Published July 1999

THE ART AND SCIENCE OF RS-485

by Bob Perrin

Start • Arm Yourself • RS-485 101 • Getting Grounded • Shielding • Topology • Termination • Idle-state Biasing • Transients • Review Time • Sources

RS-485 101

Before delving into the nitty gritty, let's first examine some general characteristics of a network built with drivers and receivers compliant with TIA/EIA-485-A.

RS-485 is a half-duplex multidrop network, which means that multiple transmitters and receivers may reside on the line. Only one transmitter may be active at any given time. TIA/EIA-485-A says nothing about the communications protocol to be used on the network. The software engineer has the liberty to implement whatever type of network protocol is deemed applicable for the current project.

RS-485 transmission lines are differential in nature. There are two wires—A and B. The driver generates complementary voltages on A and B. Figure 1 shows how EIA-485-A defines V_OA, V_OB, and V_O. When V_OA is low, V_OB is high; when V_OA is high, V_OB is low. Most physical parts also have the ability to tristate both A and B.

Figure 1—The relationship between VOA, VOB, and VO is carefully spelled out in TIA/EIA-485-A.

Signals A and B are complementary, but this doesn’t imply that one signal is a current return for the other. RS-485 is not a current loop. The drivers and receivers must share a common ground. This is why "two-wire network" is a misnomer when applied to RS-485. More on this later.

Receivers are designed to respond to the difference between A and B. V_O is the differential voltage. Receivers must be sensitive to a 200-mV difference between V_OA and V_OB. Anything less than 200 mV is indeterminate.

RS-485 can support networks up to 5000’ long and bit rates of up to 10 Mbps. Data rate must be traded off against cable length [1]. Figure 2 shows a graph fairly typical of the bit rates and line lengths you can expect. Performance will vary depending on cable type, termination, drivers and receivers used, EMI coupled into the system, and the physical geometry of the network.

Figure 2—Trading data rate for cable length is the unfortunate consequence of finite propagation delay on the transmission line.

TIA/EIA-485-A defines a unit load (UL) and declares that an RS-485 driver must be able to drive 32 ULs. The standard’s authors anticipated that device manufacturers would implement receivers and transceivers (with the driver in the high-Z state) to present a single UL load to the line.

The natural conclusion and often-repeated myth is that an RS-485 network can only support 32 nodes. This is not true. Device manufacturers now sell ¹/₄ UL transceivers (DS1487) and even ¹/₈ UL parts (MAX1482).

Assuming each node presents ¹/₈ UL to the transmission line, an RS-485–compliant network may sport as many as 256 nodes (32 UL × 8 UL/node = 256 nodes).

By using repeaters, multiple networks can be chained together to accommodate virtually an unlimited number of nodes. The propagation delays will become significant for large networks with multiple repeaters and long transmission lines, and the data rate may become unacceptably low.

Some drivers are designed to have slow edge times. These are often referred to as slew-rate limited drivers. Slow edges have reduced high-frequency components associated with them. Longer edge times permit the use of longer cables and reduce the amount of EMI emitted by the network.

Now that we have a general understanding of what an RS-485 network is, let’s examine some common pitfalls.

PREVIOUS • NEXT

Circuit Cellar provides up-to-date information for engineers. Visit www.circuitcellar.com for more information and additional articles.
For subscription information, call (860) 875-2199, subscribe@circuitcellar.com or subscribe online. ©Circuit Cellar, the Magazine for Computer Applications. Posted with permission.