There is an alternative to the common +10-V analog signal and proprietary drive-to-controller communication networks — open digital networks (buses). But there’s also much confusion about what is an “open” digital bus; what features, capabilities, and benefits such a bus provides; and which one is best.

Plug-n-play or plug-n-pray?

Digital buses have become popular recently because most claim they are open. Users should be aware, however, that there is no generally agreed upon definition of what is meant by open. Manufacturers tend to use the word for their own purposes. At one end of the definition spectrum, open can mean that the communication network or bus is a recognized standard. At the other end, it can mean that the manufacturer will supply the customer with specifications on the bus or network, but it will be the customer who must make the interconnection between products or systems.

Customers want open to mean that systems from manufacturers will easily interconnect. They perceive that with an open motion system, they:

• Save installation and setup time because products, regardless of vendor, are compatible.
• Gain more independence in selecting drives.
• Need less support from manufacturers because of the compatibility.
• Obtain best-in-class product solutions.
• Can standardize their production facilities, which can reduce total life cost of a system.

Before engineers choose a network or bus, they should make sure their chosen supplier agrees on what both parties mean by open.

“Technically, though, there’s no real advantage of open networks over proprietary networks,” says Larry Voss, applications engineer, Atlas Copco Controls Inc. “Both offer high speed and reduce wiring costs. But, people want to have this feeling that with an open network they have options.”

“In addition,” says Bryan McGovern, marketing manager, Emerson Electric Co., “as motion systems get more complicated, users would like to standardize their equipment to possibly reduce the total costs of these systems. So far, though, there’s no one best standard out there for all applications. And I don’t foresee one coming for awhile.”

“Customers think they want open motion systems because it will let them ‘plug-n-play,’” says Scott Hibbard, vice president, engineering, Indramat Div., the Rexroth Corp. “But open doesn’t necessarily mean plug-n-play. Most vendors and customers agree that open means a product that is well defined using standards that no one company owns. If anything, the standard is owned by a neutral party. Updates to the standard are carefully controlled by a standards committee. To some, plug-n-play means something has been designed well enough that all of the parameters are known ahead of time so that when you fasten the devices together, you need little or no configuration (and its headaches) to get the two devices to run together. In that sense, plug-n-play can be applied to a closed (proprietary) system.”

Plug-n-play, to others, requires a specific level of openness of all products in an industry. Right now in the motion-systems industry, openness is at a beginning level, more plug-n-pray than plug-n-play. And as Mr. McGovern alluded to, it will be a while before this industry settles on one or more specific standards.

For now, whether they choose an open or proprietary communication system, users do not have all the independence they may want in drive selection and they must still work closely with the manufacturer to get a system working properly.

And even with the availability of technical specifications (which constitutes openness for some) of these buses, users cannot choose any drive, or any amplifier, or any control to operate with a particular bus in a system.

Some of the open networks allow drive vendors to add in features and functions that they can use to differentiate (customize) their products from the competition. For example, one network offers standard drive parameters and commands, such as command and actual speed, position, and so on. “If manufacturers use only the standard instructions,” says Mr. Voss, “then the drives compatible with that network are interchangeable. But a manufacturer can put in one or more non-standard instructions, which are allowed in this network’s free message area. If customers use these vendor specific drive instructions in their applications, then they have “closed” this open bus and can no longer simply swap one network-compliant drive with another.” Thus, even among one network’s compliant products, the user must be careful.

This then, raises the question, “Which digital network is better?”

“It all depends on your application,” says Mr. Voss. “It really comes down to customer preference. And some customers choose a network because they may have heard of it more than another.”

Agrees Bryan Lawson, drives specialist, Baldor Motors and Drives, “It’s totally application dependent. We currently have projects that involve two different networks because each has different application requirements.”

Before narrowing your choice to a specific digital network, it’s important to consider the following criteria:

Data throughput rate. Not all digital networks offer the same throughput rate. Data throughput rate is the time it takes for the control to read data and write commands to the drive. Tightly synchronized, coordinated motion can require data to move at a rate of several bits per microsecond. By contrast, conveyor control and integrated machining applications operate properly when data moves at 1Mbs or less. Pump and fan applications do not need data to move even that fast.

The required data rate depends on the type of transmitted information. If torque commands are sent by the controller, which performs position and velocity loop closure, then the speed of the feedback data must be faster then the speed of the position commands sent by the controller.

Throughput is not the same as speed. Networks, whether for motion control or computer transmission, all have this definition problem. Most manufacturers will claim their network can send data at many millions of bits per second — if you don’t count how many nodes are on the network, the length of the cable between nodes, the length of the message, whether the data are sent on an asneeded basis or queried by a master control, and so on.

However, the data throughput rate of the network should not be the main concern. Typically, a motor has the slowest response in a network. “Sometimes the network system is limited by the performance capability of the drive or the control,” says Mr. Hibbard. “That is, can the drive physically accelerate or physically hold a tolerance? Or, can the control feed data to the network fast enough? For example, in machine tool cutting, where you have thousands of points on a curve to calculate for axis motion, the problem is often whether the CNC can prepare that data fast enough to send to the network.”

Number of axes. An application can have too few axes to take full advantage of a digital bus. “If you have an application with two axes, for example,” says Ann Hamil, product line manager, multiaxis controls, Automation Intelligence Inc., “and they’re in the same control cabinet, you’re not gaining as much by using a digital bus. You’re not losing anything, but you’re not taking advantage of the physical and cost benefits of the bus. You still gain the benefits of diagnostics and setup that the network provides.”

As for a large number of axes, motor and drive manufacturers say the average big customer does not have an application with more than 32 axes on a bus, a number both digital buses reportedly can handle. Both buses, supporters also claim, handle over 250 nodes. (Depending on the bus specification, a node can be one axes or several axes under the direction of a control, or some combination.)

Centralized or distributed control. Which control scheme will your drive system use? Centralized control, as defined by most people in the motion industry, means a system has one master controller that sends velocity or torque commands or signals to one or more drives. Sometimes, there are several cells, each consisting of several drives reporting to a control. Each cell control then reports to the one master controller.

Distributed control means that each drive has its own source of intelligence, closing position velocity and torque loops. In addition, it communicates with its own controller, which issues position or high-level commands.

Benefits of centralized control:
• Drives do not require an additional processor for position-loop control because a central control executes this function.
• The drives can execute tight synchronization and coordination among axes.
• Depending on the application, one operator can set up or maintain a centralized motion control system.

Disadvantages:
• Slow throughput on a drive bus limits the ability to distribute drives throughout a plant facility.

Benefits of distributed control:
• Frees a motion system controller to do other tasks than position, velocity, and torque loop closure. The position loop can be closed in the drive.
• Reduces wiring and installation costs.
• Handles complex motion tasks. For example, a distributed control is ideal for coordinating the movement of multiple machines and part-handling equipment in an automated assembly cell or process, where machines are interdependent.
• Offers diagnostic information on motors and drive performance.

Disadvantages:
• Handshaking and connectivity issues. Customers must ensure that other devices on a distributed system have the same or compatible data protocols and are integratable.

Lastly, whichever bus is chosen, be sure it can meet your future as well as your present motion needs.

See Associated Figure

If this article is helpful, please circle 330 on the reader service card.

Why digital?

“As customers use more digital drives and amplifiers in their motion systems, the analog bus, despite its speed, has become a restriction on system performance and information flow,” says Mr. Hibbard.

A digital network:

• Reduces the amount and the cost of wiring.
• Eliminates A/D and D/A conversion.
• Provides immunity to most electrical noise in a facility because many of these networks use fiber optic cabling.
• Enables engineers to gather diagnostic information on motor and drive performance.

Sercos and Macro use one fiber optic cable daisy-chained between the drive and the control. Signals sent through this cable include command (position, current, velocity), encoder, and sometimes I/O signals, as well as diagnostic information from motors and drives.

By installing one cable, total wiring costs are reduced. “According to one of our customers,” says Jeff Pinegar, marketing manager, Automation Intelligence Inc., a Div. of Pacific Scientific Co., “with an analog interface, it cost $9.00 every time they had to attach a wire. And there might be 20 wires between the analog drive and the control when you consider the quadrature encoder wires, the ±10-V command, fault signals, enable signals, and so on.”

Several manufacturers offer amplifiers that close their current loops digitally. The use of a digital communication bus eliminates the need to convert from analog to digital then from digital back to analog. This type of bus also eliminates the errors often found in A/D and D/A conversion, ground loops, and voltage offset and shift problems.

“A digital current loop in the amplifier reduces setup and generally improves performance,” said Dave Bedrosian, application engineering manager, Delta Tau. “Commutation on brushless motors and vector drives is now done digitally, and often in the amplifier. The position controller has historically closed its loops digitally. With an analog interface between the position controller and the amplifier, however, you have the cost of converting a command to analog, piping it across a noise-sensitive wire, then converting it back to digital for current loop closure. Rather than do this, people want a digital link that enables the amplifier and position control to talk together without conversion.”

Digital networks can also provide information that maintenance or operators can use to diagnose trouble in a system. “You can get a lot more information in and out of the drive than just setting the velocity and finding out if there’s a fault,” says Ms. Hamil. “You can set parameters, set up current limits, prevent a motor from running backwards regardless of what signal was sent, set up forward and reverse travel limits, set torque limits, determine if the power block has blown up or if the heat sink is too hot, or if the motor went faster than it was supposed to.” Engineers can also read current draw, and set up a drive to transmit specific fault codes that tell exactly what’s wrong.