By David Stephan, engineering manager, PCT Ebeam and Integration
As industrial automation and drive systems evolve, the transition from DC (Direct Current) to AC (Alternating Current) motors is becoming increasingly more common. DC systems are being phased out in many industries in favor of modern AC drive technology. While AC motors offer numerous advantages, companies need to be aware that the upgrade process still involves significant engineering considerations. This paper aims to assist engineers and system integrators in making informed decisions by highlighting several performance characteristics, design differences and upgrade challenges that frequently arise during the DC to AC transition.
In the early 1900s, DC motors were the default choice for industry [1]. DC motor speed is controlled via voltage change, a technology that was widely available and understood at the time. The complexity of a DC motor lies in its hardware design, not its controls. AC motor speed, however, is tied to the frequency of an applied voltage. Altering an AC motor’s speed via frequency was more difficult in the early 20th century, especially before the invention of transistors. Additionally, DC motors provided rated torque down to zero speed, allowing for better line control at slow speeds, a capability AC motors at the time lacked.
As the 20th century progressed, so too did the controls technology needed to make the AC motors a viable choice [2, 3]. Today, AC motors are standard in many applications, primarily due to maintenance benefits. Unlike DC motors, which have relatively simple controls but complex motors, AC motors have more sophisticated controls but are mechanically simpler and more reliable (see Figure 1). An AC motor consists of two main parts: (1) a ball of wire that rotates, called the rotor, which is surrounded by (2) a stationary ball of wire, called the stator.
DC motors are more complicated, featuring an armature and field (equivalent to a rotor and stator) along with a commutator. The commutator is a split-ring copper component that contains brushes. It is a wear point in the motor, requiring periodic brush replacement and occasional grinding. This maintenance makes DC motor upkeep more of a hassle and increases downtime compared to almost maintenance-free AC motors.
Modern AC motors, with solid-state devices and sophisticated digital controls, can achieve the same performance as DC motors. As the industry has shifted to favor AC technology, vendor availability and motor options have increased for AC motors and decreased for DC motors. Moreover, this shift has been ongoing for several decades.
Despite the declining popularity of DC motors, some companies choose to retain them. Older DC motors were built ruggedly. Replacing larger motors (i.e., ≥500 HP) also is more expensive, and the risk associated with changing them increases with motor size. Companies with large DC motors, such as steel rolling mills, often choose to invest in training plant personnel with DC expertise to maintain their knowledge base rather than risk an AC upgrade.
Key upgrade considerations
It’s easy to assume that all a successful DC to AC upgrade requires is to match the specifications of the new AC system to the existing DC system. However, between oversimplifying what specifications need to be considered and not factoring in the differences between AC and DC systems (e.g., physical size, how they work, etc.), what looks like a simple swap on paper quickly can become a headache during the upgrade.
What follows is a list of commonly overlooked upgrade considerations, starting with those that cause the most significant problems. Experienced integrators might find it a convenient checklist for projects. End-users hopefully find this list helpful as a method of sussing out knowledgeable integrators.
Motor sizing
An exceptionally important, and frequently missed, engineering issue when upgrading is sizing motors. Motors are rated in horsepower (hp), but horsepower is only one part of the equation. Consider, for example, a 40-hp tractor compared to a 40-hp motorcycle. Both motors have the same horsepower, but their performance is drastically different – the tractor motor is built for pulling, whereas the motorcycle motor is built for speed. Similarly, motors for industrial applications can have the same rated horsepower but different performance specifications. For this reason, it should be illegal to size motors in horsepower alone!
As shown in Equation 1, horsepower (hp) is related to torque (lb-ft) and motor base speed (rpm) [2].
EQUATION 1.
In roll-to-roll manufacturing lines, appropriate torque is needed to accelerate and decelerate rolls of material and to provide tension. Factors that influence the base speed requirements include line speed, roll size and gear ratio. Two of the three factors – horsepower, torque and base speed – must be specified to size a motor correctly. Otherwise, a tractor/motorcycle scenario could result: A motor may have enough torque but not enough speed, or enough speed but not enough torque, to meet the line specifications.
While identifying the correct motor size always is important, it is a particularly easy misstep to make during a DC to AC upgrade because of the possible differences in base speed. A 4-pole AC motor, which has a base speed of approximately 1750 rpm, is standard for many industrial applications. In contrast, there tends to be a variety of base speed options for DC motors. Since Equation 1 applies to both AC and DC motors, a difference in base speed when switching from AC to DC will result in a difference in torque for a given horsepower. For example, a 40 hp/500 rpm motor will produce 3.5x more torque than a 40 hp/1750 rpm motor.
Cost is another factor that can lead an engineer astray. Comparing two 40 hp motors, the lower-cost option is enticing. Yet, here again, the difference often lies with the motor torque and base speed specifications. A lower base speed/higher torque motor will be more expensive due to the higher current draw, which requires more copper (see Equation 2).
torque ∝ amps ∝ copper ∝ $$$
EQUATION 2.
The need for higher torque often is seen with unwind and rewind motors, so it is common for them to be more expensive than other drive motors on the line. Understanding the relationship between horsepower, torque and base speed is vital for proper motor sizing as well as cost accounting.
Overload capabilities
Older DC motors and drives often were built with very high overload capacities, sometimes as much as 250% for three minutes. These capabilities frequently were leveraged to increase line speed or tension without upgrading the motor or drive. A DC motor used as a rewind, for instance, could produce a larger diameter roll or higher tensions by operating in an overload condition for a few minutes at the end of a run.
Unlike DC motors and drives, AC systems typically have low overload ratings (e.g., 110% for one minute/160% for three seconds). This difference between DC and AC systems means that if the DC overload capacity is being used, even correcting sizing an AC motor based on the existing DC nameplate specifications (and Equation 1) may leave the line short of its previous capabilities.
Figure 2 is a simplified depiction of two examples of the current draw of a rewind motor over time. In the first example, Load #1, the original line capabilities are represented, and the rewind is stopped just before overload capacity is needed. Load #2 represents an adjustment (maybe made by an eager plant engineer) to the line to allow running more tension or a larger diameter, and it does require the use of the overload capabilities. Load #2 highlights the mismatch in overload capacities between DC (blue box) and AC (red box) motors. An AC motor sized to match the rated torque of the existing DC motor (recall Equation 2) won’t be fully capable of Load #2 and will be forced to stop shy of four minutes into a five-minute run. An appropriate AC replacement for the DC motor, in this instance, needs to be sized to include the overload capacity being used.
Whether or not overload capacity is being used on an existing DC system can be determined by reviewing documentation (if it exists) or, better yet, taking field readings to make sure the system has not been modified from the original line specifications.
Frame sizes
AC and DC frame sizes, which include information such as motor shaft height, bolt patterns and mounting base dimensions, are not the same. If the shaft to foot dimension is taller on an AC motor (see Figure 3), the mounting base will need to be modified. If the motor is mounted in steel, it is relatively easy to modify, but it can be an expensive endeavor when the base is concrete. In either case, be aware that these modifications will require a mechanical design and labor effort.
With a large portion of a DC to AC upgrade being electrically focused, it is easy to forget or underestimate the mechanical contributions; however, mechanical rework can add significant cost and time to a project and should be included in the budget. In addition to mounting base modifications, other common mechanical rework during DC to AC upgrades may include alignment and shaft dimension adjustments and/or changes to the motor/drive train interface (i.e., gearbox).
Inverter duty-rated motors and cable
The voltage output of DC drives is steady at a constant speed; it does not put any undue stress on the system and requires no particular accommodation. The same cannot be said about AC drives.
The output of an AC drive is a square-wave with a varying width, called pulse width modulation (PWM, see Figure 4). This PWM waveform looks like a sine wave to the motor. Special transistors, called insulated gate bipolar transistors (IGBTs), in PWM drives switch very fast. This can lead to a transmission line effect which results in waveforms that double the voltage seen by the motor and wire. Additionally, the fast voltage rate of change (dv/dt) of the PWM stresses the motor and wire insulation.
Both of these issues lead to premature insulation failure in non-inverter duty motors and cable. The solution is simple: Use inverter duty motors (meeting the NEMA MG1, Part 31 standard) and cable. While that is a simple solution, it is not a cheap solution. Inverter duty-rated motors and cable are more expensive, but worth the cost as they inevitably will prevent unexpected downtime and replacement motor and cable costs. If the upgrade already has been made to an AC motor that is not inverter-duty, chokes or terminating resistors can be used to mitigate this problem. In addition to better insulation, inverter duty-rated cable also is shielded. The PWM waveform is exceptionally noisy, and the cable run from the drive to the motor basically is a long transmission tower in the plant. Without proper shielding, the waveform is likely to cause interference with communications and even digital I/O signals.
Regeneration
Regenerating loads, such as unwinds in web-handling lines, pull tension against the rest of the line. As the unwind is overhauled (forced forward as it pulls backwards), it becomes a generator of electricity. That excess power must go somewhere. If it is not dissipated somewhere appropriate, the excess power can build up and cause a
drive fault, or worse.
DC drives typically have a reverse bridge to take the power generated by the unwind and send it back into the power system. AC drives most often do not have a reverse bridge. AC drives use what’s known as a “front end” to convert AC power to DC, then the IGBTs invert the DC power back to AC. The trouble is that the front end generally is made up of diodes, which allow power to flow only one way. More manufacturers are coming out with AC drives that have an active front end (AFE), capable of two-way flow to allow power to transfer back to the plant, but this is not yet standard.
An AFE AC drive is one solution for regeneration in AC systems, but it is not the only option. Most AC drives have a chopping transistor (also called a braking transistor or braking IGBT) to dissipate excess energy. An appropriately sized resistor, purchased separately, can be used to handle this excess energy from the transistor. This resistor often is large and needs to be mounted externally to the drive enclosure. Another option is to set up the drive system as a common or shared bus to allow all the other (non-unwind) drives to use the energy generated by the unwind [5].
Encoders
Speed feedback (aka closing the speed loop) is an important component for smooth roll-to-roll operations with both DC and AC drives, and one method of providing that feedback is to use an encoder. AC drives, however, also use the encoder feedback to know the position of the rotor relative to the stator’s magnetic field to accomplish vector control of the drive. Vector control allows the drive to produce rated torque down to zero speed. Without an encoder, running in a sensorless mode, AC drives cannot produce rated torque at zero speed – a limitation, recall, that gave DC systems an advantage in the early 20th century.
While encoders are beneficial for precise control and low-speed performance, they sometimes are not included on industrial lines because they can be prone to failure. Companies assume that they are better off without an encoder, especially those with process lines that require lengthy start ups, because they can avoid unexpected downtime as a result of encoder failure. To ensure that low-speed performance doesn’t suffer after the upgrade and that downtime due to the encoder is prevented, the drives can be programmed to switch to sensorless control upon encoder failure. Speed and torque control may not be as good in the interim, but the line can keep running until maintenance can be scheduled and the replacement encoder installed.
The bottom line: Use an encoder.
Inertia matching
A rule of thumb for designing drive systems is that the motor inertia (JM) should match the load inertia (JL). This improves dynamic system performance, leading to better speed and tension control. It is especially important for higher performance applications, including web-handling systems, precision indexing applications and tasks with high acceleration and deceleration rates.
It is unlikely an AC motor will be physically the same as the DC motor it is replacing, so it also is unlikely the motor inertia of the new motor will be the same. Ideally, the motor and load inertia values are equal, but it also is acceptable to match the inertia ratio (JM: JL) that previously existed with the DC motor, assuming line performance was satisfactory.
Holding brakes
Last, but not least, don’t forget about holding brakes. These brakes typically are on unwinds and rewinds, sometimes nip rolls, but they could be anywhere, and they are easy to forget about when looking at electrical schematics (another reason to have a mechanical engineer around). Brakes can be expensive – costing as much as the motor itself – and they generally can’t be swapped between DC and AC motors. Failure to capture this cost is going to ruin someone’s day.
Don’t forget about brakes. If the line already has them, it’s probably going to need them in the future.
Post-upgrade expectations
After the upgrade is complete, it is easy to think everything will be peachy and the process line will run like it never has before – it is an upgrade after all. The reality, however, is that a DC to AC upgrade is not a cure-all (far from it), and there are few expectations that should be managed going into the process.
Line performance
The purpose of a DC to AC upgrade is to update antiquated technology, not to upgrade line performance. New edge guides, larger motors, better resolution load cells – these things don’t magically appear when the switch is made from DC to AC.
If the line needs to perform beyond its original specifications, more engineering will be needed to make that happen. It’s not a bad idea to incorporate performance upgrades into a DC to AC upgrade, since there already will be an outage on the line, but these changes need to be factored in from the beginning to ensure a smooth and efficient upgrade process.
With that said, sometimes just getting a line back to its original specifications can seem like an improvement, since over the last decade or two everyone has forgotten what the line originally was capable of. Freshly tuned controls also only help the situation. Maybe the current integrator on the project has a trick or two (e.g., better inertia compensation skills) that the last one did not.
One downside, however, to restoring a process line to its original specifications is that sometimes line components begin to break. If a line has been routinely run at 600 fpm for years, and now the line is running its original maximum speed of 1,000 fpm, that bearing that hasn’t been oiled in a while might not survive the increase in speed. Be wary of components that have been added over time to the line, as well. These components might have been rated for the speed or tension the line was capable of at the time of their installation, not necessarily the original line specifications.
Technology learning curve
After a DC to AC upgrade, there also is a learning curve for maintenance personnel. Going from an old drive with a few potentiometers and switches to a modern digital drive with thousands of parameters and its own unique software package to configure and troubleshoot is a big step. This is a good time to lean on the integrator and local distributors for help and training. Have plant personnel involved during start up to get some hands-on training. Interesting (i.e., weird) things happen during start up, and it’s a great time to see how to troubleshoot new drives.
If training is needed, be sure it is included in the quote.
Safety procedures
Any major work or upgrade on a process line, safety should be considered before, during and after. One specific example of where safety procedures might need to be updated after a DC to AC upgrade is if a common and shared bus system is used to mitigate regeneration. When the drives share a common power source (the DC bus), they all need to be locked out together. If there is a procedure where the rewind needs to be locked out, but the rest of the line stays powered up, in a common or shared bus architecture, that is difficult to do. If safety had been considered in the planning stage, maybe a common/shared bus wouldn’t have been the best solution.
AC drive shelf life
It is standard practice for some companies to have spare parts on the shelf to shorten downtime after a part fails, and, after a DC to AC upgrade, one of those spare parts may be an AC drive. It’s important to know that the capacitors in AC drives have a shelf life. They need to be reconditioned if the drive has been sitting on a shelf for an extended period. Reconditioning involves slowly applying voltage to the drive. It’s not overly difficult to recondition a drive but not doing it will lead to spectacular failures (the capacitors blow up).
Conclusions
In summary, the industrial world is leaning away from DC and toward AC; almost all newly installed equipment is going to be AC, and most motor and drive upgrades are very likely to be AC. There are a lot of good reasons to go AC, but – like a lot of things – there also are key considerations to be aware of. From correct motor sizing to regeneration to remembering to include holding brakes, the physical and fundamental differences between DC and AC can affect multiple aspects of the line. A good integrator can manage these differences and explain how they impact process line function and safety, as well as upgrade costs. It takes a thoughtful and thorough evaluation to implement an AC upgrade that performs to the same level as the existing DC line, so be sure to have a competent partner when upgrading.
References
- Bose, B. K., Power Electronics, 2014. https://ethw.org/Power_electronics
- Wilson, J., Solid State Drive Fundamentals. TM-1800-A. Reliance Electric.
- Drives Engineering Handbook. Rockwell Automation. https://literature.rockwellautomation.com/idc/groups/literature/documents/at/drive-at001_-en-p.pdf
- Electric Motor Simulation: Power Tool for Design Optimization, Neural Concept. https://www.neuralconcept.com/post/electric-motor-simulation-powerful-tool-for-design-optimization
- Drives in Comon Bus Configuration. Allen-Bradley. 2024. https://literature.rockwellautomation.com/idc/groups/literature/documents/at/drives-at002_-en-p.pdf
David Stephan is the engineering manager for the System Integration team and an owner of PCT Ebeam and Integration. PCT provides both electrical and mechanical engineering services, primarily for customers that produce metals and films. PCT also manufactures industrial electron beam systems used for cross-linking film, curing inks and coatings and sterilization. Stephan’s personal areas of expertise include automation control systems, coordinated drive systems, custom machine design and fabrication, combustion systems and safety. He has extensive experience with drive systems, primarily in the film and metals industries, and has been with PCT for over 30 years since graduating from Iowa State University with a BSEE. He can be reached at 563-285-7411, email: david.stephan@pctebi.com, www.pctebi.com.