing speeds of signal transmission as a function of frequency) and within-pair skew (offset of differential signal due to differing electrical lengths of the wires within the differential pair). The “eye” of the input signal is compared to the “eye” of the signal at the far end. The height and width of the “eye” (referred to as the “eye opening”) show how the signal has been distorted and can be used to analyze if the user’s receiving circuit can reliably distinguish the logic 1’s and 0’s emerging from the cable assembly. Ideally, the eye pattern would be a square that is the same height and width as the input signal. The X-axis plots time and the Y-axis plots amplitude. Referring to the illustration, 1, 2,…8 are “snapshots” of the signal at equal lengths of time apart (X number of clock cycles) and these “snapshots” are superimposed one on top of another yielding the eye-pattern in the lower left of the illustration. (Illustration courtesy of Agilent Technologies and Electronic Design )

Far end – refers to the non-driven end of a transmission line (the receiver end)

Far-end (forward) crosstalk – the crosstalk measured near the end opposite the driven end of the active signal line(s).

FC (Fibre Channel)  – highest end storage serial IO standard

FCIA (Fibre Channel Industry Association) – Business Consortium

Forward Crosstalk – see far-end crosstalk

FPDP (Front Panel Data Port) – VITA-system parallel IO standard

Giga – prefix denoting billions or 1,000,000,000 or 10 to the 9th power.

Ground bounce – the “raising” or “lowering” of the voltage on a ground plane or ground pin due to the inductance of the ground path. This can cause the receiving circuitry to misinterpret a 1 or 0.

HDMI (High Definition Multimedia Interface) – TV display IO standard

Hertz vs. bits/sec vs. Bytes/sec – Hertz refers to frequency at which a signal repeats itself; Mbps refers to how many bits per second are transferred; MBps refers to how many bytes per second are transferred. Generally a byte is 8 or 16 bits long. In general, MHz equals one-half of Mbps, (Therefore 400Mhz = 800 Mbps. If a byte is 8 bits long, then 800 Mbps = 100 MBps.)

HSSDC2 – High Speed Serial Digital Connection style 2, 1x IO standard

HT (HyperTransport) – High end server parallel IO standard, AMD led, 1x-32x, 2.6Ghz per diff pair

IB (InfiniBand) – High end server IO standard

IBTA (InfiniBand Trade Association) –

Impedance – the opposition of an electronic component to the flow of current, measured in ohms. Most commonly thought of as a resistor, however, all devices have impedance associated with them. In transmission lines, it is a measure of the characteristic impedance of a cable, a PCB trace or an interconnect. The impedance of a transmission line (ZO ) is defined as: ZO = √(L/C), where C = capacitance and L = inductance. As capacitance decreases orinductance increases per unit length, the impedance of the cable or connector will increase. Conversely, as capacitance increases or inductance decreases per unit length, the impedance of the cable or connector will decrease. Typical transmission lines may have impedance of 50 or 100 ohms. When the impedance of a connector is stated, the risetime of the pulse used to determine that impedance should also be stated.

Inductance – refers to the ability of a device to store energy in the form of a magnetic field. Cables have a certain inductance per foot. In its simplest form, an inductor consists of a wire loop or coil. The standard unit of inductance is the henry, abbreviated H. This is a large unit. More common units are the microhenry, abbreviated µH (1 µH =10 -6 H) and the millihenry, abbreviated mH (1 mH =10-3 H). Occasionally, the nanohenry (nH) is used (1 nH = 10-9 H). The reactance of an inductor increases as frequency increases.

XL = 2πfL, where XL = inductive reactance

f = frequency and L = inductance

Insertion loss – refers to the losses resulting from the insertion of a cable, connector or other device into an electrical system. Includes, but is not limited to, DC losses from the resistance of the conductor and skin effect losses.

Interlaken – new protocol and physical layer spec, from Cisco

Jitter – the shifting or displacement of some aspect of the pulses in a high-frequency digital signal. As the name suggests, jitter can be thought of as shaky pulses. The deviation can be in terms of amplitudephase timing, or thewidth of the signal pulse. Another definition is that it is “the period frequency displacement of the signal from its ideal location.” Among the causes of jitter are electromagnetic interference (EMI) and crosstalk with other signals. Jitter can cause a display monitor to flicker; affect the ability of the processor in a personal computer to perform as intended; introduce clicks or other undesired effects in audio signals, and loss of transmitted data between network devices. The amount of allowable jitter depends greatly on the application.

Kilo – prefix denoting thousands or 1,000 or 10 to the 3rd power.

Mega – prefix denoting millions or 1,000,000 or 10 to the 6 th power.

Microstrip – a specific transmission line on a PCB where the signal trace is on an outside surface of the PCB and is spaced above a ground plane by the dielectric material, such as FR-4.

Modeling – refers to the creation and simulation of electronic circuits on a computer. SPICE is the most common circuit simulator in use. IBIS is a newer version that semiconductor manufacturers prefer to use because it protects their proprietary designs. The models can be created using various methods including actual measurements and computer programs. A model of a connector may contain the inductance, capacitance, impedance and time delay of the connector. These numbers can be used in SPICE (and other) circuit simulation models to simulate how the connector will perform in the user’s circuit. A “lumped” model uses R, L, and C components to model the behavior of the device. A lumped model is usually accurate then the device being modeled is much “shorter” than the rise time of the signal. A “distributed” model is a higher order approximation that can more accurately simulate the real world device.

Mutual capacitance and mutual inductance – the capacitances and inductances related to two or more conductors in close proximity to each other. These can cause interaction among the conductors in the form of crosstalk .

MicroTCA (Micro TeleCom Architecture) – new picoCell blade/backplane/chassis standard using edge connectors

MIL-1533 – 1Mbs military device IO standard – new 10Mbs revision

Myrinet – High end server serial IO standard; 10G w/4 lanes x 3.125G

Near end – refers to the driven end of a transmission line

Near-end (backward) crosstalk – the crosstalk measured on the victim line near the driven end of the active signal line(s).

Noise – refers to unwanted signals or energy present on a data line. Noise can be due to many outside influences such as electromagnetic interference from transformers or lightning. It can also be due to nearby circuits, such as crosstalk.

NPF (Network Processing Forum) – business consortium –

NUMA-Link (Non-Uniform Memory Architecture) – wide-parallel semi-custom IO interface for highest end servers

Odd mode coupling – the coupling between two signal paths within a pair that is driven differentially

Odd mode impedance  – the impedance of a single line in a coupled-line pair when a differential signal drives both conductors within the pair.

OIF (Optical Internetworking Forum) –

OIF CEI (OIF Common Electrical Interface) – cu backplane standard

Packet – the unit of data that is routed between an origin and a destination on the Internet or any other packet-switched network. When any file (e-mail message, HTML file, GIF file, URL request, and so forth) is sent from one place to another on the Internet, the Transmission Control Protocol (TCP) layer of TCP/IP divides the file into “chunks” of an efficient size for routing. Each of these packets is separately numbered and includes the Internet address of the destination. The individual packets for a given file may travel different routes through the Internet. When they have all arrived, they are reassembled into the original file (by the TCP layer at the receiving end).

A packet-switching scheme is an efficient way to handle transmissions on a connectionless network such as the Internet. An alternative scheme, circuit-switching, is used for networks allocated for voice connections. In circuit-switching, lines in the network are shared among many users as with packet-switching, but each connection requires the dedication of a particular path for the duration of the connection.

Packet-switched – describes the type of network in which relatively small units of data called packets are routed through a network based on the destination address contained within each packet. Breaking communication down into packets allows the same data path to be shared among many users in the network. This type of communication between sender and receiver is known as connectionless (rather than dedicated). Most traffic over the Internet uses packet switching and the Internet is basically a connectionless network.

Contrasted with packet-switched is circuit-switched, a type of network such as the regular voice telephone network in which the communication circuit (path) for the call is set up and dedicated to the participants in that call. For the duration of the connection, all resources on that circuit are unavailable for other users. Voice calls using the Internet’s packet-switched system are possible. Each end of the conversation is broken down into packets that are reassembled at the other end.

Another common type of digital network that uses packet-switching is the X.25 network, a widely installed commercial wide area network protocol. Internet protocol packets can be carried on an X.25 network. The X.25 network can also support virtual circuits in which a logical connection is established for two parties on a dedicated basis for some duration.

PCI (Peripheral Component Interface) – parallel bus standard

PCI-SIG (PCI Special Interest Group) –

PCIe (PCI Express)  – serial link standard 2.5, 5Gbs per diff pair, 8x, 16x, 32x links

PDH (Plesiochronous Digital Hierarchy) – legacy telephony IO standard

Phase timing – the timing relationship between the clock signal edges and the data signals edges

PICMG (PCI Industrial Manufacturer Group) – embedded computer business consortium

pRIO – Mature parallel Rapid IO

Propagation delay – the speed or rate with which a signal travels between the input and output of a transmission line, usually expressed in nanoseconds per foot. The materials surrounding the conductor affect this speed. Space or vacuum is the fastest medium (1 nsec/ft), whereas air is 1.0167 nsec/ft. Teflon is slower than air (1.45 nsec/ft), and FR-4 is slower than Teflon (2.12 nsec/ft). See also electrical length.

PD (nsec/ft) = (√K)/Co , where PD = propagation delay

Reactance – the resistive properties of capacitors and inductors at various frequencies. See capacitance andinductance.

Remote IO – IBM high end parallel/serial server IO standard; sometimes known as RIO

Return loss – refers to distortion measured at the near end of the transmission line due to reflections at discontinuities (changes in characteristic impedance) within the transmission line. These losses are the result of part of the signal being sent back to the source.

RIO (Rapid IO) – high end serial switch fabric IO standard;

Risetime degradation – refers to the increase in the risetime and falltime of a pulse as it travels through a transmission line. This means that a pulse with a 100 psec risetime entering a cable or connector will exit that cable or connector with a risetime longer than 100 psec.

RPR (Resilent Packet Ring) – IEEE-802.17 MAN IO standard

SAS (Serial Attached SCSI) – mid level storage IO standard

SATA (Serial ATA) – the lower end storage IO standard

SCI (Scalable Coherent Interface) – IEEE-1596 server parallel IO standard

SCSI (Small Computer System Interface) – storage parallel IO legacy standards

SDH (Synchronous Digital Hierarchy) – European standard equivalent to SONET

SDI (Serial Digital Interface) – the SMPTE-292 video IO standard

SFPDP – serial FPDP IO standard,

SID (Society of Information Displays) – International business consortium – promotes new display technologies but not interfaces

Signal – typically composed of voltage and current travelling on a conductor (wire) to convey information, data, or commands; has two main characteristics: namely, amplitude and time base

Signal attributes:

Amplitude – indicates voltage (or current) levels (“height” of signal)

Pulse Width – length of time that digital signal is high (or low) measured at 50% of amplitude

Risetime – length of time for digital signal to go from low to high; usually defined as the period of time that it takes to go 10-90% or 20-80% of the high level

Rule of thumb :  10-90% risetime » 1.33 times the 20-80% risetime

Frequency – number of times per second that a repeating signal (i.e., sine wave, clock) repeats itself; usually expressed as kHz (thousands of cycles per second), MHz (millions of cycles per second) or GHz (billions of cycles per second)

Period – length of time before a repeating signal (i.e., sine wave, clock) repeats itself. Period=1/frequency

Data transfer rate – see bit rate

Wavelength the physical length of a period of a signal. In air, the formula is l = c/f, where l = wavelength, f = frequency and c = the speed of light in air. In another dielectric, the wavelength scales as 1/ Ö K, where K = dielectric constant. Therefore, in any dielectric, l = c/(f x Ö K). Wavelength in air at 100 MHz = 120 inches; at 1 GHz = 12 inches; at 5 GHz = 2.4 inches

Signal integrity (SI) – a term used to refer to the quality of an electrical signal after it has been sent down a conductor. It encompasses many parameters such as distortiondelayattenuation, ringing, crosstalkdispersionimpedance, etc.

Single-ended transmission – also called single mode, uses two conductors, one carries the signal, and one provides the return path. Typical examples: coax, twin lead, twisted pair, ribbon cable

Skew – the difference in the time that a pair of identical signals takes to get from point A to point B going down two different paths. Skew is the result of different electrical lengths which in turn are the result of different physical lengths (i.e. 12 inches vs. 12.5 inches) and different velocities of propagation. This property is important in relation to the synchronization of data signals to clock signals. See phase timing.

Skin effect – refers to the phenomena whereby the signal traveling through a conductor will be conducted only on the outer surface (skin) of the wire as the frequency increases. At lower frequencies the current travels through the entire cross section of the conductor. As the frequency increases, the current only travels nearer the outer surface of the conductor. The thickness of this “skin” scales as 1/√f, where f = frequency. At 1 GHz the signal travels only on the outer 2 microns (0.00008 inches) of a copper wire.

SMPTE (Society of Motion Picture and Television Engineering) – Society of Motion Picture and Television Engineering – international technical consortium for broadcast IO standards

SONET (Synchronous Optical Network) – WAN comm IO standard

Speed – can refer to frequency, bit rate, or velocity of propagation. High speed is ambiguous; high speed to a PC designer may be 400 MHz but to a telecomm designer it may be 2.5 GHz. As velocity of propagation, it refers to the speed at which the signal travels from one point to another along a conductor. The materials surrounding the conductor affect this speed. For example, in air, the velocity of propagation is 300 million meters per second and in FR-4 printed circuit boards it is about half that speed. See dielectric constant.

Speedsignal = Co /(√K), where Co = the speed of light and K = dielectric constant

Speedsignal » 1 foot/nsec in air and 6 inches/nsec in FR-4

Therefore, a 1nsec risetime is 6 inches long in a PCB and 12 inches long in air

sRIO (Serial RIO) – Sometimes known as SRIO

SSPI (Scalable SPI) – new high end comm IO standard with NPF and OIF

STA (SCSI Trade Association) – business consortium –

Stripline – a specific transmission line on a PCB where the signal trace is buried within the PCB and is spaced above and below a ground plane by the dielectric material, such as FR-4.

TDRtime domain reflectometer, TDR, is a combination of a sampling oscilloscope and a fast risetime signal source that can be used to measure impedance, velocity of propagation, electrical length, propagation delay, risetime degradation and skew of cables and connectors.

TDM (Time Division Multiplexing) – a mature telephony interface

Transmission line – commonly used to denote a controlled impedance conductor path that has a defined velocity of propagation. Geometry and dielectric materials determine many of the properties of a transmission line. Examples are coaxial cable, twinax or parallel paired cable, twisted pairs, parallel traces on a PCB.

UDI (Universal Display Interface) – new display IO standard with DDWG

Utopia – mature intra/inter-rack telephony wide parallel IO standard

VAUI (Vitesse Attachment Unit Interface) – similar to XAUI

Velocity of propagation – the actual speed at which the signal travels from one point to another along a conductor. The materials surrounding the conductor affect this speed. For example, in air, the velocity of propagation is 300 million meters per second and in FR-4 printed circuit boards it is about half that speed. It is usually expressed as a percentage of the speed of light. Air is the fastest medium (100%), Teflon is slower than air (70%), and FR-4 is slower than Teflon (47%). Typically connectors are faster than printed circuit boards; those connectors with the most air being fastest. Velocity of propagation affects skew. See dielectric constant.

VP (%) = 1/(√K) x 100

where VP = velocity of propagation and K = dielectric constant

VESA (Video Electronic Standards Association) – display IO standards

VITA (VME International Trade Association) –

VME (VERSA Modular Eurocard) – European military IO standard

VXS (VME Switched Serial) – a 4x 2.5 10Gbs VITA-41 IO standard

XAUI – 10Gbs attachment unit interface – 10G Ethernet interface

XAUI-2 – 6Gbs per pair XAUI interface from Intel 4x – SFF-8470 four lane, 8 diff pair, IO connector used with IB and SAS