BLDC motor control questions 直流无刷控制常见问题

Q&A: BLDC Motor Control Class Questions
Q) Which method is more power efficient? Six step or sinusoidal?
A) Sinusoidal control will be more power efficient as it will directly control the current flowing through the motor.
Q) Does sinusoidal control necessitate FOC (field-oriented control)?
A) There are methods of using sinusoidal control combined with six-step, but FOC has many advantages.
Q) When the rotor has multiple poles, how do you define the direction of the rotor magnetic field? Doesn't it have multiple magnetic field directions?
A) In the examples we were only talking about a single pole-pair motor. In motors with multiple pairs, there will be additional instances of the 3 motor phases as well.
Q) Why are Delta and Wye topologies used, as opposed to using 3 separate coils?
A) Those two topologies are used since the three motor phases need to be connected at the neutral point so that the current will return to ground.
Q) If you have very accurate angular information of rotor position, and you have fine-grained control of current in each coil, is FOC necessary/does it provide any advantage?
A) FOC is simply a way of achieving those objectives.
Q) Is FOC a form of sensorless control? Or are they two different things?
A) FOC could be sensored or sensorless. FOC is the algorithm used to control the motor; it could use external sensors or not.
Q) During start of motor, if the rotor is in between two phases, then how do you excite the stator?
A) For trapezoidal (six-step), it is unlikely that this would happen. The worst case would be if when the system starts, it might off by one Hall state for a very brief amount of time. It probably wouldn't cause any startup issues.
Q) Normally lots of noise is introduced on Hall sensor signals in motor. How do you tackle this issue? A) Digital filtering is a good way. Many microcontrollers have inputs with built-in debounce or input qualification.
Q) How do you judge which type of control (six-step or field-oriented) to implement for BLDC motors? A) It really depends on your application requirements. FOC is more complex but has many advantages for dynamic performance, while Trap is simpler and works well for basic applications.
Q) What is the optimum time delay between each step in the six-steps techniques?
A) The optimum time-delay will depend on how fast your motor should be spinning. In most applications, there is a speed controller that will dictate how fast each step is advanced.
Q) What is the relationship for torque (similar to point ignition on car) between the advance and retard of commutation?
A) Advancing and retarding the commutation can help with torque ripple it can also help make your controller more efficient as it will push your system so that it is changing commutation at the ideal time.
Q) How does delta or Wye wound stator affect things?
A) Delta vs Wye has no impact on the control of the system. The biggest impact is if you manually measure the motor electrical parameters.
Q) Does sinusoidal control assumes sinusoidal back-EMF?
A) It does, but the vast majority of BLDC motors have sinusoidal back emf.
Q) What makes the Back EMF Trapezoidal or Sinusoidal?
A) The shape of the magnets and the shape of the stator windings. Their interaction will dictate the shape of the back emf.
Q) What about Trapezoidal control?
A) We covered Trapezoidal control in more detail in the second class.
Q) How is Bsta and Brot angle be measured?
A) Bsta & Brot don't need to be measured directly. We use them to describe the behavior of the motor but all of the measurements happen on the motor phases.
Q) Resolver?
A) Resolvers are able to provide excellent source of rotor angle feedback for FOC control.
Q) Do you need FOC to implement six-step control?
A) Nope, they are separate control techniques each with their own advantages and disadvantages.
Q) Is zero crossing possible using Delta wound motor?
A) Yes, the diagram shown is great for teaching but isn't the most accurate with respect to the actual construction of motors. Both delta & wye work great for zero cross detection.
Q) Is there a type of motor that would be more efficient with a six-step control than a sinusoidal control?
A) A motor with extremely trapezoidal back-EMF could be more efficient with six-step, but that is a rare case. Most motors will be more efficient with sinusoidal control.
Q) Is there more switching loss using Sinusoidal control than Trapezoidal control. If so, would that make Trapezoidal control more power efficient than Sinusoidal control.
A) There can be more switching losses, but often thses losses will be less than the efficiency gains from using sinusoidal control.
Q) Can you guarantee start rotation with FOC sinusoidal?
A) When using a sensor you can guarantee startup. When using sensorless you can't guarantee startup
but there are many ways to help ensure the motor starts. One of the ways is to use additional current or a more advanced algorithm such as HFI.
Q) Is it correct to state that adding ID changes the magnetic field of the stator, which changes the voltage-constant of the motor?
A) It changes the magnetic field of the rotor. Which does change the voltage-constant of the motor. This is how field weakening works.
Q) How do you know the initial position of the rotor before the motor starts in order to know which phase to energize for proper rotation direction?
A) Hello Kevin, you don't know it unless you have some sensors. That's why we always "align" our rotor before starting the motor. Align means that we apply a known vector that will move the rotor and align it to the vector we are generating. Then, we can start our algorithm.
Q) Can you use six step AND FOC to control the same motor?
A) You can use either FOC or six-step to control the same motor.
Q) Is PWM implementation of a sinusoidal can fed directly to the motor windings in order to achieve precise control or do we need in some way to convert the PWM into clean sinusoidal before being fed to motor windings.
A) The PWM frequency is much greater than the frequency of the sinusoid that you are trying to produce. Since the motor is simply an RL filter modulating the PWM in a sinusoidal pattern will create a clean sinusoidal signal in the windings.
Q) How is sinusoidal control manifest compared to 6 step? Does it appear that the same motor has a higher motor constant?
A) Sinusoidal control is simply an alternative control technique; it doesn't impact the properties of the motor.
Q) Is sinusoidal control possible for trapezoidal BLDC motor?
A) Yes, sinusoidal control can be used on those motors.
Q) How is speed/torque control implemented with six-step control?
A) Speed control requires a separate controller that will adjust the amount of voltage applied to the motor. The amount of voltage will be proportional to the desired amount of torque.
Q) Does the commutation happen in between the sector or at the start/end of the sector?
A) The commutation happens when the Hall sensors indicate that the motor phases need to be switched. This happens at the boundary between two sectors.
Q) How is the initial position of the rotor found?
A) For sensorless, the initial position is assumed and the technique will be to use additional current to force the motor to work. There are advanced techniques that will detect the initial position of the rotor by injecting high frequencies.
Q) How is id and iq reference values chosen?
A) The reference value for iq is the amount of torque that is requested from the speed controller. The reference value for id is typically 0, but for field weakening that value will often be negative to change the voltage constant of the motor.
Q) How can the PI controllers be tuned?
A) The PI controllers are tuned using a process called pole-zero cancellation where the PI gains are determined by the motor electrical parameters.
Q) Why do so many semiconductor vendors claim to have proprietary algorithms for BLDC motor control?
A) This is an area where they can provide value to customers, so naturally there is research done in this space.
Q) https:///wiki/Y-%CE%94_transform
A) Thanks for providing this link. It gives great examples showing these circuits are equivalent.
Q) If these sinusoidal waves are generated with a PWM signal, what would be the suggested frequency, or is it dependent on the BLDC motor itself?
A) The PWM frequency depends on your application. Usually for electric motors we can use a PWM frequency between 5kHz - 20kHz. Remember that sometimes you don't want to hear the noise of your motor (so you use a high PWM frequency) but this will affect power losses of your electronic components
Q) If you are losing too much torque using sinusoidal control, is it because it may be out of phase, or the speed too low? In this case is it then better to use Field Oriented Control?
A) In sensored BLDC control, no matter FOC or 6-step are selected, the speed loop and torque-loop can be closed since 0 RPM. In this way, no additional torque required. Regarding sensorless BLDC, there is open-loop period from 0 RPM to certain speed, At open-loop, additional torque is usually needed by additional PWM duty cycle. When the control loop is closed, the current consumption is reduced.
Q) What is the difference between a BLDC motor and PMSM?
A) The biggest difference between a BLDC and a PMSM is that BLDC motors have concentrated windings that tend to produce more trapezoidal back-EMF but they both can be controlled via the same methods.
Q) How much does the number of rotor poles impact the torque ripple on six-step control?
A) Increasing the number of rotor poles will decrease the amount of torque ripple as the electrical speed of the motor increases for the same mechanical speed and torque ripple is less pronounced at high electrical speeds.
Q) How do you find the location of the rotor using sensorless algorithm?
A) The back-EMF of the motor is estimated and can provide the rotor angle information. This is covered in more detail in the third class.
Q) If you apply three 120 degree out of phase sinusoidal current to each phase, will it start the motor? Or do we need a current boost?
A) Assuming the motor has no inerta, it should start the motor. However if the effective stator magentic field is in a position where it is very out of alignment it could cause issues with starting.
Q) What it is HFI method?
A) High frequency injection (HFI) is a method where high frequencies are injected onto the motor phases so that the subtle magnetic differences in the rotor can be determined which will help with getting a good initial guess as to the motor position.
Q) How can we determine rotor position in sensorless configuration?
A) The back-EMF of the motor is estimated and can provide the rotor angle information. This is covered in more detail in the third class.
Q) For which applications can we use sensorless control and for which can we not use it.
A) Any application that requires a large (>50%) of torque during startup or requires the motor to operate at less than approximately 10% of rated speed will be difficult to use sensorless.
Q) i need board
A): Further email communication will cover the board order information.
Q) In sinusoidal excitation, the three windings are excited simultaneously, how can then the back-EMF be measured?
A) The back-EMF can be measured based on the amount of voltage used to excite the motor windings and the current flowing through the motor. This is covered in greater detail in the third class.
Q) In six-step control, how do you know how to start commutation since when the motor is stopped, there is no back-EMF?
A) When doing sensorless six-step, you guess as to the first commutation step and start stepping through the commutation table at a slow rate to encourage the rotor to rotate with the stator.
Q) How do you start the motor using sinusoidal control, if the rotor is in between two phases and there is no Hall effect sensors?
A) In order to start, you assume a rotor angle and begin rotating in the intended direction of travel. You end up using more current than what is required but the goal is to drag the rotor around until it is spinning at a high enough speed such that you can use the back-EMF in order to estimate the rotor angle.
Q) For the motor construction, can you highlight the differences between I-PMSM and S-PMSM motor?
A) S-PMSM is the more common and feature magnets on the surface of the rotor. I-PMSM feature magnets that are interior to the rotor and require some special considerations in order to run at maximum efficiency.
Q) Do you have any efficiency differential comments of six-step vs. sinusoidal?
A) Sinusoidal control will be more efficient than six-step control.
Q) How do you ascertain the alignment of the Hall sensor placement relative to the stator pole locations?
A) The motor datasheet provides information about how the Hall sensors are aligned with the stator phases.
Q) For six-step sensorless, using zero crossing gives only four points per pole pair or two points per pole pair?
A) Six-step sensorless gives you six points per pole pair.
Q) Can you give example of update rates for FOC stages?
A) Typically, FOC algorithms should run at the PWM frequency approximately 10kHz.
Q) Is it possible to brake (slow down / stop) the motor with the control techniques discussed?
A) Yes, slowing down or braking is simply applying less torque or possibly torque in the opposite direction.
Q) Adding more poles in six-step control, will that reduce torque ripple?
A) Yes, adding motor poles will reduce torque ripple as this will increase the electrical speed of the motor for a given mechanical speed and torque ripple is reduced at faster electrical speeds.
Q) What kind of sample time is used for the feedback cycle?
A) For FOC, a sample time of 0.1ms is typical. For six-step sample times between 0.1ms and 1ms are common.
Q) Which control technique best controls motor drift when stopping the motor?
A) FOC will do a better job of stopping the motor more accurately and tracking the rotor angle if it does move while not under power.
Q) You mentioned startup when using Hall effect sensors. How do you start-up a sensorless design? How do you determine alignment if you don't have Hall position sensors?
A) The six-step startup is performed using an "align and go" technique where the stator is held at a fixed position in order to align the rotor to it. Then the commutation table is manually stepped through in order to accelerate the motor until the zero-cross detection can begin working.
Q) What type of motor has trapezoidal back-EMF?
A) Not very many motors have a true trapezoidal back-EMF, Those motors will have very concentrated windings.
Q) Are there microcontrollers with a built-in FOC block, or are there separate ICs to handle the FOC for us?
A) I'm not aware of microcontrollers that have a built-in FOC block, most microcontroller manufacturers provide an FOC algorithm that is optimized for their specific microcontroller.
Q) Does back-EMF vary with motor load or speed? How would you account for this in algorithm-based sinusoidal control?
A) Back-EMF varies with motor speed. The faster the motor spins, the more back-EMF there is. You don't need to account for this fact in sinusoidal control; it merely means you have more signal to work with.
Q) How does this translate to motors with multiple rotor pole pairs?
A) Multiple pole pairs means that the motor magnetic field will be spinning faster than the mechanical rotor, this helps improve torque ripple.
Q) With sensorless six-step control, it is difficult to sense rotor position at low RPM. Does field control also suffer from this?
A) Yes, all sensorless techniques have difficulty at low speeds. Typically anything less than about 10% of rated speed is difficult for sensorless.
Q) Are there not two different brushless motors -- AC and trapezoidal? Can you use sinusoidal with a trapezoidal motor?
A) All brushless motors are AC motors. The difference between a BLDC motor and a PMSM motor is the shape of the stator windings. You can use either control with either motor.
Q) How is field weakening done with field-oriented control?
A) Field weakening is done by making the Id reference a negative value. This will suppress the back-EMF of the motor and allow it to spin faster.
Q) For Hall-less sensoring using the stator windings, is there a lot of noise (ringing) you need to deal with?
A) There isn't too much noise when running at a speed above about 10% since that is where the back-EMF will become larger and less subject to noise.
Q) Would Id be non zero on an induction motor?
A) Yes, this is how FOC creates the rotor magnetic field in an induction motor.
Q) When you refer Space Vector is that "Space Vector PWM control"? Thanks
A) Yes. applied voltage space vector is generated by U, V, W PWM control.
Q) How do you place the Hall effect sensors accurately outside of a BLDC motor?
A) Hall sensors are inside BLDC motor. In other way, the Hall sensors was put inside BLDC motor by motor vendor.
Q) Some papers show trapezoidal commutation is good for BLDCs (no torque ripple) and sinusoidal control is good for PMSMs (again no torque ripple). So does the commutation method really depend on the type of motor (either BLDC or PMSM)?
A) BLDC and PMSM are not apple-to-apple classification. For example, many motors are BLDC PMSM. Usually, sinusoidal control has less torque ripple and trapezoidal has worse torque ripple. Smooth current means less torque ripple.
Q) The Speed Vs Current plot of sinusoidal and block commutation looks more or lessly same for some
controllers. How do you say which controller is better in such a situation?
A) Sinusoidal usually reach better performance (less torque ripple and less noise).
Q) Does a sinusoidal driven BLDC motor have a higher efficiency than a commentated BDC motor.
A) Sinusoidal control will be more power efficient as it will directly control the current flowing through the motor.
Q) What is the difference between sinusoidal PWM and space vector modulation?
A) Space vector modulation is a method to produce sinusoidal currents in a motor. So it is an implementation of sinusoidal PWM.
Q) How critical is the angle theta measurement?
A) For FOC control it is quite critical. If the value is off by more than about 5-10% it can cause a lot of efficiency loss and could lead to motor stall.
Q) Can you implement 12-step control by adding steps such as U-in, with W&V out, etc. -- as a compromise between 6-step vs. continuous control?
A) Six-steps are driven by back-EMF or Hall sensor feedback. Both feedback are at 60-degree resolution. So, only 6-step makes sense.
Q) Can induction motors be controlled in a similar manner?
A) Yes.
Q) Any comments on regenerative braking?
A). Energy storage device has to be well designed. When too much energy is stored, the storage device (e.g. DC bus capacitor or battery) may be blown up.
Q) Are there any considerations when changing direction (ie, how to offset motor momentum in the opposite direction)? Other question: what are new/key features used by MC ASIC companies to differentiate themselves?
A). It is safer to slow down to zero speed and then accelerate to the target speed at difference direction.
Q) What is vector driving of the motor?
A) Vector control is a different term for field-oriented control.
Q) How can you make a motor stall at precise external resistance/torque?
A) If your motor is running in torque control, it will stall at a precise load torque applied.
Q) Which complementary technique/measurement would you recommend to use when controlling the speed at low RPMs?
A) Sensored control, either six-step or FOC, will work well at low speeds. Between six-step and FOC, FOC will probably be better and smoother.
Q) ....and starting in the sensorless design?
A) Open-loop starting up is often used in start-up of sensorless design. (Reliable but not the most power-efficent.)
Q) Do you want the 90 degree lead angle to be smaller when under no or low-load conditions?
A) No, you want to maintain the 90 degree lead angle in FOC so that you are being as efficient as possible. To deal with no or low load situations you mealy reduce the amount of current you are providing.
Q) What's the difference between BLDC sinusoidal controlled and induction motor with sinusoidal vector control?
A) AC induction motors for a particular power are much larger than their BLDC/PMSM equivalent They are also less efficient and the FOC is more complex, which requires additional CPU resources. They work well for some applications, but in many ways BLDC/PMSM are superior.
Q) Can you use the Vector Control scheme (sinusoidal) to commutate the motor at low speeds?
A) Yes, vector (FOC) control works very well at low speeds assuming that you have the motor angle information.。