Slla 067 A
Slla 067 A
Slla 067 A
ABSTRACT
The strong interest in this application report leads us to publish a second edition. In addition
to changes and updates made throughout all sections, new parts have been added, such as
M−LVDS, CAN, Clock Interfacing, and VME.
This report is intended to be a reference tool for finding the most appropriate bus interface
solution for today’s advanced system architectures. It focuses on bus interface solutions for
different transmission concepts, and gives an overview of the different bus solutions available
from Texas Instruments.
Today, advanced bus system designers have a dilemma: many different solutions can be
used to solve the same problem. The task for the designer is to figure out the optimum solution
for their special application. In order to ease this choice, this application report has been
created. The key bus interface characteristics are presented in a way that developers may
easily find the optimum bus solution for their systems.
The report is organized into product family sections to allow easy distinction among various
solutions. Each family section of this report covers details on the electrical parameters and
appropriate protocols, as well as application and feature-benefit information on the chosen
product family. All sections are set up in the same order to aid comparison.
Keywords: ABT, ABTE, AHC, ALVT, BTL, CAN (ISO-11898) , CDC, CompactPCI, FB+,
Firewire, FlatLink, Gigabit Ethernet/10G Ethernet, GTL, GTLP, HSTL, IEEE1284,
IEEE1394, LVDM, LVDS, LVT, M-LVDS, PCI, TIA/EIA-232, RS-232, TIA/EIA-422, RS-422,
TIA/EIA-485, RS-485, TIA/EIA-644, TIA/EIA-899, SONET, SSTL, USB, VME, TMS320
DSP Family.
1
SLLA067A
Contents
1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2 Single-Ended Versus Differential Data Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1 Single-Ended Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.1 Advantages of Single-Ended Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.2 Disadvantages of Single-Ended Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2 Differential Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2.1 Advantages of Differential Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.2 Disadvantages of Differential Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3 Standard Switching Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4 Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5 Parallel Versus Serial Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5.1 Advantages of Parallel Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5.2 Disadvantages of Parallel Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5.3 Advantages of Serial Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
5.4 Disadvantages of Serial Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
6 Data Transmission Topologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
6.1 Point-to-Point (Simplex) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
6.2 Multidrop (Distributed Simplex) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
6.3 Multipoint (Multiplex) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
7 IEEE 1394 Cable and Backplane Applications (FireWire) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
7.1 Electrical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
7.2 Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
7.3 Applicability and Typical Application for IEEE 1394 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
7.4 Applicability and Typical Application for IEEE 1394 Backplane . . . . . . . . . . . . . . . . . . . . . . . . . 20
7.5 Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
8 Universal Serial Bus (USB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
8.1 Electrical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
8.2 Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
8.3 Applicability and Typical Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
8.4 Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
9 TIA/EIA-232 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
9.1 Electrical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
9.2 Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
9.3 Applicability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
9.4 Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
10 TIA/EIA-422 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
10.1 Electrical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
10.2 Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
10.3 Applicability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
10.4 Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
11 TIA/EIA-485 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
11.1 Electrical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
11.2 Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
11.3 Applicability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
11.4 Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
List of Figures
1 Signaling Rate Versus Cable Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2 Single-Ended Transmission with Parallel Termination at Line End . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3 Differential Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4 Switching Levels of Single-Ended Transmission Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5 Different Interconnection Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
6 Principle of Parallel and Serial Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
7 Point-to-Point Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
8 Multidrop Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
9 Multipoint Connection Using Multiple Transceivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
10 Possible Interconnections Using 1394 as Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
11 Principle Application Setup Using 1394 Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
12 USB - Tiered Star Topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
13 Null Modem Application Using RS232 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
14 Supply Current Versus Switching Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
15 Type 1/Type 2 Switching Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
16 Parallel Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
17 Serdes Interface Using Two Transceivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
18 PCI Card for Personal Computer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
19 CompactPCI Backplane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
20 PCI Bus System With 8 PCI Buses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
21 Typical Application Showing a 1284 Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
22 Equivalent Circuit of a Single Backplane Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
23 Typical Backplane Application Using Several Plug-In Cards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
24 Principle Setup of an Open Collector Bus System Using BTL Devices . . . . . . . . . . . . . . . . . . . . . . . 56
25 Principle Setup of an Open Collector Bus System Using GTL Devices . . . . . . . . . . . . . . . . . . . . . . . 58
26 RT vs Slot Spacing With GTLP Medium and High Drive Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
27 Typical Output Buffer Environment, Class II of SSTL Standard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
28 DDR SDRAM Memory Interfacing Solution Using the SN74SSTL16857 . . . . . . . . . . . . . . . . . . . . . . 66
29 Input and Output Waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
30 Static Phase Offset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
List of Tables
1 Bus Interface Selection Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2 1394 PMD Connection Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3 Top IEEE 1394 Link Layer Device List (PC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4 Top IEEE 1394 Integrated Device List (PC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5 Top IEEE 1394 Integrated Device List (Non-PC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
6 Top IEEE 1394 Link Layer Device List (Non-PC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
7 Top IEEE 1394 Physical Layer Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
8 Top USB Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
9 Top TIA/EIA-232 Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
10 Top TIA/EIA-422 Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
11 Example TIA/EIA-485 Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
12 Example CAN Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
13 Top LVDS Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
14 Top LVDM Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
15 Top M-LVDS Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
16 Multiplexing Ratios for LVDS Serdes and FlatLinkE Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
17 LVDS Serdes Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
18 Serial Gigabit Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
19 Gigabit Ethernet/Fibre Channel Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
20 SONET/SDH Device List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
21 Top Device List − PCI Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
22 Top Device List − IEEE 1284-Compatible Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
23 Selected Characteristics for General-Purpose Logic Families . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
24 Top Feature List of Advanced System Logic by Logic Family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
25 Top Device List − Backplane Transceiver Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
26 Top Device List GTL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
27 Top Device List GTLP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
28 Maximum Data Transfer Speeds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
29 Top Device List VME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
30 Top Device List for Stub Series Termination Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
31 Recommended Clock Drivers for Serial/Gigabit Transceivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
32 Clock Drivers for Memory Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
1 Overview
In today’s information-hungry society, transmitting data over several inches between computer
memory and display screen is just as critical as sending it halfway around the globe. Over the
past three decades, Texas Instruments has combined its expertise in high-speed digital and
analog technologies to address these needs. TI is constantly pushing the capabilities and
extending the performance parameters of practically every data transmission standard.
This application report provides the reader with an overview of the different wired bus systems
and should give enough insight into which standard or which bus configuration would suit the
reader’s needs. Before we discuss each standard, its technical features, and the products that
TI offers, it is important to understand the fundamentals behind the different bus configurations
available.
Data transmission, as the name suggests, is a means of moving data from one location to
another. Several important parameters define how the information is transferred. Two of these
are the distance, i.e. the space between the sending and the receiving systems and speed, i.e.
the rate at which data has to be passed to the receiving device. Different transmission stan-
dards, such as TIA/EIA−232, IEEE 1394 and LVDS, provide solutions for various needs in terms
of speed and line length as defined by Figure 1.
10000
CML/ECL
LVDS
CML/ECL
1000
M-LVDS
LVDS
Signaling Rate − Mbps
LVDS
100 GTLP
485/422
M-LVDS
GTL/BTL
CMOS/TTL
10
1
485/422
0.1
0.1 1 10 100 1000 10000
Distance − Meters
It can be seen from Figure 1, that as the cable length increases, the speed at which the informa-
tion is transmitted must be lowered in order to keep the bit-error rate down. Therefore, it is very
important to choose the correct standard covering the required communication distance and the
needed data rate. Once the choice of standard has been made, the selection of the device re-
quired for the specific application can begin. This is certainly not an easy task.
The matrix on the following page has been generated to give a more accurate overview of the
different kind of transmission types, the modes, standards, distances, data rates, its benefits,
and which TI family would suit which purpose. For every family, Texas Instruments has created a
section within this application report describing the characteristics of the standard, and whether
or not there is a software overhead. There is, of course, also a family-dedicated web page, con-
taining further information on product data sheets, additional application reports, and the possi-
bility of sampling devices via the Internet. More information is available at TI’s web page:
http://interface.ti.com.
There are currently two forms of operation for electrical interface circuits, single-ended (or unbal-
anced) and differential (or balanced) data transmission. The difference between single-ended
and differential transmission is described below.
Single-ended transmission is performed on one signal line, and the logical state is interpreted
with respect to ground. For simple, low-speed interfaces, a common ground return path is suffi-
cient; for more advanced interfaces featuring higher speeds and heavier loads, a single return
path for each signaling line (twisted pair cable) is recommended. Figure 2 shows the electrical
schematic diagram of a single-ended transmission system.
Driver Receiver
VCM
Signal Line
Data IN Data OUT
VO VIN
VGND
The advantages of single-ended transmissions are simplicity and low cost of implementation. A
single-ended system requires only one line per signal. It is therefore ideal for cabling, and when
connector costs are more important than the data transfer rate, e.g. PC, parallel printer port or
serial communication with many handshaking lines, e.g. EIA-232. Cabling costs can be kept to a
minimum with short distance communication, depending on data throughput, requiring no more
than a low cost ribbon cable. For longer distances and/or noisy environments, shielding and
additional ground lines are essential. Twisted-pair cables are recommended for line lengths of
more than 1 meter. The additional efforts might outweigh the previously described
cost-advantages.
The main disadvantage of the single-ended solution is its poor noise immunity. Because the
ground wire forms part of the system, transient voltages or shifts in voltage potential may be in-
duced (from nearby high frequency logic or high current power circuits), leading to signal degra-
dation. This may lead to false receiver triggering. For example, a shift in the ground potential at
the receiver end of the system can lead to an apparent change in the signal, sufficient to drive
the input across the thresholds of the receiver, thus increasing its susceptibility to electromag-
netic fields.
Crosstalk is also a major concern especially at high frequencies. Crosstalk is generated from
both capacitive and inductive coupling between signal lines. Capacitive coupling tends to be
more severe at higher signal frequencies as capacitive reactance decreases. The impedance
and termination of the coupled line determines whether the electric or the magnetic coupling is
dominant. If the impedance of the line is high, the capacitive pickup is large. Alternatively, if the
line impedance is low, the series impedance as seen by the induced voltage is low, allowing
large induced currents to flow. Single-ended transmission is much more susceptible to external
noise and the radiation of EMI is increased compared to differential systems.
In addition to above described noise sources, line reflections have to be considered as well. This
is not only important for long transmission lines, short line lengths show critical over- and under-
shoot on non terminated lines as well. The higher the speed and the faster the rising and falling
edges of the output signal are, the more critical is the line length and even lengths of some
centimeters might need to be terminated.
These problems will normally limit the distance and speed of reliable operation for a
single-ended link.
For balanced or differential transmission, a pair of signal lines is necessary for each channel. On
one line, a true signal is transmitted, while on the second one, the inverted signal is transmitted.
The receiver detects voltage difference between the inputs and switches the output depending
on which of the two input lines of a channel is more positive. As shown in Figure 3, a signal re-
turn path is required.
VCM
Diff. Driver Diff. Receiver
Signal Line
VGND
Balanced interface circuits consist of a generator with differential outputs and a receiver with dif-
ferential inputs. Better noise performance stems from the fact that noise is coupled into both
wires of the signal pair in the same way and is common to both signals. Due to the common-
mode rejection capability of a differential amplifier, this noise is rejected. Additionally, since one
signal line emits the opposite signal of the other, the emissions cancel each other. This is true in
any case for crosstalk from and to neighboring signal lines.
The twisted-pair cable used in these interfaces, in combination with a correct line termination to
avoid line reflections, allows very high data rates. Signaling rates of up to 10 Gbs are possible
with differential signaling.
3.0 −3.6 V
3 VIH
VCC
2.4 VOH VCC 2.4 VOH
2.4 VOH 0.8 V CC VOH
2.0 VIH 2.1
2.0 VIH
0.7 x V CC VIH 1.6 VIH 1.62 VIH
1.5 Vt 1.5 Vt 1.5 Vt 1.55 Vt 1.5 VOH
vcc 1.4 1.47 VIL VIH
Vt VIL 1.2 1.05
1.1 VOL VIH Vt
2 1 0.85 1
0.8 VIL 0.8 VIL 0.8 Vt 0.95 VIL
0.2 x V CC VIL
0.6 VOL 0.75 VIL 0.55 VOL
0.4 VOL 0.4 VOL 0.4 VOL
0.2 V CC VOL
0 GND 0 GND 0 GND 0 GND 0 GND 0 GND 0
GND
TTL LVTTL CMOS ETL BTL/FB GTL GTL+GTLP
−3 VIL
VCC = Supply Voltage
VOH = Minimum High-Level Output Voltage
VIH = Minimum High-Level Input Voltage
Vt = Nominal Input Voltage Threshold
VIL = Maximum Low-Level Input Voltage
VOL = Maximum Low-Level Output Voltage
GND = Ground
−5 VOL
TIA/EIA−232−F(RS−232)
4 Connectivity
When designing a data transmission system, bus-system designers have numerous ways of
connecting the transmitters, receivers, boards, backplanes and racks. Figure 5 shows how, in a
typical backplane rack, the cards can be linked up. The backplane itself consists of a parallel
bus type configuration with each card slotted onto that bus via a connector. In some situations, it
is necessary to connect either individual cards together, or a backplane to a card, etc. It would
be very inconvenient to connect these devices together via a parallel cable. Serial techniques
supporting a very high data rate cope with this kind of need.
Between
Cards
On the
Backplane
Between
Cabinets
Between
Racks
Backplane
Connection
LVDS and IEEE1394b are two of the latest standards that have been developed for high-speed
point–to–point serial transmission. What are the benefits of serial transmission compared to par-
allel transmission? The next section explains the principle behind the two different techniques
and outlines the benefits and shortcomings of both of them.
In the case of serial transmission, the data must first be converted to a serial stream. This is
called the serialization. The serial data is then transmitted at high speed along the line to the re-
ceiver, which must then deserialize the information back into the original parallel data. In order
for the serial technique to achieve the same data rates as the parallel one, the data must be sent
along the line at a much higher speed than on the parallel bus. Both methods have their advan-
tages and disadvantages.
Parallel Data Parallel Data
bit1
bit2
bit3
Parallel Data
bit1 bit
n−1
bit2 bitn
bit3
bit Serialization
n−1 Serial Data
D R D R
D D R
R R R R R R
D D D D
R R
R R
R R R
D D D R R R
D D D
7.1 Electrical
The 1394 standard is a packet-transaction-based technology for cable- or backplane-based en-
vironments. Both chassis and peripheral devices can use this technology. The 1394 serial bus is
organized as if it were memory space interconnected between devices, or as if devices resided
in slots on the main backplane. Device addressing is 64 bits wide, partitioned as 10 bits for bus
ID, 6 bits for node ID and 48 bits for memory addresses. The result is the capability to address
up to 1023 buses, with each having up to 63 nodes, each with 281 terabytes of memory.
Memory–based addressing, rather than channel addressing, views resources as registers or
memory that can be accessed with processor-to-memory transactions. Each bus entity is termed
a unit, to be individually addressed, reset, and identified. Multiple nodes may physically reside in
a single module, and multiple ports may reside in a single node.
Some key features of the 1394 topology are multi-master capabilities, live connect/disconnect
(hot plugging) capability, genderless cabling connectors on interconnect cabling, and dynamic
node address allocation as nodes are added to the bus. Another feature is that transmission
speed is scalable from approximately 100 Mbps to 800 Mbps. Future devices shall support as
well 1.6 Gbps and 3.2 Gbps.
Nodes may act as a repeater, allowing nodes to be chained together to form a relatively unre-
stricted topology. 1394-1995 and 1394a-2000 based cable PHYs (physical layer devices) use
cables with power wires. Equipment/devices meeting the 1394 standard that are not switched on
can still power their PHYs from the 1394-bus, thus acting as a repeater or hub. Equipment with
low power consumption can be powered directly from the 1394 bus. By using the maximum
number of 16 hops (1394a-2000 allows 24 hops) with regular 4.5-m cables, the maximum end-
to-end distance of 72 m is reached. 1394b-2002 defines new transmission media and modula-
tion types. Data can now be transmitted over several types of optical cables and low cost twisted
pair cables. Heterogeneous networks allow optimum solutions in terms of cost (use electrical
cable where possible; use optical cables where needed). While remote powering across optical
cable sections is no longer possible with 1394b, you may ask if this standard extension is still
backward compatible to 1394-1995. Well, you cannot connect an optical cable with an electrical
outlet. In such configurations you need a simple cable converter. Bilingual PHYs like the
TSB81BA3 are used as translators between the 1394-1995 and 1394a-2000 world (that use DS
coding) and the 1394b world (that use 8B10B coding).
DS coding in 1394-1995 and 1394a is a method that uses common mode DC signals for ar-
bitration and speed signaling. An encoded clock that can easy be recovered simplifies fast
detection of data with different speeds.
8B10B coding in 1394b allows DC-free data transmission which is needed when transformers
or capacitors (to separate ground domains) are used in the signal path. The bit representation is
the same for electrical or optical, it just needs an optical transceiver to go optical (LED driver and
LED on TX, photodiode and sense amplifier on RX).
It’s simple! if you can connect it with cables, it is compatible back and forth (from 100 Mbps to
3.2 Gbps).
Speed Reach 100DS 200DS 400DS 100 200 400 800 1600
STP 4.5m 1394- 1394- 1394- 1394b 1394b 1394b
(1394cbl) 1995 1995 1995
UTP−5 100m 1394b
POF 50m 1394b 1394b
HPCF 100m 1394b 1394b
MMF 100m 1394b 1394b 1394b
Remote- na yes yes yes on on on
Power STP STP STP
7.2 Protocol
Both asynchronous and isochronous data transfers are supported. The asynchronous format
transfers data and transaction layer information to an explicit address. The isochronous format
broadcasts data based on channel numbers rather than specific addressing. Isochronous pack-
ets are issued on the average of each 125 µs in support of time-sensitive applications. Providing
both asynchronous and isochronous formats on the same interface allows both non-real-time
and real-time critical applications on the same bus.
The cable environment’s tree topology is resolved during a sequence of events triggered each
time a new node is added or removed from the network. This sequence starts with a bus reset
phase, where previous information about a topology is cleared. The tree ID sequence deter-
mines the actual tree structure, and a root node is dynamically assigned, or it is possible to force
a particular node to become the root. After the tree is formed, a self-ID phase allows each node
on the network to identify itself to all other nodes. During the self-ID process, each node is as-
signed an address. After all of the information has been gathered on each node, the bus goes
into an idle state waiting for the beginning of the standard arbitration process.
An additional feature is the ability of transactions at different speeds to occur on a single device
medium (for example, some devices can communicate at 100 Mbps while others communicate
at 200 Mbps, 400 Mbps or 800Mbps). Use of multispeed transactions on a single 1394 serial
bus requires consideration of each node’s maximum capabilities when laying out the connec-
tions to ensure that the path between two higher–speed nodes is not restricted by a device with
lower–rate capabilities.
The backplane physical layer shares some commonality with the cable physical layer. Common
functions include: bus state determination, bus access protocols, encoding and decoding func-
tions, and synchronization of received data to a local clock.
The TSB14AA1 and SN74GTLP1394 provide a way to add high-speed 1394 connections to
almost any backplane.
PHY Chip
Link Chip
Traces DATA
Terminators Connectors STROBE
Backplane-Specific
Transceivers
7.5 Features
• Real-time data transfer for multimedia applications
• 100, 200, and 400 Mbps data rates for high-speed applications in cable environments at
1394.a (2000) and 800 Mbps data rates at 1394.b
• 25 and 50 Mbps data rates for backplane environments
• Live connection/disconnection without data loss or interruption
• Automatic configuration supporting plug and play
• Free network topology allowing branching and daisy-chaining
• Assured bandwidth assignments for real-time applications
8.1 Electrical
The USB physical interconnect is a tiered-star topology. A hub is at the center of each star. Each
wire segment is a point-to-point connection between the host and a hub or function, or a hub
connected to another hub or function. The USB transfers signal and power over a four-wire
cable. The signaling occurs over two wires on each point-to-point segment.
There are three data rates: full-speed signaling bit rate at 12 Mbps, a limited capability low-
speed signaling mode at 1.5 Mbps and high-speed mode at 480 Mbps. All modes can be sup-
ported in the same USB bus by automatic dynamic mode switching between transfers. The low-
speed mode is defined to support a limited number of low-bandwidth devices, such as a mouse,
because more general use would degrade bus utilization. The clock is transmitted, encoded
along with the differential data. The clock encoding scheme is NRZI with bit stuffing to ensure
adequate transitions. A SYNC field precedes each packet to allow the receiver(s) to synchronize
their bit recovery clocks. The cable also carries VBUS and GND wires on each segment to deliv-
er 5 V power to devices.
8.2 Protocol
The USB is a polled bus. The host controller initiates all data transfers.
All bus transactions involve the transmission of up to three packets. Each transaction begins
when the host controller, on a scheduled basis, sends a USB packet describing the type and
direction of transaction, the USB device address, and endpoint number. This packet is referred
to as the token packet. The addressed USB device selects itself by decoding the appropriate
address fields. In a given transaction, data is transferred either from the host to a device or from
a device to the host. The direction of data transfer is specified in the token packet. The source of
the transaction then sends a data packet or indicates it has no data to transfer. The destination,
in general, responds with a handshake packet indicating whether the transfer was successful.
The USB data transfer model between a source or destination on the host and an endpoint on a
device is referred to as a pipe. There are two types of pipes: stream and message. Stream data
has no USB-defined structure, whereas message data does. Additionally, pipes have
associations of data bandwidth, transfer service type, and endpoint characteristics like
directionality and buffer sizes. Most pipes are created when a USB device is configured. One
message pipe, the default control pipe, always exists once a device is powered, in order to
provide access to the device configuration, status, and control information.
The transaction schedule allows flow control for some stream pipes. At the hardware level, this
prevents buffers from underrun or overrun situations by using a negative acknowledgment
handshake to decrease the data rate. When negative acknowledged, a transaction is retried
when bus time is available. The flow control mechanism permits the construction of flexible
schedules that accommodate concurrent servicing of a heterogeneous mix of stream pipes.
Thus, multiple stream pipes can be serviced at different intervals and with packets of different
sizes.
USB Hub
Monitor USB Hub
Keyboard
Mouse
Printer
Scanner
Modem Joy Stick
Phone/Answering
8.4 Features
• Single PC supports up to 127 devices
• Data rates of 1.5 Mbps,12 Mbps, and 480 Mbps supported
• Ends confusion of multiple add-in cards for ease of use
• Universal connectors and cables for all devices and applications
• Auto configuration upon connection for real plug and play
• Assured bandwidth for real-time applications
9 TIA/EIA-232
Usually, it is based on or is identical to other standards, e.g., EIA/TIA-232-F. The following sec-
tion gives a closer view of these specifications, their benefits, and applicability. Where appropri-
ate, the protocol is briefly described.
TIA/EIA-232, previously known as RS-232 was developed in the 1960’s to interconnect layers of
the interface (ITU−T V.11), but also the pin-out of the appropriate connectors (25-pin D-type or
9-pin DB9S) (ISO 2210) and the protocol (ISSUED-T V.24). The control lines data carrier detect
(DCD), data set ready (DSR), request to send (RTS), clear to send (CTS), data terminal ready
(DTR), and the ring indicator (RI) might be used, but do not necessarily have to be (for exam-
ple, the PC serial mouse utilizes only RI, TD, RD and GND). Although the standard supports
only low-speed data rates and line length of approximately 20 m maximum, it is still widely used.
This is due to its simplicity and low cost.
9.1 Electrical
TIA/EIA-232 has high signal amplitudes of ± (5 V to 15 V) at the driver output. The triggering of
the receiver depends on the sign of the input voltage: that is, it senses whether the input is
above 3 V or less than −3 V. The line length is limited by the allowable capacitive load of less
than 2500 pF. This results in a line length of approximately 20 m. The maximum slope of the
signal is limited to 30 V/µs. The intention here is to limit any reflections that can occur to the rise-
and fall-times of the signal. Therefore, transmission line theory does not need to be applied, so
no impedance matching and termination measures are necessary. Due to the voltage swings of
-5...15 to 5...15 V, a dual supply voltage was necessary in the past. Many devices operate with
single supplies, generating the large positive and negative driver output voltage swings with
integrated charge-pumps.
9.2 Protocol
Different from other purely electrical-layer standards, TIA/EIA-232 defines not only the physical
layer of the interface (ITU-T V.11), but also the pinout of the appropriate connectors (25-pin D-
type or 9-pin DB9S) (ISO 2210), and the protocol (ITU-T V.24). The interface standard also
specifies handshake and control lines, in addition to the unidirectional receive data line (RD) and
the transmit data line (TD).
9.3 Applicability
TIA/EIA–232 is historically associated with computers interfacing with peripherals at low speed,
short distance, for example, mouse, modem, joystick, etc., or to interconnect two PCs (that is,
null modem, Figure 13). Today other equipment also uses the TIA/EIA-232 I/O. One such
example is for programming purposes.
NULL
MODEM
9.4 Features
• Established standard
• Inexpensive
• Universally used
10 TIA/EIA-422
TIA/EIA-422 (RS-422) allows a multidrop interconnection of one driver, transmitting
unidirectional to up to 10 receivers. Although it is not capable of bidirectional transfer, it is still
applicable and used for talker-audience scenarios.
10.1 Electrical
TIA/EIA-422 (ITU-T V.11) is limited to unidirectional data traffic and is terminated only at the line
end opposite to the driver (refer to the multidrop explanation at the beginning of this report). Line
length to 1200 m is possible and, over shorter distances, signaling rates of 10 Mbps are
common. TIA/EIA-422 allows up to ten receivers (input impedance of 4 kΩ) attached to one
driver. The maximum load is limited to 80 Ω while the output needs to supply an amplitude of 2 V
minimum. The receivers of TIA/EIA-422 detect the bus state with as little as 200 mV differential
and up to 7 V of common-mode signal.
10.2 Protocol
Not applicable/none specified
10.3 Applicability
TIA/EIA-422 is most commonly used in DTE-to-DCE interface in telecommunications but has
general applicability to any point-to-point or multidrop bus where a direct connection is desired in
the presence of ground noise.
10.4 Features
• Established standard
• Good for multidrop over long distances
• High noise immunity and low radiated emissions
11 TIA/EIA-485
TIA/EIA-422 was published before TIA/EIA-485. Due to the lack of bidirectional capabilities al-
lowing for multipoint connections, the TIA/EIA-485 standard was created to add this feature. The
new standard (TIA/EIA-485 or ISO/IEC 8284) defines the electrical characteristics of the inter-
connection, including driver, line, and receiver. It allows data rates in the range of 35 Mbps and
above, and allows line lengths of up to 1200 m. Of course, both limits cannot be reached at the
same time. Recommendations are given regarding wiring and termination. The standard does
not specify the connector or any protocol requirements.
11.1 Electrical
TIA/EIA-485 describes a half-duplex, differential transmission method designed for twisted-pair
cables and other balanced media. The standard requires drivers to deliver a minimum
differential output voltage of 1.5 V with up to 32 unit loads of about 12 kΩ each, plus termination
resistors at each end of the bus. Connection of more than 32 nodes is possible if fractional
unit-load devices are used. The common-mode voltage levels on the bus may vary between
–7 V and 12 V and receivers must be sensitive enough to determine the bus state based on a
differential signal level of 200 mV.
11.2 Protocol
No particular protocol is specified in the standard. However, many popular protocol standards
reference RS-485 as an electrical-layer solution. These include Profibus (EN 50170), Interbus-S,
and MODBus.
11.3 Applicability
Due to its differential transmission form, 485 is robust against electrical noise corruption. Due to
its wide common-mode voltage range, it is tolerant to ground potential shifts between nodes. For
both of these reasons, this standard is perfectly suited for applications requiring low noise
emissions and susceptibility. This is especially valuable in long lines. In most applications, the
signaling rate is sufficient, for example, to control a process line. TIA/EIA-485 is also used in
backplane connections, as its high node-count and data integrity are necessary features for
these applications.
11.4 Features
• Very robust interface (common-mode range: −7 V to 12 V)
• Low radiated emissions
• High noise tolerance
• Multiplex mode of operation
• Up to 256 nodes with 1/8th UL transceivers
Table 11. Example TIA/EIA-485 Device List
DEVICE FUNCTION POWER PACKAGE
SN65HVD1x Single channel transceivers VCC: 3.3 V 8–pin SOIC
I/O: (LV)TTL/RS–485
SN65HVD2x Single channel, wide common mode transceiver, SN65HVD23, VCC: 5 V 8–pin SOIC
SN65HVD24 with receiver equalization I/O: (LV)TTL/RS–485
CAN is ideally suited in applications requiring a large number of small messages in a short
period of time, with high reliability in rugged operating environments. Since CAN is message
based and not address based, it is especially suited when data is needed by more than one
location and system-wide data consistency is considered mandatory.
12.1 Electrical
The CAN standard ISO 11898 defines half-duplex, differential transmission on cable lengths up
to 40 m, a maximum stub length of 0.3 m, and a maximum of 30 nodes at signaling rates up to
1 Mbps. However, with careful design, longer cables, longer stub lengths, and many more nodes
can be added to a bus—always with a trade-off in signaling rate.
A single shielded or unshielded twisted-pair cable with a 120-Ω characteristic impedance (Zo) is
specified as the interconnect. This interconnecting cable is terminated at both ends with a
resistor equal to the Zo of the line. Nodes are then connected to the bus with unterminated drop
cables, or stubs.
Data transmission circuits employing CAN are used in applications requiring a rugged
interconnection with a wide common-mode voltage range. Therefore, shielded cable is
recommended in these electronically harsh environments, and when coupled with the
Standard’s −2-V to 7-V common-mode range of tolerable ground noise, helps to ensure data
integrity.
12.2 Protocol
The CAN communications protocol, ISO 11898, describes how information is passed between
devices on a network, and conforms to the open systems interconnection (OSI) model that is
defined in terms of layers. ISO 11898 defines the functions of the lower layers of the OSI model
that are performed within a CAN controller. These CAN controller functions are designed for in-
teraction with a higher layer protocol such as CANopen to complete master and slave node
communication relationships for bus operation. There is an abundance of plug-n-play CAN pro-
gramming routines available on the internet from different vendors.
12.3 Applicability
The robust features of CAN make it ideally suited for the many rugged applications to which the
CAN protocol is being adapted. Among the applications finding solutions with CAN are
automobiles, trucks, trains, busses, airplanes and aerospace, agriculture, construction, mining,
and marine vehicles. CAN-based control systems are being used in factory and building
automation and embedded control systems for machines, medical devices, domestic appliances
and many other applications.
12.4 Features
• Established standard and standardized plug-n-play products
• Fault tolerant—very high short-circuit protection
• Self-diagnosing error repair
• Very robust interface for harsh electrical environments
• Wide common-mode range
13 TIA/EIA-644 (LVDS)
Low voltage differential signaling (LVDS) is an approach to achieve higher signaling rates on
commonly used media. Since the limitation of the previously known differential standards is
mainly related to the maximum achievable slew rate and EMI restrictions, LVDS targets
low-voltage swings that are reached much faster, enabling higher speeds, lower EMI , and lower
power consumption.
13.1 Electrical
As the name says, LVDS utilizes differential transmission mode and low signal amplitudes. The
swing is in the range of only 350 mV, generated on a 100-Ω termination resistor. The driver may
also be a current-mode driver forcing an output current between 2.47 mA and 4.54 mA into
either one of the two outputs. This benefits in higher speed, lower power consumption and
reduced EMI. More information on EMI comparisons can be found at http://www.ti.com. Search
on literature number SLLA030. The common-mode range of LVDS is specified as 0 V to 2.4 V,
but devices are available that allow for larger offsets (−4 V…5 V). The chart in Figure 14 shows
the dependency of the supply current versus the switching frequency for HVD (TIA/EIA-422),
LVDS and PECL. As can be seen, the supply current is significantly lower and also the increase
over frequency is much lower with the use of LVDS drivers.
180
PECL
140
120
I CC − Supply Current − mA
100
80
60
422
40
LVDS
20
0
0 50 100 150 200
f − Frequency − MHz
The initial specification, TIA/EIA-644, addresses only point-to-point interfaces, but it allows the
attachment of multiple receivers if attention is paid to several conditions, including stub line
length, termination, and signaling rate. Revision A of the standard includes the multidrop option.
Most recent devices support transfer rates in the Gbps range (Gigabits per seconds).
13.2 Protocol
TIA/EIA-644 does not specify protocol, but is referenced by higher-level standards.
13.3 Applicability
LVDS targets applications that transfer data point–to–point at very high speeds. It is also capa-
ble of driving multiple receivers if care is taken regarding the load configuration. LVDS is particu-
larly suitable for any application that requires low power and/or low EMI. The common-mode in-
put voltage, which is currently limited by the standard to 0 V to 2.4 V, restricts the usage of LVDS
with long line lengths (which may cause unpredictable ground shifts), as well as the use in elec-
trically noisy environments. Using Texas Instruments wide VICR devices with a significantly in-
creased input voltage range and integrated hysteresis, the applicability in those environments
becomes suitable.
In the high-speed range, TI offers repeaters and/or converters for LVDS−PECL / PECL−LVDS
translation as well as PECL−CML, LVDS−CML and vice versa. Conversion between LVDS and
other levels can be achieved with external circuitry. More detailed information can be found in
the application reports SCAA062 and SCAA059 several data sheets, e.g., http://www−s.ti.com/
sc/ds/sn65lvds104.pdf . LVDS is found in clock and data distribution, backplane and cable trans-
mission, level conversion and many more applications.
13.4 Features
• Very high speed
• Very low power consumption
• Very low EMI
• Low cost
14 LVDM
LVDM stands for LVDS-multipoint, enabling a half-duplex operation with LVDS voltage levels
and speeds. It benefits from the same advantages as LVDS, and additionally allows bidirectional
data transfer, and the attachment of several drivers, receivers, and/or transceivers. LVDM is a
LVDS-multipoint option available from TI exclusively. Meanwhile, a new standard has been re-
leased, following the industry demand for standardized Multipoint-LVDS, including definitions
such as bus contention prevention, etc., which are of high importance when mixing devices of
multiple suppliers. Please refer to next section, M-LVDS. Devices complying with the new stan-
dard provide a higher output drive capability than LVDM. However, for some applications where
power consumption and heat sinking are concerns, LVDM might still be an option over M-LVDS.
Existing products remain available.
14.1 Electrical
LVDM is LVDS with doubled driver output current. For bidirectional transfers, a termination is
needed at each end of the line. The requirement of termination matching the line impedance re-
mains. Therefore, both ends are terminated with 100 Ω. The effective termination resistance re-
sults in the parallel configuration of these two resistors, equaling half the impedance or twice the
load. To ensure the same input voltage levels, such as with LVDS (i.e. at a 100-Ω load), the out-
put current is doubled to generate the required amplitude on loads of just 50 Ω.
14.2 Protocol
Not applicable/none specified.
14.3 Applicability
LVDM devices are suitable for half-duplex (bi-directional) communications between two nodes at
up to 655 Mbps and distances up to 30 m (depending on ground noise).
14.4 Features
• Drives double terminated multipoint
• Very high speed
• Very low power consumption
• Very low EMI
• Low cost
15 M-LVDS
After recognizing the benefits of LVDS, users soon demanded support of the multipoint mode of
operation with LVDS-like signals. The technical committees of TIA responded with the
development and publication of TIA/EIA-899, multipoint-LVDS or M-LVDS. M-LVDS modifies the
electrical specifications of LVDS to accommodate the lower-impedance loads, stubs, failsafe
issues, and other problems unique to the multipoint bus structure.
15.1 Electrical
The differential output voltage of a standard M-LVDS driver is 480 mV to 650 mV across a 50-Ω
test load and, with a 50-mV receiver threshold, provides a minimum of 430 mV of differential
noise margin. This margin allows for loaded bus impedance as low as 30 Ω and reflection
coefficients of −0.25. The receiver common-mode input voltage may range from –1 V to 3.4 V.
TIA/EIA-899 specifies two types of receivers. Type-1 receivers have an input threshold near zero
volts, while the Type-2 receiver threshold is nominally 100 mV as shown in Figure 15. The option
of the Type-2 receiver allows idle-line failsafe to be implemented at the cost of noise margin.
The driver and input characteristics of M-LVDS interfaces are specified such that 32 worst-case
loads may be connected to a single bus segment. To allow for the stubs created by these
connections, the driver output transition times are limited to be no faster than 1 ns. Additionally,
the standard suggests that silicon providers offer different signaling-rate options, so customers
can choose according to their specific requirements (limit EMI, allow for longer stub lines, etc).
Type 1 Type 2
2.4 V
High High
150 mV
VO
50 mV
0V
−50 mV
Low Low
2.4 V
Transition Region
15.2 Protocol
Not applicable/none specified.
15.3 Applicability
M-LVDS allows multipoint connection of up to 32 nodes and signaling rates to 500 Mbps. These
features make M-LVDS applicable to backplane or cabled data transmission where single-ended
or TIA/EIA-485 signaling rates, power consumption, or electromagnetic interference is
unacceptable. The common-mode range of M-LVDS likely limits application to cables of
30 meters or less but depends upon the environment.
15.4 Features
• High speed
• Low power consumption
• Low EMI
• Up to 32 bus connections
• Slew-rate-limited driver outputs
• Type-2 receivers available for idle-line failsafe
• Loaded bus impedances as low as 30 Ω
• Capable of live-insertion
Table 16. Multiplexing Ratios for LVDS Serdes and FlatLink Devices
INPUT (LVTTL) OUTPUT (LVDS)
Transmitter 10 data lines + clock line 1 line with embedded clock
21 data lines + clock line 3 data lines + clock line
28 data lines + clock line 4 data lines + clock line
Receiver 1 line with embedded clock 10 lines + clock line
3 data lines + clock line 21 lines + clock line
4 data lines + clock line 28 lines + clock line
16.1 Electrical
The parallel I/O bus is LVTTL based. The high-speed serial transmission uses LVD-Signaling
while meeting or exceeding the ANSI EIA/TIA-644 Standard. The devices require a single 3.3-V
power supply.
16.2 Protocol
FlatLink devices are optimized for video data transmission (SVGS, XGA and SXGA) between
display processor and flat panel display.
There are no specific protocol requirements for using 10-bit LVDS Serdes devices. However, the
transmitters are generating a start and a stop bit between each 10 bits sent over serial link. The
start/stop bits ensure correct byte boundary on the parallel receiver bus, and sufficient transition
density for the clock recovery circuit.
16.3 Applicability
FlatLink: This technology is designed to transfer large data packets from the CPU to the
graphic LCD display in a notebook type of application. The major advantages, compared with
realizing this transfer based on standard bus logic, are in reduced number of parallel data lines,
reduced power consumption, and lower electromagnetic emission.
LVDS Serdes: The differential transmission mode provides high robustness against in-coupling
noise. Therefore, LVDS serdes devices are perfectly suited as point-to-point interconnections
between two parallel buses in telecom backplanes, or industrial applications.
Bus speed: The clock rate of those bus systems must be chosen within a defined range. The
limiting factor for minimum and maximum clock speed is the fully integrated PLL of the
transmitter that needs to lock onto the input clock. The following frequency ranges are supported
by TI solutions:
Initialization after reset/power down: As typical in all PLL based applications, the receiver
requires a certain time to lock to the clock signal. Therefore, the result on the receiver output
becomes valid after approximately 1 ms.
16.4 Features
• Data throughput up to 660 Mbps over one differential pair
• Industrial temp range (−40°C to 85°C) for LVDS serdes
• Very low power consumption and low emission (portable devices)
• Using a daisy-chain layout offers multidrop architecture ability
• Low power consumption when disabled (below 1 mW)
FlatLink
17.1 Electrical
The parallel interface of TI’s 16:1 general purpose gigabit devices consists of 16 transmit data
lines latched on rising edge of a corresponding clock signal and two control signals enabling
sending special characters. The receive path is designed in a similar manner: 16 data bits and
clock plus two status bits (Figure 16).
The 18:1 general purpose gigabit transceiver accepts 18 bits of data with clock signal on its par-
allel transmit interface. On the receiving path, 18 bits with recovered clock are present.
2 Gigabit Switch/Controller TLK2501
TXD[15:0]
TX_EN
GTX_CLK
TX_ER
RXD[15:0]
PCS/PMD
Functionality: The following steps are performed on the transmitter side of a 16-bit general pur-
pose gigabit transceiver:
1. The incoming bus signal is buffered in reference to the input clock signal
2. 8B10B encoding of the data word (ensuring sufficient transition rate of equal/above three
or more transitions on 10-bit data for clock recovery purposes, as well as DC balance of
serial data stream, allowing AC coupling)
3. Serialization of the data
4. Clock the serial data into the differential line using the CML/VML driver
The receiver side performs these operations in reverse order, regenerating the original sent
word.
Copper or
Tranceiver 1 optical Transceiver 2
Transmission media: The transmission media of the serial path can be PCB, copper cable, or
fiber-optics using an optical module between transceivers. The single-ended impedance of the
copper must be matched to 50 (75) Ω. The maximum distance to be bridged depends on the
transmission speed and the transmission media. Tests have shown proper functionality at 2.5
Gbps over distances up to 10 m of xTP Cat5 cable.
17.2 Protocol
There is no protocol necessary for usage of TI’s general-purpose gigabit transceivers.
Using special carrier words and an initialization algorithm, the on-chip state machine of a 16-bit
transceiver allows correct data transfer after reset, or data loss. It also ensures the boundary of
the parallel output data word out of the incoming serial bit stream. Status outputs on the receiver
inform the user about the actual transmission state, as it can be normal data transfer, transmis-
sion error, idle, or carrier extension.
There no specific protocol requirements for using 18-bit general-purpose serdes devices. How-
ever, the transmitter adds a start and a stop bit between each 18 bits sent over the serial link.
This start/stop bits ensure correct byte boundary on the parallel receiver bus and sufficient tran-
sition density for the clock recovery circuit.
17.3 Applicability
The differential transmission mode provides high robustness against coupled noise. Therefore,
serdes devices are suited for networking, telecommunication, or data-communication designs.
Bus speed: The clock rate of those bus systems can be chosen within a certain range. The
limiting factor on minimum and maximum clock speed is the bandwidth of the fully integrated
PLL (e.g., 80 -125 MHz).
Bus Topology: Gigabit serdes can be used for point-to-point connections.
Initialization after reset/power down: As typical in all PLL-based applications, the receiver
requires a certain time to lock to the incoming serial signal. During this time the receiver output
and control pins are held in a high-impedance state for up to one millisecond.
Adjustable voltage swing: The serial output driver offers an adjustable voltage swing being
used for optimizing a particular transmission line impedance and length, as well as for controlling
the output swing for EMI and attenuation concerns.
17.4 Features
• Data throughput up to 2.5 Gbps over one differential pair
• Very low power consumption, low emission
• External pins to signal data errors/loss of signal on receiving side.
• Integrated clock recovery
• Full duplex data transmission
• Serial link and device test using built-in random bit generator/verification function
18.1 Electrical
IEEE802.3z Gigabit Ethernet standard specifies the physical and the data link layer (in reference
to the OSI model) for a serializer/deserializer device running at 1.25 Gbps on serial link. The
parallel interface is a ten-bit interface (TBI) with 10 bits of LVTTL data aligned to the rising edge
of the clock. The data is already 8B/10B encoded. Serial link is based on LVPECL signaling.
The electrical part of ANSI X3T11 Fibre Channel specification defines using TBI interface on
parallel bus and LVPECL on serial link as well.
IEEE802.3ae 10-gigabit Ethernet standard describes, among other things, the 10-Gigabit Media
Independent Interface (XGMII) as parallel bus as well as the 10-gigabit attachment unit interface
(XAUI) as serial interface. XGMII interface consists of:
• 32 data bits of 1.5-V high-speed transceiver logic (HSTL) Class 1 signals (former drafts of
the specification prescribed using 1.8-V HSTL) in each direction
• 4 LVTTL control bits indicating that the reconciliation sublayer is presenting either valid data
or control characters
• Corresponding clock signal for sampling and driving both data and control signals, in double
data rate (DDR) mode (signals valid on rising and falling edge of clock)
XAUI interface includes four ac-coupled point-to-point differential pairs. The nominal baud rate is
3.125 Gbps with maximum differential voltage of 1.6 V peak-to-peak.
18.2 Protocol
Gigabit Ethernet, 10 Gigabit Ethernet and Fibre Channel standards prescribe using specific
protocols for controlled data transmission. Detailed description of these standards is outside the
scope of this report.
It is not necessary, however, to provide fully spec-compliant data to the transceiver device (PHY)
to ensure its functionality. It is important to know that Gigabit Ethernet and Fibre Channel
transceivers expect 8B/10B encoded data on their parallel interfaces (10-gigabit devices can
provide on-chip encoding/decoding function), and that they use special pattern − K28.5, for word
alignment purposes, on the receiver parallel bus.
18.3 Applicability
Gigabit Ethernet and Fibre Channel protocols are widely used in local area network (LAN)
applications for transferring large amounts of data over copper cable or glass fiber. 10-Gigabit
Ethernet systems are primary used in metropolitan and wide area networks.
Bus speed: Fibre Channel transceivers work at a fixed speed of 106.25 MHz on the parallel
interface; Gigabit Ethernet devices accept 125-MHz parallel data; XGMII works at 156.25 MHz,
respectively. However, TI’s Gigabit Ethernet transceivers offer extended parallel clock range.
The limiting factor on minimum and maximum clock speed is the bandwidth of the fully
integrated PLL.
Bus skew: The output data stream of a Gigabit Ethernet/Fibre Channel transceiver device is
device specific and independent of the input bus skew. This might help to overcome bus skew
problems.
Bus Topology: All Gigabit Ethernet and Fibre Channel transceivers are used for point-to-point
connections.
18.4 Features
• Data throughput up to 2.5 Gbps over one differential pair
• Predefined patterns for controlled data transmission
• External pins to signal data errors/loss of signal on receiving side
• Integrated clock recovery
• Capability to drive data at 3.125 Gbps over lines up to 80 cm of FR4 backplane
• Serial link and device test using built-in random bit generator/verification function
19 SONET/SDH Transceivers
Synchronous optical network (SONET) and synchronous digital hierarchy (SDH), are similar
standards defining the data speeds, data formats and electrical/optical performance of the link
for optical networks. SONET is a US standard while SDH is the European counterpart.
19.1 Electrical
There are several electrical interfaces defined by ANSI and OIF groups for use in SONET/SDH
systems. One of them is the Serdes-Framer Interface SFI-4 described in the OIF-SFI4-01.0
specification. This interface consists of the following, and is used by OC-192 devices:
• 16 LVDS data bits in each direction clocked in source-synchronous manner on 622.08 MHz
(optional 311.04 MHz DDR) LVPECL clock
• Framer transmit clock signal sourced by serdes device
• Receive loss of synchronization error signal
19.2 Protocol
SONET/SDH standards prescribe using specific protocols for controlled data transmission.
Detailed description of these standards is out of scope of this report.
It is not necessary, however, to provide fully spec-compliant data to the transceiver device (PHY)
to ensure its functionality. It is important to know, that SONET/SDH transceivers expect
scrambled data (scrambling polynomial: 1+x6+x7) on their parallel interfaces.
19.3 Applicability
SONET/SDH technology dominates in wide area networks (WAN). SONET/SDH links are
point-to-point optical connections.
Bus speed: SONET/SDH parallel buses work at fixed speeds, such as: 38.88 MHz
(OC-3/STM-1), 155.52 MHz (OC-12/STM-4), 311.04 MHz (OC-24/STM-8), 622.08 MHz (OC-48
and 192/STM-16 and 64).
19.4 Features
• Data throughput up to 9.953 Gbps (OC-192/STM-64)
• Worldwide synchronous data transmission enabling adding/dropping of data within one
single multiplexing process
• Forward error correction applicable (PHY devices often support FEC speed which differ from
the original frequencies)
20 PCI/CompactPCI
The primary goal of the PCI developers was a low cost, flexible, high performance industry
standard local bus architecture. To enable portable systems with PCI bus, low power dissipation
also was a requirement. Personal computers have been the first area where the PCI became the
standard for local buses. Meanwhile the PCI bus starts to become a major player in the
industrial area. For industrial applications the form factor changed to CompactPCI and hot-swap
has been included. The PCI special interest group (PCI SIG: http://www.pcisig.com) maintains
the specifications of the PCI bus and the PCI industrial computer manufacturers group (PICMG:
http://www.picmg.org/) takes care of the CompactPCI specification.
20.1 Electrical
• 32-bit address space
• 32- or 64-bit data path
• 33-MHz and 66-MHz bus clock speed
• Maximum data rate: 1 Gbps for 33-MHz/32-bit bus; 4 Gbps for 66-MHz/64-bit bus
• Two major form factors:
− Personal Computer: 62-pin connector for 32-bit (Figure 18) and 82-pin connector for
64-bit systems.
− CompactPCI: Eurocard industry standard, both 3U (100 mm by 160 mm) and 6U
(233.35 mm by 160 mm) board sizes (Figure 19)
• CMOS drivers; TTL voltage levels
• 5 V, 3.3 V interoperable
• Reflected wave switching, thus the bus is short. To increase the number of connected
devices PCI-to-PCI bridges are necessary. Figure 20 shows a PCI system with 8 PCI buses.
• Direct drive—no external buffers
PCI Connector
CPU
L2 Cache (SRAM)
PCI to PCI Bridge −1 Load PCI Bus 7
PCI −ISA ISA Slot VGA PCI to PCI Bridge −1 Load PCI Bus 2
Bridge Controller
1 Load ISA Slot 1 Load
PCI to PCI Bridge −1 Load PCI Bus 1
ISA Slot
20.2 Protocol
20.2.1 Initialization
PCI devices are not configured after the supply power is switched on. None of the devices have
a memory or address range assigned, and no interrupts are available. To operate a PCI bus, all
connected devices (master and slaves) have to be configured. Through the configuration space
the controller reads the requirements of a device and is able to assign memory-, I/O-space and
Interrupts. This enables plug-and-play systems.
In large systems with multiple PCI buses, the configuration software needs an enumerator to
number all buses in the system and to program the PCI-to-PCI bridges accordingly.
20.2.2 Operation
During operation each device on the bus may become the master and can transfer data to any
other bus member. The bus design requires burst transfers to work efficiently. Still single byte
transfers are still possible.
20.3 Applicability
The PCI bus is a computer bus system. The first and biggest success was the personal
computer. Nearly every system in the world is equipped with a PCI bus system. High data rates,
low cost, multimedia support, and scalability have been key features of the PCI bus that led to
the success.
Notebook computers make advantage of the low-power consumption. The reflected wave
switching design does not require any termination or additional bus-drivers. The disadvantage is
a short bus length that can be increased using PCI-to-PCI bridges. Notebook docking stations
often use PCI-to-PCI bridges to generate a new PCI bus in the docking unit.
In the industrial area CompactPCI gains more and more market share. CompactPCI comes in a
Eurocard industry standard rack and supports hot-swap. According to the requirements of the
industrial environment, high quality connectors with hot-swap support are used for CompactPCI.
Besides hot-swap capabilities, the electrical specifications and protocols are identical to PCI.
Especially in Telecom and server applications 66 MHz and 64-bit systems are used
20.4 Features
• Processor independent
• Multimaster; peer-to-peer
• Supports memory, I/O, and configuration space
• Data bursting as normal operation mode—both read and write—variable burst length
• Low latency guarantees for real-time devices
• Initialization hooks for auto-configuration
• Arbitration: central,k access oriented, and hidden
• 64-bit extension transparently interoperable with 32 bit
21.1 Electrical
The IEEE1284 specification defines the physical set up of the 1284 interface including wiring
diagram, minimum drive capabilities and termination considerations.
21.2 Protocol
The protocol is defined in the IEEE1284 standard.
21.3 Applicability
The ’1284-compatible devices are widely used from computer and peripheral manufacturers,
because the ’1284 standard can communicate more than 50 times faster than conventional
parallel port interfaces.
CABLE P
1 E
8 Data 8 Data R
H 1
O 2 L6 I
P
S 8 5 Status V1 H
C2
4 Control E
4 R
T 8 A
4 L
Status
Peripheral Logic
Host Logic
21.4 Features
• Designed for the IEEE Std 1284-I (Level 1 Type) and IEEE Std 1284-II (Level 2 Type)
electrical specifications
• Adds bidirectional capabilities to the existing Centronics parallel interface
• Supports 5 modes of data transfer (Centronics; Nibble; Byte; EPP; ECP)
• Advanced operating mode can reach speeds of 16 to 32 Mbps
• New electrical interface, cabling and connector for improved performance and reliability
while retaining backward compatibility
• 50 to 100 times faster than the original parallel port (Centronics)
VCC VCC
d = 2 cm
R1 R1
R2 R2
R1+R2 R1+R2
R1||R2 LO LO R1||R2
R2 R2
VCC VCC
CO CL CO CL CO CL CO CL
Standard logic devices can be used for either point-to-point connections or to realize backplane
buses, which consist of many drivers and receivers along the bus, as shown in Figure 22. Both
solutions using standard logic devices are usually set up as parallel buses; up to 36 bits can be
switched by one logic device. The operational frequency can be chosen within a range from a
few MHz up to the clock frequency of about 50 MHz, such that the data-throughput per device is
in the range of 1 to 2 Gbps.
22.1 Electrical
The electrical specification of LVTTL levels meets the TTL specification. The combination of 5-V
TTL with LVTTL is possible without any additional effort. 5-V CMOS levels are not compatible
with LVTTL levels: 5-V tolerance is mandatory for unidirectional; and the use of 5-V level shifters
is required for bidirectional data transfer when combining a 3.3-V system with a 5-V supplied
system part. Table 23 shows the key parameters for 5-V and 3.3-V advanced system logic
families. Another important feature is the live insertion capability of a logic family, which enables
the user to insert and remove modules during operation. The important parameters are IOFF,
Powerup/down 3-state, and precharge functionality as shown in Table 23. To get more
information about this topic, refer to the application report Live Insertion (literature number
SCZAE07).
A maximum bus length is not specified for backplanes. However, in practice, the bus length of
parallel backplanes does not exceed about 50 cm.
22.2 Protocol
Not strictly specified for standard logic families. VME is applicable. ABTE supports VME64.
22.3 Applicability
The backplanes are not limited to any special domain. They are used in telecom, computer, and
industry application, wherever several system parts are connected using a backplane or a
memory bus.
22.4 Features
• 8-, 16-, and 32-bit devices enable parallel operation on the backplane/memory bus
• Boundary scan devices (JTAG - IEEE 1149.1) available in LVT and ABT enable easy
testability during design and production
• Bus hold feature eliminates external pullup resistor
• Series damping resistors enable improved signal integrity in point-to-point buses
• ABT, LVT, ALVT and LVCZ incorporate power up 3-state outputs, enabling support of hot
insertion/removal
• ABTE supports precharge feature, enabling support of hot insertion/removal
• LVC, ALVC are specified down to 1.8 V, further reducing power consumption
Table 24. Top Feature List of Advanced System Logic by Logic Family
AHC AC ABT ABTE (A)LVC LVT ALVT AVC AUCΦ
Gates n n n N/A n N/A N/A N/A n
Flip-flops n n n N/A n n n N/A n
Drivers n n n n n n n n n
Transceivers n n n n n n n n n
UBT† N/A N/A n N/A n n n n n
Bus hold‡ N/A N/A n n n n n n n
26 Ω Series resistors§ n n n n n n n DOC n
SCOPE¶ N/A N/A n N/A N/A n N/A N/A N/A
† The universal bus transceiver (UBTTM) combines D-type latches and D-type flip-flops for operation in transparent, latched or clocked mode.
‡ Bus hold on data inputs eliminates the need for external pullup resistors.
§ 26-Ω series resistors are included in the output stages, in order to match bus impedance avoiding external resistors
¶ SCOPE products are compatible with the IEEE Standard 1149.1–1990 (JTAG) test access port and boundary scan architecture
A bus system with reduced voltage swing solves a lot of problems. The BTL bus realizes a bus
in open collector mode, as shown Figure 24. In this case, the falling edge is actively generated
from the driver. Only a low impedance driver can switch the bus with the incident wave.
VCC
2.1 V
R = ZO
VREF = 1.55 V
BUS
Line Driver Receiver
Figure 24. Principle Setup of an Open Collector Bus System Using BTL Devices
23.1 Electrical
The physical layer of the Futurebus is called backplane transceiver logic (BTL) and works with a
voltage swing of 1.1 V only, using an open collector bus system. The saturation voltage of the
pulldown transistor and the forward voltage of the serially connected diode generates the output
low level voltage of 1 V. The high level of 2.1 V comes from the termination resistor connected to
the termination voltage of 2.1 V. The value of the termination resistor is equal to the impedance
of the bus line and therefore the bus line is terminated correctly. For safe detection of the logic
levels, the inputs are designed with differential amplifiers and a threshold at 1.55 V, exactly in
the middle of the voltage swing.
To reduce ICC current spikes, the fall time is defined to be 2 ns or slower. The rise time is not
generated by active electronics, but by the pullup resistor.
23.2 Protocol
The FutureBus + logical layer specification, according to IEEE896.2, describes the node
management, live insertion, and profiles. However, the physical layer may also be used
stand-alone without the logical layer.
23.3 Applicability
The target area for BTL devices is the telecom sector, where live insertion capability is
mandatory.
23.4 Features
• Reduced voltage swing: VL = 1 V; VH = 2.1 V generates low switching noise 10 Ω
(20 Ω || 20 Ω) line impedance
• Correct line termination by a pullup resistor at the line end avoids line reflections
• Decoupling diode reduces output capacitance to < 5 pF, increases line impedance
• Maximum output edge rate 2 ns; trapezoidal waveform reduces system noise
• Supports live insertion/withdrawal
R=Z
Line Driver
VREF = 0.8 V
Figure 25. Principle Setup of an Open Collector Bus System Using GTL Devices
Because there is no diode in the open collector/drain outputs (compared to the BTL- solution)
the low level is 0.4 V. With a chosen high level of 1.2 V, the voltage swing is reduced to 0.8 V
only. Again, the threshold is in the middle of the voltage swing at 0.8 V.
With a drive capability of GTL outputs up to about 40 mA, the GTL devices are able to drive a
termination resistor of 0.8 V/40 mA = 20 Ω. If the bus line is terminated correctly, the lowest
impedance that can be driven by a GTL driver in the middle of a bus is 40 Ω (effectively the
driver sees: 40 Ω || 40 Ω = 20 Ω). As a result of the 0.8 V swing and the 40 mA IOL, the
maximum power dissipation of one output is 16 mW. Therefore, it is possible to integrate these
low power drivers into ASICs.
24.2 Electrical
The gunning transceiver logic (GTL) devices support two different logic level specifications: GTL
(according EIA/JEDEC Standard EIA/JESD8-3) and the GTL + levels. The bus system, similar to
the BTL bus, is realized as an open collector bus. Because no diode is included in the open
collector/drain output stage of the GTL devices, the output low level can be reduced to 0.4 V
(GTL+: 0.55). With a chosen high level of 1.2 V for GTL (GTL+: 1.5 V), the voltage swing is
reduced to only 0.8 V (0.95 V). The threshold is in the middle of the voltage swing at 0.8 V
(GTL+: 1 V).
GTL+ is becoming more and more a standard in the industry due to the enlarged noise margin
of GTL+ levels. For example GTL+ levels are being used on the Intel Pentium Pro (P6)
processor to address this noise margin concern. Using GTL+ levels instead of GTL, the margin
is increased about 16 %.
24.3 Protocol
Not specified.
24.4 Applicability
GTL was originally designed for a small bus on a board, for example, between a processor and
its memory modules. Because the target application for GTL is not a backplane bus, but a bus
on a board, no requirements for live insertion/withdrawal have been included in the specification.
With reduced output levels and state-of-the-art designs, the consequences are reduction of
power consumption, higher speeds, and improved signal integrity compared to the BTL-bus,
such that GTL+ backplane optimized-drivers are a premium solution for heavily loaded bus
systems. Live insertion capabilities and an increased drive for low impedance backplanes are
met with the GTL1655 device.
24.5 Features
• Differential amplifier specifies stable threshold voltage of the receiver
• Low voltage swing generates low switching noise
− GTL: VL = 0.4 V; VH = 1.2 V
− GTL+: VL = 0.55 V; VH = 1.5 V
• High drive capable option available, enabling incident wave switching as low as 10 Ω
(20 Ω || 20 Ω) line impedance
− GTL/GTL+: Low drive capability, Iolmax = 40/50 mA
− GTL1655: High drive capability, Iolmax = 100 mA
• Correct line termination using a pullup resistor at the line end avoids line reflections
• Edge rate control output circuit of GTL1655 enables variable output slew rate depending on
load condition for maximum data throughput.
45
RT(Min)
40 ZO = 65 W
(Medium Drive)
Termination Resistor − Ω
35
30
ZO = 50 W
25
RT(Min)
20 (High Drive)
15
2.5 3.5 4.5 5.5 6.5
Slot Spacing − cm
Figure 26. RT vs Slot Spacing With GTLP Medium and High Drive Devices
To enhance the data throughput of a GTLP backplane a source synchronous clock layout is
advised. Due to the fact that the clock signal is sent from the same source, the flight time over
the bus can be eliminated in the timing budget. This roughly doubles the data throughput of the
backplane. TI offers special devices the GTLPH1627 and GTLPH16927 which have one bit
integrated for the source synchronous clock. It also features a flexible setup time adjustment
(FSTA), which offers the designer an easy way of implementation of a source synchronous bus.
25.1 Electrical
Optimized for the GTLP signal level specifications, also operates at GTL (according JEDEC
Standard JESD8-3) or GTL+ signal levels. The bus system, identical to the GTL bus, is realized
as an open drain bus. The GTLP voltage swing is from 1.5 V to 0.55 V with ± 50 mV around the
VREF threshold of 1 V.
25.2 Protocol
Not specified.
25.3 Applicability
GTLP is used where the major concerns are higher data throughput, live insertion capability or
lower power consumption in parallel backplane architectures. GTLP offers up to four times the
performance of TTL devices in backplane upgrade applications.
25.4 Features
• 3.3-V operation with 5-V-tolerant LVTTL inputs/outputs which allow the devices to act as 5-V
TTL to GTLP as well as 3.3-V LVTTL to GTLP translators
• Significantly improved output edge control (OEC) circuitry on the rising and falling edge of
the GTLP outputs reduces line reflections, electromagnetic interference (EMI) and improves
overall signal integrity allowing clock frequencies in excess of 80 MHz, with Source Synchro-
nous Clock Layout even in excess of 120 MHz.
• Fully supports live insertion with Ioff, PU3S and BIAS VCC circuitry
• Edge rate control (ERC) circuitry on high drive devices allows fast or slow edge rates.
• Output edge control (OEC) circuitry on the rising and falling edge of the GTLP outputs
• CMOS construction for 1/3 the static power consumption of BiCMOS logic devices
• A-port (LVTTL side) balanced drive of ±24 mA with optional bus-hold circuitry
26 VMEbus
The VMEbus is an asynchronous bus, which operates in a master/slave architecture. In 1987,
the VMEbus specification was originally introduced as IEEE1014, but was altered several times
to adjust to newer technologies and to increase the data throughput. New standards became the
VME64, VME64x and finally the VME320 specification. VMEbus standards are governed by the
VITA committee. TI was asked by the VITA committee to jointly develop a bus driver complying
with the VME320 specification, but would be still backward compatible.
26.1 Electrical
The VMEbus uses push-pull totem pole outputs. For the VME320, LVTTL levels are used with a
±48-mA output drive. For the implementation of the VME320 bus a special star interconnection
method is used to reach higher speeds.
26.2 Protocol
26.3 Applicability
The VMEbus is often used in a wide variety of applications. The main areas are still industrial,
such as factory automation, and robotics; or in aerospace, e.g., fly-by-wire systems.
26.4 Features
• Live insertion capability
• Up to 21 slots in backplane
• Asynchronous data transfer
• Error detection
• Up to 64 address and data bits
RT 25 Ω
RS
VOUT
25 Ω VIN VREF
VSS
Class I specifies an acceptable value of 50 Ω for the termination resistor RT, and Class II
specifies an acceptable value of 25 Ω. Figure 27 shows the typical dc environment for the output
buffer (Class II); in this case an additional series resistor RS is specified at 25 Ω. In order to meet
the 400 mV minimum requirement for VIN, a minimum of 8 mA must be driven into RT, if RT
equals 50 Ω (Class I); or 16 mA if RT equals 25 Ω (Class II). The standard stipulates that for
each value of RT, a capacitive load equal to 10 pF or 30 pF can be used. The SSTL16837A
supports both SSTL and LVTTL switching levels. Although the data sheet provides specifications
where SSTL levels are used for the input and output levels, the device can operate under any
combination of SSTL/LVTTL levels for the inputs and the outputs. When SSTL levels are applied
to the device, it functions approximately 2 ns faster than when using LVTTL levels.
27.1 Electrical
The stub series terminated logic (SSTL) interface standard is intended for high-speed memory
interface applications and specifies switching characteristics such that operating frequencies up
to 200 MHz are attainable. The input high and low voltage levels (VIH and VIL) are VREF +
200 mV and VREF − 200 mV, resulting in a worst case noise margin of 25%. This seems to be a
relatively small noise margin, but because it is a terminated bus, the actual noise source would
have to be a fairly high current to produce 400 mV across the relatively low-impedance
termination.
All totem-pole outputs of the SSTL compatible devices have a dedicated VDDQ supply that (as
stated in the SSTL_3 and SSTL_2 JEDEC standards) can be lower than or equal to VDD, but
never greater than VDD. This feature allows for the internal circuitry supply voltage to be raised
to 3.6 V for maximum speed performance, while lowering VDDQ to prevent the device from
dissipating large amounts of power in the output stage.
Irrespective of the input and output switching levels however, the characteristic high-level and
low-level output drive current of 20 mA is maintained.
27.2 Protocol
Not specified.
27.3 Applicability
Figure 28 shows a complete DDR SDRAM (SDRAM II) memory interfacing solution offered from
Texas Instruments. Using a specially designed register, low-voltage bus switches, and a
differential clock, DDR SDRAM modules can achieve double the memory data rate by allowing
the SDRAMs to operate at twice their system frequency using the rising and the falling edge of
the system clock.
The 14-bit registered buffer SSTL16857 is designed for 2.3-V to 3.6-V VCC operation and
differential data input and output levels. It buffers DDR SDRAM address and control signals. The
CDC857 differential clock completes the TI solution for 184-pin DDR SDRAM modules.
Optionally, the SSTL_2 optimized CBTLV3857 provides bus isolation in the DDR SDRAM DIMM
application and reduces the capacitive loading on the data lines.
Front View
Back View
Figure 28. DDR SDRAM Memory Interfacing Solution Using the SN74SSTL16857
27.4 Features
• Maximum frequency of 200 MHz enables fast DDR SDRAM memory buses
• Supports both SSTL and LVTLL switching levels, which enables any combination of SSTL
and LVTTL levels for inputs and outputs
• SSTL and LVTTL levels for inputs and outputs
• Outputs have dedicated VDDQ, which can be lower or equal to VDD. This enables the internal
circuitry supply voltage to be raised to 3.6 V for maximum speed, while lowering VDDQ to
prevent the device from large power dissipation in the output stage.
Table 30. Top Device List for Stub Series Termination Logic
DEVICE SPECIFICATION DESCRIPTION PACKAGE STATUS
SN74SSTL16837A SSTL_3 20-Bit SSTL_3 interface universal bus driver with 3-state outputs 64-pin TSSOP Available
SN74SSTL16847 SSTL_3 20-Bit SSTL_3 interface buffer with 3-state outputs 64-pin TSSOP Available
28.1 Electrical
This type of circuits has no industry standard except clock circuits for DIMM (SRAM, DDR and
Rambus).
28.1.4 Synthesizer
The PLL-based clock circuitry receives low frequency signals from the crystal or other source
and delivers different frequency output signals according to system needs.
28.2 Protocol
No protocol is specified for CDC devices. This is dependent on applications. However, the
following definition of some parameters are related to clock drivers.
Definitions of some basic parameters related to clocks:
Propagation Delay: Propagation delay is the time between the specified reference points on the
input and output voltage waveforms with the output changing from one defined level
(high-to-low) to the other (low-to-high) defined level.
Output Skew: Output skew is the difference between the propagation delays of any two outputs
of the same device at identical transitions.
Pulse skew: Pulse skew is the magnitude of the time difference between the high-to-low and
the low-to-high propagation delays when a single switching input causes one or more outputs to
switch.
VDD
CLKIN 50% VDD
0V
tPLH = Low to High Propagation delay
tPLH tPHL
tPHL = High to Low Propagation delay
Y0 50% VDD tsk(o) = Output Skew
tsk(p) = Pulse Skew = |tPLH − tPHL |
tsk(o)
Y1 50% VDD
Static Phase Offset: Static phase offset is the time difference between the averaged input
reference clock and the averaged feedback input signal when the PLL is in locked mode.
CK
CK
FBIN
∑ n=N
1 t(φ) n
t(φ) = (N is a Large Number of Samples)
N
Dynamic Phase Offset: Dynamic phase offset is the phase difference between input clock and
feedback input clock due to the inability of the PLL to instantaneously update the feedback clock
when the period of the input clock changes.
Jitter: Jitter is the dispersion of a time parameter of the pulse waves in a pulse train with respect
to reference time, interval, or duration.
Peak-to-Peak Jitter: Peak-to-peak jitter is defined as the upper and lower bounds of a large
number of samples of cycle-to-cycle period measurements from ideal.
Cycle-to-Cycle Jitter: Cycle-to-cycle jitter is the difference in the period of successive cycles of
a continuous clock pulse train.
PLL Loop Bandwidth: PLL loop bandwidth is that region where the PLL is able to track phase
and frequency modulation of the reference signal as long as the modulation frequencies remain
as an angular frequency band roughly between zero and the natural frequency of the PLL.
Spread Spectrum Clock (SSC) Compatible: A PLL clock driver is SSC compatible if the PLL is
able to track the modulation profile, and frequency is minimum tracking skew.
28.3 Applicability
28.3.1 Clocks Drivers for TI Serdes
All serializer-deserializers (serdes) require a low jitter clock to perform its operation correctly.
Since the input duty cycle and jitter are very critical to serdes operation, TI clocks are readily
available to meet these requirements. The peak-peak jitter requirement for the TLK1201,
TLK1501, TLK2201, TLK2501, TLK2701, TLK2711, and TLK3101 is less than 40 ps; and
CDCV304 is an excellent clock driver for these serdes, adding just 20 ps peak-peak jitter, on
average.
The RMS jitter requirement of SLK2501, and SLK3104 is less than 3 ps. The CDCVF111 and
CDC111 add very negligible jitter to its output. This differential clock driver has nine outputs and
very low output skew.
Table 31. Recommended Clock Drivers for Serial/Gigabit Transceivers
SERIAL/GIGABIT TRANSCEIVER RECOMMENDED CLOCK DRIVER
SLK2501 CDCVF111, up to 622.08 MHz
CDC111, up to 500 MHz
TLK3104SC CDCVF111, up to 622.08 MHz
CDC111, up to 500 MHz
TLK3114SA CDC111/CDCVF111
TLK1201, TLK1501, TLK2201, TNETE2201B, TLK2501, TLK2701, TLK2711, TLK3101 CDCV304, up to 140 MHz
29 Summary
This Comparing Bus Solutions application report shows that Texas Instruments provides a vast
portfolio of data transmission products covering most of the commonly used communication
standards. The offering ranges from typical backplane logic families like ABT, LVT and GTLP to
more advanced serial transmission families like LVDS and IEEE1394, while still providing
solutions for the mature TIA/EIA (formerly RS) standards such as 232 and 485.
By referring to this application report, choosing the correct standard and therefore the
appropriate family, has been made easier. Simply by selecting the type of transmission, e.g.,
serial or parallel, the distance and the data rate, a designer can very quickly find what is needed
and has an introduction to the family contained in the report, with Internet pages waiting to be
browsed for additional information.
Texas Instruments is committed to maintaining a leadership position in the field of data
transmission, and TI continues to increase the range of advanced products available to
engineers.
30 References
1. EIA/TIA–422–B, ITU–T V.11: Electrical characteristics for balanced double-current interchange
circuits operating at data signaling rates up to 10 Mbit/s
2. EIA/TIA-485, ISO/IEC 8482: Information Technology—Telecommunications and information
exchange between systems—Twisted pair multi-point interconnections
3. CAN, ISO 11898, Controller Area Network
4. ISO/IEC 9316, Information Technology, Small Computer System Interface–2
5. EIA/TIA-232-E, ITU–T V.28
6. TIA/EIA-644, Electrical Characteristics of Low Voltage Differential Signaling (LVDS) Interface
Circuits
7. TIA/EIA-899, Electrical Characteristics of Multipoint-Low Voltage Differential Signaling
(M-LVDS) Interface Circuits for Multipoint Data Interchange EIA/JESD8–A, Interface
Standard for Nominal 3 V/3.3 V Supply Digital Integrated Circuits
8. EIA/JESD8–3, Gunning Transceiver Logic (GTL), Low-Level, High-Speed Interface Standard
for Digital Integrated Circuits
9. IEEE1394–1995, High Performance Serial Bus
10. IEEE P896.1, Futurebus+ Logic Layer Specifications
11. IEEE P896.2, Futurebus+ Physical Layer and Profile Specifications
12. EIA/JESD8–8, Stub Series Terminated Logic for 3.3 V (SSTL_3)
13. EIA/JESD8–9, Stub Series Terminated Logic for 2.5 V (SSTL_2)
14. IEEE802.3z Gigabit Task Force
15. GMII Gigabit Media Independent Interface
16. USB Universal Serial Bus Specification Revision 1.1 (Compaq, Intel, Microsoft, NEC)
31 Glossary
ABT Advanced BiCMOS technology, 5-V logic family with
TTL-compatible input and output specification
ABTE Enhanced transceiver logic. 5-V logic with reduced noise margins
to achieve higher speeds on the backplane, TTL-compatible
Arbitration The process by which nodes compete for ownership of the bus.
The cable environment uses a hierarchical point-to-point algorithm,
while the backplane environment uses the bit-serial process of
transmitting an arbitration sequence. At the completion of an
arbitration contest, only one node is able to transmit a data packet.
DDR SDRAM Double data rate synchronous dynamic random access memory
Downstream The direction of data flow from the host or away from the host. A
downstream port is the port on a hub electrically farthest from the
host that generates downstream data traffic from the hub.
Downstream ports receive upstream data traffic.
GTL Gunning transceiver logic. GTL+ and GTLP are derivatives of GTL
that operate at enhanced noise margin signal levels (VTT = 1.5 V,
VREF = 1 V and VOL = 0.55 V). GTLP is normally associated with
optimized edge rate devices that allow high frequency operation in
heavily loaded backplane applications.
Live Insertion/Removal The ability to attach and remove devices while the system is in
operation
Link layer The layer, in a stack of three protocol layers defined for the serial
bus, which provides the service to the transaction layer of one–way
data transfer with confirmation of reception. The link layer also
provides addressing, data checking, and data framing. The link
layer also provides an isochronous data transfer service directly to
the application.
Node An addressable device attached to the serial bus with at least the
minimum set of control registers. Changing the control registers on
one node does not affect the state of control registers on another
node.
Packet (1394) A serial stream of clocked data bits. A packet is normally the PDU
for the link layer, although the cable physical layer can also
generate and receive special short packets for management
purposes.
Physical layer The layer, in a stack of three protocol layers defined for the serial
bus, which translates the logical symbols used by the link layer into
electrical signals on the different serial bus media. The physical
layer ensures that only one node at a time is sending data and
defines the mechanical interface for the serial bus. There is a
different physical layer for the backplane and for the cable
environment.
RS Recommended standard
Termination resistor The termination resistor is used at the end of a line in order to avoid
reflections of the transmitted signal. If the termination resistor is
chosen equally to the line impedance the line is optimally
terminated.
Transfer One of the four USB transfer types. Isochronous transfers are used
when working with isochronous data. Isochronous transfers provide
periodic, continuous communication between host and device.
Upstream The direction of data flow towards the host. An upstream port is the
port on a device electrically closest to the host that generates
upstream data traffic from the hub. Upstream ports receive
downstream data traffic.
32 TI Contact Numbers
INTERNET Asia
TI Semiconductor Home Page Phone
www.ti.com/sc International +886-2-23786800
TI Distributors Domestic
www.ti.com/sc/docs/distmenu.htm Australia 1-800-881-011
PRODUCT INFORMATION CENTERS TI Number -800-800-1450
Americas China 10810
Phone +1(972) 644-5580 TI Number -800-800-1450
Fax +1(972) 480-7800 Hong Kong 800-96-1111
Email [email protected] TI Number -800-800-1450
Europe, Middle East, and Africa India 000-117
Phone TI Number -800-800-1450
Deutsch +49-(0) 8161 80 3311 Indonesia 001-801-10
English +44-(0) 1604 66 3399 TI Number -800-800-1450
Español +34-(0) 90 23 54 0 28 Korea 080-551-2804
Francais +33-(0) 1-30 70 11 64 Malaysia 1-800-800-011
Italiano +33-(0) 1-30 70 11 67 TI Number -800-800-1450
Nordic Countries New Zealand 000-911
Fax +44-(0) 1604 66 33 34 TI Number -800-800-1450
Email [email protected] Philippines 105-11
Japan TI Number -800-800-1450
Phone Singapore 800-0111-111
International +81-3-3457-0972 TI Number -800-800-1450
Domestic 0120-81-0026 Taiwan 080-006800
Fax Thailand 0019-991-1111
International +81-3-3457-1259 TI Number -800-800-1450
Domestic 0120-81-0026 Fax 886-2-2378-6808
Email [email protected] Email [email protected]
Texas Instruments Incorporated and its subsidiaries (TI) reserve the right to make corrections, modifications,
enhancements, improvements, and other changes to its products and services at any time and to discontinue
any product or service without notice. Customers should obtain the latest relevant information before placing
orders and should verify that such information is current and complete. All products are sold subject to TI’s terms
and conditions of sale supplied at the time of order acknowledgment.
TI warrants performance of its hardware products to the specifications applicable at the time of sale in
accordance with TI’s standard warranty. Testing and other quality control techniques are used to the extent TI
deems necessary to support this warranty. Except where mandated by government requirements, testing of all
parameters of each product is not necessarily performed.
TI assumes no liability for applications assistance or customer product design. Customers are responsible for
their products and applications using TI components. To minimize the risks associated with customer products
and applications, customers should provide adequate design and operating safeguards.
TI does not warrant or represent that any license, either express or implied, is granted under any TI patent right,
copyright, mask work right, or other TI intellectual property right relating to any combination, machine, or process
in which TI products or services are used. Information published by TI regarding third-party products or services
does not constitute a license from TI to use such products or services or a warranty or endorsement thereof.
Use of such information may require a license from a third party under the patents or other intellectual property
of the third party, or a license from TI under the patents or other intellectual property of TI.
Reproduction of information in TI data books or data sheets is permissible only if reproduction is without
alteration and is accompanied by all associated warranties, conditions, limitations, and notices. Reproduction
of this information with alteration is an unfair and deceptive business practice. TI is not responsible or liable for
such altered documentation.
Resale of TI products or services with statements different from or beyond the parameters stated by TI for that
product or service voids all express and any implied warranties for the associated TI product or service and
is an unfair and deceptive business practice. TI is not responsible or liable for any such statements.
Following are URLs where you can obtain information on other Texas Instruments products and application
solutions:
Products Applications
Amplifiers amplifier.ti.com Audio www.ti.com/audio
Data Converters dataconverter.ti.com Automotive www.ti.com/automotive
DSP dsp.ti.com Broadband www.ti.com/broadband
Interface interface.ti.com Digital Control www.ti.com/digitalcontrol
Logic logic.ti.com Military www.ti.com/military
Power Mgmt power.ti.com Optical Networking www.ti.com/opticalnetwork
Microcontrollers microcontroller.ti.com Security www.ti.com/security
Telephony www.ti.com/telephony
Video & Imaging www.ti.com/video
Wireless www.ti.com/wireless