TIA/EIA-422-B Overview

National Semiconductor
TIA/EIA-422-B Overview Application Note 1031
Michael R. Wilson
February 1996

ABSTRACT shaded region in Figure 2. The operating range of receiver is

This application note covers topics associated with concerns defined between ± 10V and is represented by the shaded
for implementing a balanced interface circuit utilizing the TIA/ area in Figure 2.
EIA-422-B (formerly RS-422-A) electrical interface standard.
The items designated by bullets below indicate the topics
covered within this application note.
• Cable Length and Data Rate
• Termination
• Failsafe
• Configuration
• ESD Protection
• Live Insertion
• By-Pass Capacitors
• Stub Lengths AN012598-2
• Receiver Power Off Characteristics FIGURE 2. Receiver Operating Range
• Typical Cable Media
Also, RS-422 receivers have a ± 200 mV threshold over the
Each topic has an independent section. The sections are
entire common mode range of ± 7V. A differential noise mar-
identified by bold all upper case titles. Subsection titles are
gin ≥1.8V is guaranteed between the driver’s differential out-
bold only.
put swing and the receiver’s threshold.
INTRODUCTION/OVERVIEW RS-422 drivers and receivers are designed for point-to-point
TIA/EIA-422-B (RS-422) is an industry standard specifying and multi-drop configurations but not multi-point. For
the electrical characteristics of a balanced interface circuit. multi-drop configurations, a daisy chain is the recommended
Other prefixes are commonly used with the RS-422 stan- interface configuration.
dard. These include EIA, EIA/TIA, and RS; although previ- Caution, at long distances or high data rates, termination is
ously correct, today TIA/EIA is the correct prefix. However, recommended to reduce reflections caused by a mismatch
for simplicity, RS-422 will be used throughout the rest of this in the impedance of the cable and the impedance of the re-
application note. Also, a suffix letter denotes the different re- ceiver’s input. Refer to the section entitled “Termination” for
visions of the standard but various prefixes with the same further information.
suffix reference the same identical standard. Significantly, the RS-485 differential interface standard is
RS-422 was introduced to solve the limitation problems of very similar to RS-422. However, there are differences that
single-ended standards like TIA/EIA-232-E. Single-ended in- distinguish the two standards from one another; which in-
terfaces lack common-mode noise rejection capability; ideal clude the output stage of the driver, the common mode range
for noisy environments. Also, data rates are usually limited to of the interface, the input resistance of the receiver, and the
less than 0.5 Mbps. A RS-422 interface may be implemented drive capability of the driver. For more details concerning the
to overcome these limitations. comparison of RS-422 and RS-485, please reference Na-
A RS-422 driver can drive up to ten unit loads (i.e., 4 kΩ to tional’s application note AN-759.
circuit common is one unit load). The driver is capable of
CABLE LENGTH AND DATA RATE
transmitting data across 4000 feet (recommended limit) of
cable; but not at maximum data rates (see Figure 3 ). Stan- Cable Length and Data Rate have an inverse affect on each
dard RS-422 drivers are guaranteed to source and sink a other. When operating at either the recommended maximum
minimum 20 mA across a 100Ω load. This corresponds to a cable length or data rate the other can not be obtained. For
minimum differential output voltage, VOD, of 2V across the instance, it is not possible to operate at 4000 feet when op-
load (see Figure 1). erating at 10 Mb/s or vice-versa.

AN012598-1

FIGURE 1. Terminated Configuration

AN-1031

The complement RS-422 receiver must be equal to or less

than one unit load. This is represented by the slope of the

TRI-STATE ® is a registered trademark of National Semiconductor Corporation.

TransZorb ® is a registered trademark of General Semiconductor Incorporated.

© 1998 National Semiconductor Corporation AN012598 www.national.com

 A chart displaying the recommended operational region of a also shown for comparison. The curves were obtained from
typical RS-422 standard interface is shown in Figure 3. empirical data using a 24 AWG, copper, 16 pF/ft, twisted-pair
Other electrical interface standards operating regions are cable parallel terminated with a 100Ω load.

AN012598-3

FIGURE 3. Cable Length vs. Data Rate

TERMINATION also leads to a smaller differential output voltage and lower

Termination is recommended for use when designing a DC noise margins than with a series termination. Also, if the
RS-422 interface that is considered to be a transmission line. receiver has built-in failsafe circuitry, a known output can not
An interface is considered to be a transmission line if the be guaranteed with this type of termination scheme unless
propagation delay of the cable is greater than 1⁄8 the transi- the receiver specifically says that it supports “terminated”
tion (rise or fall) time of the signal. This is a time-domain failsafe. The cost associated with purchasing the one exter-
analysis. The same is also true in frequency-domain. nal component should also be considered.
For clarification, the transition time for the time-domain Other schemes exist to choose from and each type has its
analysis is measured from zero to one hundred percent of advantages and disadvantages. Please refer to application
the transition. The rising or falling edge may be used for note AN-903 for more detailed information on parallel and
analysis, whichever edge is the fastest. other termination schemes.
Transmission lines may restrict the use of a multi-drop con-
FAILSAFE
figuration and limit the maximum data rate of the RS-422 in-
terface. For typical RS-422 interfaces, open, terminated, and shorted
input are the three types of failsafe to consider as shown in
Transmission line theory will not be discussed in this applica-
Figure 5. A receiver with full failsafe protection guarantees a
tion note but is useful knowledge. For more details on trans-
known receiver output for all three types failsafe conditions.
mission line theory, please refer to National application notes
AN-806, AN-807, and AN-808.
Open Input Failsafe
Parallel Termination Open input failsafe is the condition when the receiver’s out-
put is known when its inputs are left open or floating. The
Parallel termination, a very popular form of termination, has
driver is not connected but the receiver is powered. A re-
the advantage of allowing higher data rates and longer cable
ceiver that provides failsafe for an open input condition, may
lengths than an interface using some other termination
provide failsafe for idle (driver’s output idle) and TRI-STATE ®
schemes because transmission line effects are minimized.
(driver’s output disabled) conditions if the bus is not parallel
This is possible because the termination resistor (Rt) is cho-
terminated. Therefore, the output state of the receiver may
sen to closely match the cable impedance (ZO) (see Figure
be known for both idle and TRI-STATE bus conditions in
4). The cable impedance can be obtained from the cable
some instances.
manufacturer. The cable impedance may also be measured
using TDR, time domain reflectometry, techniques. Addition- A receiver is able to provide open input failsafe with internal
ally, because of the minimized affect of the transmission line, pull-up and pull-down resistors, typically > 50 kΩ. Some-
multi-drop configurations with good signal quality are also times only one bias resistor is used on one input and the
possible, as long as the stubs (discussed later) are not trans- other input biased to a voltage reference point. Note that any
mission lines themselves. receiver that does not have built-in failsafe, may use external
pull-up and pull-down resistors to provide failsafe protection.
A disadvantage of parallel termination is the high power dis-
sipation associated with the heavy termination load. This

www.national.com 2
 AN012598-4

Rt = ZO
where:
Rt = termination resistor
ZO = characteristic impedance of cable
FIGURE 4. Multi-Drop Application with Parallel Termination

Terminated Input Failsafe CONFIGURATION

A receiver has terminated failsafe when its output can be de- For RS-422 interface devices, usually three types of configu-
termined while its inputs are under terminated conditions. rations are commonly used.
This must be valid for various types of termination schemes;
otherwise the receiver does not have terminated failsafe. Ad- Point-to-Point Configuration
ditionally, terminated idle line and terminated TRI-STATE line First, point-to-point, is a one driver and one receiver system.
conditions should also be supported. For RS-422 receivers Point-to-point applications may be thought of as using
with built-in open input failsafe, the protection circuitry does single-ended standards like TIA/EIA-232-E because this is
not guarantee terminated failsafe operation. If a receiver the configuration single-ended standards are popular for.
does not have this feature, external bias resistors may be However, differential standards are not restricted from use in
used to provide terminated failsafe protection. For more in- point-to-point applications. A typical point-to-point system is
formation please reference application note AN-847 that is shown in Figure 7.
devoted to this topic.
Multi-Drop Configuration
The second configuration, multi-drop, is one driver with two
or more receivers normally connected in a daisy chain lay-
out. For RS-422, the maximum number of receivers is 10 if
the receiver’s input impedance (RIN) is equal to 4 kΩ or one
unit load. If a receiver’s RIN is equal to 8 kΩ then that re-
ceiver is equal to 1⁄2 a unit load. Therefore a RS-422 driver
that can drive 10 unit loads can drive 20 receivers with a RIN
= 8 kΩ. An example of a multi-drop application is shown in
Figure 4.

Multi-Point
The last type of configuration is multi-point, which uses two
or more drivers connected to one or more receivers (see Fig-
ure 6 ). RS-422 drivers are normally not designed into this
type of configuration. However, a multi-point system can be
accomplished if certain issues are addressed. The three is-
AN012598-5 sues are ground potential differences between drivers, con-
FIGURE 5. Types of Failsafe tention between drivers, and the drive capability of the driv-
ers. Therefore, RS-485 devices are recommended for
multi-point applications.
Shorted Input Failsafe For further details and illustrations concerning these issues
Shorted failsafe is when the receiver output is known when please reference application note AN-759.
the receiver inputs are shorted together. If a receiver does
not have this type of failsafe, external pull-up and pull-down ESD PROTECTION
resistors alone may not help. However, using an Electrostatic Discharge (ESD) is normally an event of very
alternate-failsafe termination technique will provide protec- high potential for a short period of time. This may be damag-
tion. Please refer to application note AN-903 for details about ing to some integrated circuits (IC). ESD is not limited to a
alternate-failsafe termination. one time occurence in the life of an IC, unless it is fatal the
National’s DS36276 and DS36277 are interface devices that very first time. It may be an ongoing transpiration that can
provide all three forms of failsafe protection with no external wear down an IC until it eventually fails fatally.
components required. To accomplish this, the threshold point The RS-422 standard does not specify requirements for
of the receiver was shifted, violating the RS-422 ± 200 mV ESD protection. However, the industry has developed a de-
threshold specification.

3 www.national.com
 facto minimum standard of 2,000V ESD protection under hu- tion. The first method is built-in IC protection circuitry that re-
man body model (HBM) conditions. However, many systems quires no external components. This type of protection is the
today require much higher levels of ESD protection. Possibly result of the manufacturer’s IC design. In the early years of
as high as 10 kV or 15 kV. IC manufacturing, ESD was not as well publicized or stan-
ESD protection may be enhanced in different ways. IC pro- dardized as it is today. Today, IC designers strive to achieve
tection circuitry, TransZorbs ® , and protected connectors are five, ten, even fifteen thousand Volts of ESD protection.
three possible solutions for increasing an IC’s ESD protec-

AN012598-6

FIGURE 6. Example of Multi-Point Configuration

AN012598-7

FIGURE 7. A Point-to-Point RS-422 System with TransZorbs

The later two solutions, TransZorbs and protected connec- line. This solution could become expensive as the number of
tors, are both external to the IC and usually implemented, by interface lines requiring protection increases. Figure 7 shows
the system manufacturer, to provide additional protection, if an example of a RS-422 point-to-point system with
needed, beyond that which is supplied by the IC itself. TransZorbs.

Integrated Circuits Protected Connectors

IC protection circuitry is normally designed into the die at the Protected connectors, another solution for increasing ESD
input and output stages of the device. This is because these protection of an IC, provides the same type of clamping pro-
are the stages that connect to the outside world (via cables tection that TransZorbs do. However, only one device is
or other media). Thus, these are the locations damaged needed per interface connector. Therefore, if you have a
most often by ESD phenomenon. The purpose of the cir- DB-25 pin connector, all 25 lines receive the same increase
cuitry is to, one, be able to drain off very large amounts of in ESD damage resistance. Additionally, designers do not
current very quickly, and two, keep the high current away need to worry about lack of additional PC board space, since
from sensitive areas of the IC. An analogy may be made with the device is built into the connector. Also, the protected con-
lightning rods that channel away large amounts of current, nector only adds an additional 5 pF or 6 pF of capacitance to
caused by a very large electric field (potential), from your the signal load. Designers will be able to find protected con-
house. nectors in a variety of connector sizes. Figure 8 shows a dia-
gram of two connectors with built-in ESD protection. For
TransZorb more information on protected connectors please refer to ap-
TransZorbs are like back-to-back diodes. They are con- plication note AN-878.
nected between the interface line that needs protection and In the later two solutions, the system manufacturer provides
ground. They act like voltage clamps, clamping voltages that additional ESD protection. Ideally, system manufacturers
are above the TransZorb’s specified reverse stand-off volt- would like the IC manufacturers to provide all the ESD pro-
age (Vrs). Take care when selecting a TransZorb, if the Vrs is tection required for their systems, internal to the device. Na-
too high, the IC may become damaged before the TransZorb tional Semiconductor’s Interface Group recognizes the im-
ever turns on. One TransZorb should be used per interface

www.national.com 4
portance of ESD protection and has released the DS36276 voltage (VCC). First, this may cause biasing of diodes and
and the DS36277 with ESD protection up to 7,500V. Future bus clamping, or result in large current faults that damage
Interface products will trend toward higher ESD protection. devices. Second, the potential difference between ICs con-
nections and the interface connector may be large enough to
LIVE INSERTION cause ESD strikes which may be harmful to the IC. In addi-
RS-422 does not specify how to insert an IC (driver or re- tion, the IC may transmit signals when the supply voltage is
ceiver) into a live interface. The same is true for removing below minimum operating level which may result in data
the IC. PC board connectors, cable connectors, and sockets transfer errors. On the other hand, when removing an IC
are likely interface points. A device may be inserted live via from an active interface, the VCC should be within normal op-
one of these interfaces. Live insertion is a larger concern, erating levels with no device pins higher than the supply pin.
since removing an IC, from a live interface, normally is less Also, the potential difference at the instance of the discon-
damaging to an IC. For live insertion, the device may not be nect is minimized since the potential is the equivalent just
powered up when being inserted. Thus, the RS-422 device prior.
may receive bus potentials that exceed the power supply

AN012598-8

FIGURE 8. Example of Protected Connectors

Recommendations for implementing a live interface are, STUB LENGTH

one, use an IC that supports the feature. Look for this feature Since RS-422 is a multi-drop standard, receivers may be
in the device’s datasheet. However, it will not always appear connected to the bus via a stub. The length of the stub is im-
there so be sure to ask the chip manufacturer’s technical portant because of the affect it may have on the signal. As
support group, to be certain. These ICs usually have spe- the stub length is increased, its characteristics begin to act
cially designed input and output structures that prevent dam- as a transmission line. When is a stub considered to be a
age to the device even if bus voltages exceed VCC. Two, transmission line? A typical guideline for stubs and transmis-
since removal is less harmful than inserting, try to create sion lines is stated here:
similar conditions for inserting the chip that would exist when
Time Domain:
removing the chip. In other words, it is ideal to have the de-
vice powered and referenced to the same ground potential If the propagation delay of the stub (one way trip) is greater
as the interface before the inputs and outputs make physical than 1⁄8 the transition time, measured from 0% to 100% of the
contact. signal transition, then the stub may be considered a trans-
mission line.
For a PC board, a staggered connection where the ground
trace is the longest, so that it makes contact first, and the The boundary conditions at which a stub begins to act as a
VCC trace is the second longest followed by the input traces transmission line are not precise. Therefore, designers may
then the output traces last would suffice. Live insertion via use slightly different ratios.
cables or sockets may require special design but should Furthermore, the maximum length of the stub depends on
make contact in the same pattern. the transition time measured at the point of the stub intercon-
nect. This is very important to remember because if the total
BY-PASS CAPACITORS length of the cable is 1,000 feet, a longer stub can be hung
By-pass capacitors help reduce transients on output signals. off the cable at 750 feet away from the driver than at 75 feet
Therefore, by-pass capacitors are recommended for better away from the driver. This is because the cable capacitance
signal quality. One 0.1 µF capacitor is recommended for slows the transition time of the driver’s output as it propa-
each powered IC in the system. If a device has more than gates down the cable and the transition time is longer.
one power supply (i.e., VCC and VEE) then use one capacitor The length of the stub may be increased by slowing down
for each supply. The capacitor should be placed between the the transition edge at the stub interconnect. This can be
power supply pin and ground. Place the capacitor as close to done using a bulk capacitance load or a RS-422 driver with
the VCC pin of the device as possible. Additionally, a 10 µF output wave shape control like National’s DS3691, DS3692,
capacitor may be used near the main path where VCC is de- or DS36C280.
livered to the system for bulk charge storage. If a system is
large, additional bulk capacitance may be distributed across
the system.

5 www.national.com
 To implement a RS-422 interface with stubs of equal length. Receiver Input Impedance
Apply the stub guideline rule to the stub(s) closest to the driv- The receiver input impedance curve is identical for both in-
er’s output and use this length for the maximum length for all puts while the receiver is powered off. The characteristic
stubs on the interface bus. curves also pass through the (0V, 0 mA) coordinate (see Fig-
If a stub is too long and causes a noticeable reflection, it will ure 10). While the receiver is powered on, the impedance of
be measured positive on the stub because the voltage re- both inputs are the same but the input impedance curves do
flection coefficient is positive since the load impedance ZL is not pass through the (0V, 0 mA) point. Additionally, the
larger than the stub impedance Zo. The formula for the re- curves may or may not cross the x and y axis at the same
flection coefficient at the load is shown in equation 1. points (see Figure 10). For receivers with built-in open input
failsafe, the curves will differ by at least 200 mV over operat-
ing range.
The input impedance of a RS-422 receiver is guaranteed
from −10V to +10V which is the operating range for RS-422
Eq. 1. Load Reflection Coefficient receivers. The 10V is equal to the 7V common mode voltage
Note, the load impedance is equivalent to the input imped- plus 3V offset voltage.
ance of the receiver which is equal to or greater than 4 kΩ.
The stub impedance is typically about 100Ω–130Ω. The re-
flection created at the stub will propagate in both directions
away from the stub (see Figure 9). Keep in mind that reflec-
tions are time dependent events.

RECEIVER POWER OFF CHARACTERISTICS

The receivers characteristics powered down are similar to
those when powered up (see Figure 10). Therefore, the re-
ceiver, when physically connected, may be powered up or
down transparent to the RS-422 driver. AN012598-9

FIGURE 9. Stub Reflections

AN012598-10 AN012598-11

FIGURE 10. Receiver ON/OFF Characteristics

Receiver Breakdown Voltage Cable Size
The receiver breakdown voltage is guaranteed to be greater The defacto standard gauge size is 24 AWG. A range of
than ± 10V whether the receiver is powered on or off. The 22 AWG to 28 AWG is acceptable for most applications. For
breakdown point may be found by curve tracing the device cable length see “Cable Length and Data Rate” section in
out past ± 10V. Breakdown voltages may be more than twice this application note.
the maximum operating voltage.
More on Cables
TYPICAL CABLE MEDIA For more details on cable selections, including information
concerning shields, insulation, and cable characteristics, re-
Cable Type fer to application note AN-916.
Flat or round cables may be used when connecting a
RS-422 driver and receiver(s). Twisted pair or non-twisted SUMMARY
conductors may be used. Twisted pair cables have an ad- RS-422 is a standardized differential electrical interface ca-
vantage over non-twisted when transmitting differential sig- pable of transmitting data in point-to-point and multi-drop ap-
nals. Twisted pair cables couple noise more symmetrically plications. When implementing a RS-422 interface, make
than non-twisted pair cables. This helps preserve the noise sure design-in concerns are addressed at the early stages of
rejection limits of the RS-422 receiver ( ± 7V).

www.national.com 6
design to eliminate problems later in the application, which Referenced Application Notes:
may be a lot more costly. Understanding RS-422 means un- AN-759 RS-422 vs. RS-485
derstanding the issue associated with the interface.
AN-806 Transmission Lines
RS-422 is a well-defined standard and has many applica-
AN-807 Transmission Lines
tions. However, RS-422 has been superseded by RS-485
which builds onto the existing foundation and creates an AN-808 Transmission Lines
even more rugged standard. For more information on AN-847 Failsafe
RS-485, please reference AN-979. AN-878 ESD
AN-903 Termination
REFERENCES
AN-916 Cables
1. TIA/EIA-422-B, Electrical Characteristics of Balanced
Digital Interface Circuits, Electronic Industries Associa-
tion Engineering Department. Washington D.C. 1994.
2. INTERFACE: Data Transmission Databook. National
Semiconductor Corporation. Santa Clara, CA 95052.
1994.

7 www.national.com
TIA/EIA-422-B Overview

LIFE SUPPORT POLICY

NATIONAL’S PRODUCTS ARE NOT AUTHORIZED FOR USE AS CRITICAL COMPONENTS IN LIFE SUPPORT DE-
VICES OR SYSTEMS WITHOUT THE EXPRESS WRITTEN APPROVAL OF THE PRESIDENT OF NATIONAL SEMI-
CONDUCTOR CORPORATION. As used herein:
1. Life support devices or systems are devices or sys- 2. A critical component in any component of a life support
tems which, (a) are intended for surgical implant into device or system whose failure to perform can be rea-
the body, or (b) support or sustain life, and whose fail- sonably expected to cause the failure of the life support
ure to perform when properly used in accordance device or system, or to affect its safety or effectiveness.
with instructions for use provided in the labeling, can
be reasonably expected to result in a significant injury
to the user.

National Semiconductor National Semiconductor National Semiconductor National Semiconductor

Corporation Europe Asia Pacific Customer Japan Ltd.
AN-1031

Americas Fax: +49 (0) 1 80-530 85 86 Response Group Tel: 81-3-5620-6175

Tel: 1-800-272-9959 Email: [email protected] Tel: 65-2544466 Fax: 81-3-5620-6179
Fax: 1-800-737-7018 Deutsch Tel: +49 (0) 1 80-530 85 85 Fax: 65-2504466
Email: [email protected] English Tel: +49 (0) 1 80-532 78 32 Email: [email protected]
Français Tel: +49 (0) 1 80-532 93 58
www.national.com Italiano Tel: +49 (0) 1 80-534 16 80

National does not assume any responsibility for use of any circuitry described, no circuit patent licenses are implied and National reserves the right at any time without notice to change said circuitry and specifications.