The great debate: Slip control versus field-oriented control
Two leading motor control technologies campaign for their competing approaches.
Author's note: With last year's hard-fought election still in recent memory, the last thing you might want to read is an article reminiscent of that political process. However, I think we can have a little fun with this topic by casting two competing motor control technologies as candidates, pitting them against each other in a heated debate, and letting you decide which technology wins.
Moderator: Good evening ladies and gentlemen, and welcome to this final debate between our two motor control candidates. To your left, please welcome the incumbent technology, Constant Slip Control, or simply “Slip” as he prefers to be called. And on the right, please meet our challenger, Field-Oriented Control, better known by his nickname, “Vector.”
I would like to address my first question to Slip. Considering that you are “slipping” so far behind in the motor “poles,” do you really believe you are a serious contender for variable speed drive applications?
Slip: Despite my opponent’s current lead in the “poles,” let me make one thing perfectly clear: I bring many years of tested experience to this discussion and represent the preferred choice of design engineers worldwide when it comes to controlling AC induction motors. My opponent makes many promises, including increased transient response. But are we ready to embrace the risky schemes represented by these claims? By the end of this debate, it will become apparent why I represent the most economical, robust, and energy-efficient choice for induction motor control.
Moderator: Is that true, Vector? Do you really represent a more risky approach?
Vector: Not at all. You can obviously see that this is a desperate attempt by my opponent to smear and distort my extensive record, which reveals my unconventional approach to motor control. Thousands of motor control designers around the world cannot all be wrong. As microcontroller and Digital Signal Controller (DSC) costs continue to drop, expect to see more and more motor control systems benefit from the advantages I offer over my opponent, which will become evident by the end of the evening. With a long history of transforming motor designs around the world, I am clearly the maverick you need to buck the failed policies of the past and lead us confidently into the future. In fact, it has been said that I am perhaps the most significant discovery in the motor control world since the AC induction motor itself …
Slip: I have worked with induction motors for many years. Induction motors are friends of mine. And you, sir, are no AC induction motor!
Also, you do represent a risky approach, compared to the simplicity of control that is the cornerstone of my campaign. To prove this, please refer to Exhibit 1, which shows a block diagram of a variable speed AC drive utilizing slip control. The system shown is based on voltage mode control, but current mode systems are also popular.
To keep from overfluxing the machine during transient conditions, the voltage is limited by a predetermined V/Hz law, as shown by the voltage limiter block. Instead of dual current loops with precarious frame transformations, complicated flux estimators, and risky sensorless back EMF observers, we see a kinder, gentler control system where motor voltage and frequency represent the controlled variables. The desired slip is supplied directly as an input to the system, which can be dynamically and optimally adjusted to maximize torque, efficiency, power factor, torque per amp – you name it.
In this particular system, voltage is controlled independently from frequency to provide optimum defluxing of the machine under light loads to save energy. The energy savings resulting from this approach compared to the rated flux control of my opponent are shown in Exhibit 2, where the slip in my system has been adjusted to optimize motor efficiency. In both cases, the same 3 HP induction motor model is used.
Given that motors consume more than 50 percent of all electricity generated, can we really afford the risky and wasteful energy-spending policies advocated by my opponent?
Vector: There you go again. You should really check your facts before you start talking about my energy plan.
My opponent has compared the energy savings of his plan to a field-oriented system operating with rated flux. He either doesn’t know or doesn’t want you to know that the same energy savings can be obtained in a field-oriented system by simply lowering the d axis current, which directly reduces the machine flux. In Exhibit 3, we see a field-oriented system where the motor flux and motor torque can be controlled independently, much like what can be done in a DC machine with a separate field winding.
My opponent cannot directly control motor flux; he can only affect it indirectly by changing voltage. He also mentioned that various values of slip can be commanded to optimize different motor operating parameters, including efficiency. Well, it turns out that the desired steady-state slip of an induction motor can just as easily be set in a field-oriented system by adjusting the ratio of d and q axis stator currents as shown in the following equation, where Rr and Lr are rotor resistance and rotor inductance, respectively, as shown in Equation 1.
So, everything my opponent can do, I can do better. But what he failed to tell you is that when you reduce the flux to increase efficiency, it’s like putting the motor to sleep. This significantly increases its susceptibility to sudden or unexpected torque perturbations, as shown in Exhibit 4. It can take hundreds of milliseconds to get the motor to wake up and get the flux reestablished. In some cases, you can even stall the motor. That’s why many designers leave the flux at its rated value if fast transient response is needed over a wide torque range.
But I can understand why my opponent doesn’t want to talk about transient response – because constant slip control is based on a slower, steady-state model of an induction motor. As much as he would like to dress up his performance in this area, he can’t. You can put lipstick on a pig, but it’s still a pig. Fast transient response represents a significant advantage over my opponent’s failed control policies, which in the past required DC motors to be used if fast response was needed.
Slip: First, let me address the transient response issue. It’s way overrated. Sure, you occasionally run into a rare application that needs lightning-fast torque response. But in most cases, once you have it, you spend the rest of your design time trying to mitigate its effects. Fast torque response can translate into high levels of jerk in your system, which can cause acoustic noise and premature mechanical wear.
My dear ol’ grandma always used to say, “Why pay for a 1 GHz op-amp when a 1 MHz op-amp will do the job?” Take a washing machine, for example. With all of the mass associated with a loaded drum, why in the world would you want or need super-fast torque response? All of that extra bandwidth will only get you into trouble. It’s very difficult to tame and could potentially lead to instability problems in your motor control system down the road.
But I think this whole discussion about transient response is an attempt to take the focus off of the real issue. Over my desk I have a sign that says, “It’s the economy, stupid.” Why would anyone pay for the expensive DSP required to do field-oriented control when you can meet your requirements with slip control running on a processor that costs half as much? Take a look at Exhibit 5, which shows how a slip control system can be implemented on a simple 8-bit processor.
My opponent is eager to point out that the cost of DSCs is dropping. But a simple 8-bit processor with a von Neumann architecture will always be proportionally less expensive than a 16-bit machine based on a Harvard architecture with multiple internal buses. With our economy in such bad shape, can your design budget really justify this kind of lavish pork barrel spending?
Vector: My opponent seems anxious to talk about the economy. OK then, let’s talk about the economy. Exhibit 6 shows a field-oriented system that utilizes a Freescale DSC. My opponent mentioned the lower processor cost associated with slip control. But I submit to you that the system cost is more important when considering your design budget. When you compare Exhibit 6 with Exhibit 5, you immediately notice the hardware similarity between our two designs.
The processor cost in these systems, as a proportion of the total system cost, is typically somewhere between 2 and 10 percent, depending on the motor drive horsepower rating. So the processor cost ratio that my opponent loves to cite has very little impact on total system cost. On the other hand, the selected processor and control topologies have a huge impact on system performance. So, from a system perspective, field-oriented control represents the best economical choice.
Another issue my opponent doesn’t like to talk about is other motor topologies. I realize that tonight’s debate is focused on AC induction machines. But before you vote on which control technique is best for your application, consider this: Many of the library routines that comprise field-oriented control can easily be ported over to other motor topologies. Slip routines only work with induction motors. As software resources become more precious, can you really afford to rewrite your motor control algorithms every time you choose a different motor type? Whether my opponent will admit it or not, I bring a new level of standardization to the motor control industry, representing the hope and change we can believe in for the future.
Finally, I’m getting tired of the accusations that my approach is “risky.” This is simply an attempt by my opponent to cover up the failed control policies of his past. I’m going to say this again: I have not had any relations with any motor that were considered risky in nature. Field-oriented control is not risky. It is well understood by almost everyone in the motor control community. Even Joe the Plumber could do it! Field-oriented libraries and control blocks are readily available from most semiconductor suppliers, including Freescale Semiconductor at www.freescale.com/motorcontol.
Moderator: Slip, you have one minute for rebuttal.
Slip: Field-oriented control or vector control is risky. It’s clearly more complicated and requires a higher level of motor control technical savvy than I do.
But let me go back to the economy for a moment. A slip-controlled system is cheaper than a vector-controlled system. It’s just true. We can argue about how much cheaper, but that’s not the point. Shouldn’t the responsibility of every design engineer be to design the most robust system that meets the required specs for the lowest possible cost? This is especially true for high-volume applications. When multiplied by the compounding effect of high-volume manufacturing, the cost savings per system result in quite a stimulus package for your company. But I do agree with my opponent’s comments about Freescale. They can handle both of us.
Moderator: Well, what do you know! We finally found something you both can agree on. Unfortunately, gentlemen, that’s all the time we have for tonight’s debate. I want to thank each of you for sharing your perspectives on this important topic. I also would like to thank the audience for attending our debate this evening. Finally, thanks to Industrial Embedded Systems for sponsoring tonight’s discussion.
Dave Wilson is motion products specialist at Freescale Semiconductor, based in Milwaukee, Wisconsin. He has 30 years of experience working on projects ranging from nuclear pulse processing to artificial intelligence pattern recognition, and has designed motor control systems as simple as trigger controls for power tools and as complex as a six-axis DSP servo stage controller for a scanning electron microscope. He is also the author of several articles, patents, and conference papers related to motor control. Dave holds a BSEE from John Brown University and an MSEE from the University of Wisconsin.