Motors and Drives Demystified -- Drive System Control Methods [part 1]

Home | Articles | Forum | Glossary | Books

AMAZON multi-meters discounts AMAZON oscilloscope discounts

Introduction

Up to this point, basic drive theory, component hardware, and interface devices have been discussed. It is now time to put the basics to work to develop a drive system. The following information will help tie the components together into a coordinated control system. All systems configurations would be closed loop because of the precise speed and torque regulation required.

In actuality, there would be many more pieces to the system "puzzle" than what is presented here. However, this section is meant to present a general outline of drive systems and how the pieces work together in an auto mated environment.

Coordinated Drive Systems

It is helpful to start with what could be considered a "simple" system and move to the more complex. FIG. 1 indicates one such simple closed loop system.

As seen in FIG. 1, this is a "widget" manufacturing facility. This section is the "finishing" section of the system, with proximity sensors strategically placed along the out-feed conveyor. All of the sensors are connected to an amplifier unit that sends contact signals to the drive. The drive needs to know where the widget is at all stages of the system. Therefore, the job of the proximity sensor is to send a contact closure signal to the drive. This would be considered a digital input (DI). In this case, the drive does not need to know how big, or how long the widget is, just that fact that it has arrived at a particular station on the conveyor. Once the drive has determined the widget has finished all processes, it can then send a relay output signal to the warehouse, alerting that the widget is on its way.


FIG. 1. Sensor control system

In this case, the operator sends a speed reference to the drive. The drive operates the motor at that speed until the widget reaches the first proximity sensor. At that point, the contact closure signal indicates the widget has arrived at the labeling section. The drive takes the contact closure and operates at a preset speed 1 (slower speed), so the label section has time to perform its function. Once the widget exits the label section, another proximity sensor "opens" the preset speed contact, and the conveyor returns to normal speed.

The process is repeated when the widget enters the sorting section. At this point, another digital input is closed, which signals the drive to switch to preset speed 2. At the output of the sorting section, another proximity sensor closes. This indicates that the widget is ready for the warehouse, and the drive returns the conveyor to normal speed.

This is a somewhat crude system but could be considered a coordinated system. The drive could be either DC or AC, along with the corresponding motor. Similar configurations would be seen on packaging systems, food processing systems, and any application where process speed may differ from part movement speed. Later in this section, a view of a more auto mated system will be presented.

FIG. 2 indicates a simple automatic AC drive pumping system. The heart of this system is proportional integral derivative (PID) control.

In FIG. 2, the high level is set for 10 feet and the low level (danger) is set for 2 feet. This automatic pumping system would be set to have a constant level of 10 feet at all times. If the actual value (level) is less than 10 feet, the drive responds to the corresponding "error" signal and activates the pump at a designated speed. In this case, the level is only at 90%, with a feedback voltage of -9 V, indicating a 1-V error and a corresponding increase in speed. Once the error is at zero (feedback at -10 V), the drive slows to zero speed, meaning no pumping action is required.


FIG. 2. Automated pumping application-PID

If this were a duplex pumping system, the drive could be programmed to bring online a fixed-speed "lag" pump, if the demand required it. Several of the latest AC drives on the market include the software intelligence to operate this type of system automatically. Relay outputs can be programmed for a low-level pump start and a high-level pump stop. In addition, a sleep function can be set when the drive stops pumping. The sleep function would cause the drive to release the start command but keep the microprocessor alive, waiting for the next pumping cycle. This function saves additional energy, since the IGBTs are off, as well as any other power electronics circuits.

In FIG. 2, when the level reached a "critical" stage (2-foot level), the drive could be programmed to take emergency action, such as sound an alarm. Some type of level sensor would be set at the 2-foot level. It would send a contact closure to a digital input that would be programmed for an emergency function (e.g., sound an alarm, engage full speed pumping, etc.). The sensor could also send a contact closure to a programmable logic controller (PLC), if that's what was controlling the application.

In Section 5, information on PID control was presented. FIG. 3 is a graphic representation of PID system response and how it might be programmed for the drive shown in FIG. 2.


FIG. 3. PID control-system response

Though indicating meters instead of feet, the graph indicates the response of the drive to a feedback error signal. If the drive was programmed as seen in the graph, it would take about 1 s for the drive/pump to stabilize.

A combination of gain, integration time, and derivative rate are needed to effectively tune this pumping system. By looking at the graph, the drive gain may be set too high, causing the pump to overspeed (overshoot) the desired level. An oscillation would occur until the system stabilized in about 1 s. The integration time would also need review, since the drive may be set to achieve the desired level too quickly for the conditions that exist.

The derivative function could also be reviewed, since it dictates how the amount of error is to be corrected per unit time (i.e., 10% correction per second).

FIG. 4 indicates a device called a scanner used to check the quality of output from a paper machine. The scanner would check paper-quality items such as thickness, coloration, moisture content, surface smoothness, and fiber content.


FIG. 4. Paper machine scanner system

As one unit, the scanner could be considered a sensor that feeds vital information back to the main system control. The scanner would be the input device to the main control, which would signal the drives to change speed- or torque-control output to correct for errors in quality.

Included in the scanner assembly would be a small AC drive that controls the scanner head. The scanner head-speed rate (back and forth motion) would be controlled by the small AC drive (as low as 1 HP). The upper and lower head are connected through mechanical linkages that cause both pieces to move together. The transmitting devices would be in the upper head, and the receiving sensors would be in the lower head, for example.

Tension Control

Before the web of paper arrived at this point, a series of coordinated devices would be in action. Paper, plastic, foil, or any other type of web system would require the use of tension control. FIG. 5 illustrates a basic tension-control system.


FIG. 5. Tension-control characteristics

The tension control system shown FIG. 5 would be part of a web winding system, just ahead of a winder unit. The web material is fed through the in-feed rollers, under the tension control load cell and out through the out-feed rollers.

The purpose of tension regulation is to control the surface web tension as material is wound. The same would be true if the material was being unwound. Tension regulation is achieved by using load cell tension feed back. A load cell is a pressure-sensitive sensor that decreases in resistance as pressure on the cell increases.

Differences between the tension set point and the load-cell feedback allow the tension PI regulator (proportional integral) to develop the error correction needed to maintain tension by trimming either speed or torque of the driven section. The tension PI regulator is updated in millisecond time frames to produce very responsive trim (fine tuning). The proportional gain of the tension regulator can be adapted to accommodate changes in roll diameter.

As the diameter of the roll increases, the proportional gain can be set to increase from a minimum gain at core diameter, to a maximum gain at full roll. A winder application is operated in torque control. The torque-control drive will operate at any speed necessary, as long as the torque (tension) of the web is satisfied. When operating in torque control an encoder is typically required. When using this type of control, accurate web material information is required for calculating the WK² torque value of the roll. These calculations will vary depending on the density and tensile strength of materials. If tuned properly, torque control may result in stable steady-state performance.

Dancer control is similar to tension control. The dancer unit changes out put resistance according to the tension on the web that is wrapped over the dancer roll. The changes in resistance and corresponding voltage change are fed back to the drive so proper torque control adjustments can be made. FIG. 6 shows a dancer-control scheme.


FIG. 6. Dancer position (tension) control

The purpose of dancer regulation is to control the surface-web tension as material is wound or unwound. This is done by monitoring the position of the dancer feedback device, similar to a variable resistor or pot. The dancer is loaded with web material from the in-feed rolls or from another area of the system. The load on the dancer is monitored with the feedback sent to the dancer control drive. The output of the drive is automatically adjusted to achieve desired web tension at the out-feed rolls of the system.

Web tension variations are absorbed by the dancer and cause the position of the dancer to change. The difference in dancer position feedback and the dancer position set point allow the dancer PI regulator (drive 2) to develop the error correction needed to return the dancer to the set point position by trimming the speed of the section. The proportional gain of the dancer regulator can be adapted for roll diameter changes.

As the diameter of the roll increases, the proportional gain can be set to increase from a minimum gain at core diameter to a maximum gain at full roll. The set point position for the dancer is defaulted to the center of the total dancer movement. The regulation position of the dancer can be adjusted by the operator by way of a drive parameter.

The dancer PI regulator is updated in millisecond timeframes to produce very responsive trim (fine tuning). This allows for very stable dancer position control over the entire speed range.

Remote Operator Interface

In many cases, the actual drive must be located at a distance from the motor and driven machine. This is because the drive must reside in a somewhat clean atmosphere, free from dust particles and other contaminants that may be present in the factory. Also, in many cases, the operator needs to be close to the application to verify system operation or to make machine adjustments under safe conditions.

For these reasons, some type of remote-control capability is more often than not necessary. The simplest remote operator device would be a remote start/stop pushbutton, along with an analog speed pot. A step up from the simplest devices would be an operator console, with pushbuttons for preset speeds, jog, forward/reverse, drive reset, speed increase, and speed decrease. It could be as elaborate as pilot lights to indicate each production stage, status indicators, and a full-color monitor and touch screen to enable the complete system.

For purposes of this text, the focus will be on standard hardwired control devices, followed by automated controls with serial and fiber optic communications. FIG. 7 indicates a remote operator station and an operator console.


FIG. 7. Remote operator controls

Any drive, AC or DC, needs two items to be satisfied-start command and speed reference. These remote devices would be wired into the drive's dig ital input and analog input terminals. Additional controls may be needed such as the ones listed above, but basic drive operation would require start and speed command signals.

Most drives have a control terminal block where standard I/O is connected. In addition, terminals or a removable connector or terminal block may be available for future connection to serial communications. Figure 6 8 indicates standard drive I/O connections.


FIG. 8. Standard drive I/O connections

As shown in FIG. 8, in many cases the control connections are easily identified and are laid out on the control board in a logical manner in a sequential number scheme. The analog input circuit must be matched with the signal type that is connected. In many cases, small terminal jumpers function to match either voltage-reference or current-reference inputs. Some drives include a DIP switch or rocker switch for signal matching. To function properly, this matching jumper or switch must be located in the proper location. If it is not, unstable control could result, or in the worst case, the drive would operate at maximum speed, with no speed control. The user's manual and the silk screen indication on the board are the likely places to find the correct settings.

The analog output is normally a connection for an external analog meter.

This meter could be a separately purchased and mounted device or an item included on an operator console. The normal output would be 0-20 mA and scalable to many types of values.

Digital inputs receive their name from the fact that the input is either on or off. Some drives operate on 12 VDC logic, some on 24 VDC, and still others use a 120 VAC interface option. Typically all the drive needs to see at a digital input terminal is control voltage, which is accomplished through a contact closure (a manual switch, limit switch, auxiliary con tact, etc.). Once the drive sees the control-logic voltage at the terminal, the drive performs the operation connected with that digital input (i.e., start, stop, preset speed, reverse, etc.). When control voltage is removed from that terminal, the function stops.

A word about control logic would be appropriate here. Many drives use what is called source control. That is, the terminals on the control board must see a control voltage before a function occurs. In other words, voltage must be sourced to the drive. Some drive manufacturers term this control "PNP" logic, in reference to the transistor regulator types involved with the control logic inside the drive. FIG. 9 illustrates this type of logic control.

In source logic, all of the circuit commons are tied together. The circuit is complete when the control voltage is applied to an appropriate DI.

The converse logic control is termed sink control. In this type of control, the logic voltage is actually tied to circuit common. The control logic voltage sinks to circuit common. When a contact closure is made with a DI, a circuit is closed between the terminal and circuit ground. The internal control logic energizes the function. This type of control is required by some PLC controllers, where TTL (Transistor-Transistor Logic) outputs, or external voltage sources, must be applied for control. This is also used where circuit ground is something other than earth ground. FIG. 10 shows this type of control logic.

If there is a possibility for an unsafe condition to exist, it would be in the fact that inadvertent grounding (connecting to common) of an external DI switch or contact would cause a function to occur. If for some reason the start switch was connected to ground during routine maintenance of the drive, the drive would accidentally start. Many codes or industrial control schemes require a positive voltage at the terminal block before any operation occurs. In source logic, if the start switch were connected to ground, the logic voltage would be shorted to ground and the operation would stop.


FIG. 9. Source control logic connections

It should be noted that some drive manufacturers include a logic-control power supply on the control board. No external power source is needed, only a contact closure to engage a function. If an external source is required by the drive, care must be taken to ensure that polarity and grounding of the external source matches that of the drive. Ground loops and voltage mismatch can cause many aggravating situations, especially at the low voltage or current values being used.

Many drives include several relay outputs as external-control contacts.

These contacts may be dry (not carrying a voltage) or Form C contacts.

They can be programmed for a multitude of operations. For example, the contacts may be programmed to close when a preset speed is selected, or a set speed is reached, or reverse is commanded. In addition, some drives have the capability of monitoring any type of read only information using the relay outputs.


FIG. 10. Sink control logic connections

Some manufacturers call this the supervision function. Quantities such as current, hertz, output voltage, DC bus voltage, and calculated torque can be internally connected to a relay output. When a programmed value is obtained, or exceeded, the relay would change state or energize. The SPDT (Single Pole, Double Throw) contact can be wired into an indicator light, an alarm circuit, or fed back to a system controller like a PLC. With this type of function, the drive can actually be used as a monitoring tool. It has the intelligence to compare values and make indications when values are not reached, when they are exceeded, or when they occur.

Some manufacturers use digital outputs instead of relay outputs. A digital output would supply a voltage or current value to an external device, when a programmed value is met.

Serial and Ethernet Communications

During recent years, the push has been to automate many operations to improve product quality or to maximize the efficiency of the system. Communications is a necessary factor in automated systems. Generally speaking, the most common type of drive communication is through a serial link. In this scheme, the data is transmitted in a serial fashion (bits are transferred sequentially, one after the other). When talking about drive communications, three modes currently apply: serial, fiber optic, and Ethernet (intranet) communications. The drive control schemes using each of these modes will be explored, as well as their connections to higher level control systems.

Since serial communications involves sequentially transmitted data, it would not be considered high-speed. Typical communication rates include 4800, 9600, and 19,200 baud (bits per second). This transmission speed is acceptable when communicating with an air-handling unit, which moves volumes of air in and out of the building. This speed would not be acceptable in a paper machine, where high-speed data transmission is critical to the success of paper density, coating, and composition. In many cases, this type of communication is ideal for 24-hour monitoring of drive operation. FIG. 11 indicates this type of monitoring function.


FIG. 11. Serial communications setup

In most cases, the drive serial link is an RS-485 configuration. With this type of connection, the total communication network length can be ~4000 feet. To accomplish this network matching, the computer must have an RS-485 output. If it does not, then a converter such as the one listed in FIG. 11 would be needed.

There are many different languages (protocols) available for industrial applications. The HVAC industry also has several protocols available specifically designed for air handling, cooling tower, chiller, and pumping applications. Several companies have been pioneers in the PLC market and have developed their own specific protocol. The drives pictured in FIG. 11 have the same protocol as that available in the computer. If the protocol matches, then communication is possible after the computer and all drives are programmed to accept serial communications.

In many cases, because of transmission speed and protocol limitations, the total number of drives on a serial link may be as high as 32. In practical terms, if communication speed is at issue, the total number of drives may be around half of that. With a 9600 baud rate, the communication speed starts at 100 ms and would drop from that point, with every drive added to the network.

Typical programming for the drive would include ID number, baud rate, protocol selection, fault function, and communication time out. The computer would also be programmed in a similar fashion to match the drive communication speed and inputs.

Connection to a building's Ethernet or network system is now becoming more popular. There are definite advantages to this type of communication connection. Information gathered from the drives can be easily down loaded to a mainframe network computer. Trends can be identified and analysis can quickly be done, as opposed to transferring data to an independent system and then transferring to the network. In addition, connecting to a building's drive system can easily be accomplished though the Internet, which would be available through a modem connection.

Several software companies offer software that allows the user to develop customized screens that indicate parts of the process. This would include drive operation, inputs to the system, and outputs from the system. Color graphics are fast becoming the interface media of choice. FIG. 12 gives an example of this type of network/drive communications.


FIG. 12. Ethernet/drive communications.

The Ethernet or network communication speed is higher than that of the serial link. Though the network speed is high (in the low mega-baud range), that doesn't necessarily indicate a rapid transfer of data from net work to drive, or vice versa. Information from the network must be transmitted to the drive, and the drive's microprocessor must be able to efficiently decode the information. With the drive's slower serial link rate, only a limited amount of information can be handled per second. The processing speed of many drives would be in the 25- to 50-MHz range, with 2-5 MB of memory available. By today's computer standards, this rate would be extremely slow. However, by today's drive microprocessor standards, this speed is quite adequate for processing all internal drive information. As drive microprocessor technology improves, internal processing speed will also improve.

When connecting the serial link to the drive, it is done in a "daisy chain" fashion. FIG. 13 indicates this procedure.


FIG. 13. Serial link connections

When wiring the drives, care should be taken not to "daisy chain" the shields. Many drive manufacturers include termination resistors on the control board. These resistors reduce electrical noise on the communication network. Termination resistors are to be included on the first and last drive in the network. The inner drives are not to be terminated. (Too many resistors in-series with the communications devices may cause the network to malfunction.) Termination resistors are typically placed "in circuit" by moving a pair of jumpers into position across several contact points.

Fiber-Optic Communications

The use of fiber-optic communications has steadily increased over the last few years. Optical fibers or "light pipes" are highly immune to electrical noise (EMI and RFI) and have the capability of transmitting data over long distances. Glass-constructed optical fibers, along with periodically placed amplifiers, would allow transmission over thousands of feet. The glass type of optic fiber is quite costly compared with plastic. When comparing optical fiber with hard-wired control, the user must decide between high noise immunity and higher installation costs and lower immunity and lower installation costs.

In addition to drive-to-control fiber-optic communication, an increased use of optical fibers is seen in modern drive circuitry. The advantages of fiber optics for building communications certainly holds true for internal drive control and communications. FIG. 14 illustrates this type of internal communications.


FIG. 14. Internal fiber-optic communications

In FIG. 14, the fiber-optic cables interconnect between the main control board and the other definite purpose boards. In addition to the internal communications, fiber optics can also be used to interface with monitor and programming software, as well as optional PLC modules.

In a higher-level system, fiber optics is the normal. Its high speed (~4+ mega baud) communication makes this format ideally suited for industrial applications. FIG. 15 shows an external drive interface and a PLC.

As seen in FIG. 15, the drives are connected by fiber optics in a "ring" structure. The fieldbus option module is purchased from the drive vendor.

It changes the protocol of the PLC into a language that the drive can understand.


FIG. 15. Drive and PLC interface

The only possible disadvantage to this type of communications is the fact that if one drive goes down on a fault, the entire communication network goes down with it.

Programming is similar to that of serial communications. Each drive on the network needs a unique ID, as well as baud rate, communication fault function, and communication time out.

When installing a fieldbus module to a network, care must be taken in connecting the transmit and receive fibers in the correct positions. Figure 16 illustrates this procedure.


FIG. 16. Transmit and receive fiber-optic connections (ABB Inc.)

In FIG. 16, output from the drive (TX) must be the input to the Field bus module (RX). The same holds true for the receiving optic cable. Also be sure that the cable plugs are completely seated into the receptacle by listening for the "snap" action. This allows for the maximum transfer of "light" data into the receptacle.

cont. to part 2 >>

Top of Page

PREV.   Next | Guide Index | HOME