I'm investigating HVAC fans etc to design my own system and controls, and having fully variable speed control is an important consideration to me.

I'm pretty sure there are types of AC motors that cannot be easily speed controlled with a triac or voltage control (despite the indications in the answer at https://electronics.stackexchange.com/a/2900/7517)

I feel reasonably comfortable with controlling DC motors with voltage control, but suspect this is a simplistic mis-conclusion.

Other than products that explicitly state that "Smooth or stepped speed control can be performed by external thymistor or transformer.", how can I tell if I can use these general techniques to control speed for AC and DC motors? (For extra points/comments: am I correct in assuming they're talking about a thryistor?)

Am I correct that less than 100% of AC and DC (as separate categories) are able to be speed controlled?

Bonus sub-question (for votes rather than answer): in the quote above when they say thyristor, do they really mean a TRIAC? I can't see that a uni-directional current switch would be at all useful. Unless of course the fan could accept rectified ac ... or something?

  • \$\begingroup\$ A triac is a type of thyristor. \$\endgroup\$ – Olin Lathrop Jan 15 '12 at 15:22
  • \$\begingroup\$ @Olin yes, I realise a triac is a type of thyristor. I'm just checking that a triac is the only type of thyristor that would be useful for this. i.e. I can't see why you'd only want to trigger on positive voltages on an ac waveform instead of positive and negative. \$\endgroup\$ – Stephen Jan 15 '12 at 22:10

The following answer is a bit long since I'm trying to cover motor speed control in general; however, I'll first briefly answer your questions in a correct, but perhaps not useful manner:

  • Yes. There are motors that cannot be easily controlled via a triac or voltage control methods.

  • You are correct that you may be over-simplifying.

  • No, with the proper control circuit (and controller function), any motor can be speed controlled.

  • I don't know what the speed controller you linked to is using for their thyristor. It could be a triac, it could be two SCR's connected in anti-parallel (basically a triac). It could be a single SCR (but not likely). I'm not sure that this distinction matter however...

  • Finally, for an example of how a unidirectional SCR could implement "high quality" speed control, see this discussion on how an ASCI operates. State of the art 1960's style...

I don't know what you want to control but would suggest that you consider something like this as an example of an integrated motor+drive solution or this for a separate component drive for an existing ac induction motor. Either example is high end and may be too expensive for your application, especially if this is a one-off situation. Lower cost alternatives are available, however...

(full disclosure: many years ago I worked for what was then GE Motors and my co-workers developed the ECM I linked to. Also, in my present job, we use similar ABB drives...)

Onto motor speed control in general!

Speed control can be achieved with two different techniques:

  • direct speed control of the motor where the control circuit forces the motor to turn at a specified speed (within the limitations of the control circuit, motor, etc.). Examples include variable dc voltage to a dc motor or a variable speed drive for an ac induction motor
  • indirect speed control of the motor where the combination of the control circuit action, the motor's design and the motor's mechanical load determine the operating point speed. Examples include torque modulation or slip control.

In some cases, e.g. a common dc motor, the speed controller can be as simple as a variable resistor or a variable dc voltage source. For other motors, the control circuit will be more complicated. For a few motors (e.g. stepper, switch reluctance or brushless dc), the "complicated" control circuit is required to just get the motor to turn. Since this circuit is already present in this case, speed control is almost a trivial after thought (provided that the controller designer chose to make this capability available to the user).

If you want a simple DIY motor speed control, then a dc motor with a permanent magnet field is about as simple as it gets. Provided the motor isn't overloaded (or supply current limited), the speed will track the applied voltage. Vary the terminal voltage as you wish (resistor, variable dc supply, power amplifier, etc.). Easy-peasy lemon squeezy.

Wound field dc motors generate their magnetic field via a winding and this field current will need to be controlled. For low power applications one might fix the field current and then control the supply voltage; however, in larger power motors controlling both the field current and possibly the armature voltage gives additional dynamic control performance and more opportunities for excitement: Never turn off the field current of an operating wound field dc motor! (hint, it will accelerate until turned off, the speed reaches the steady operating point determined by the residual magnetic field or a mechanical over-speed failure occurs...) You will likely not encounter wound field dc motors in any contemporary HVAC application.

All motors (ac or dc) can be indirectly speed controlled via torque modulation. Basically, this control method applies full torque for a period of time (motor is "on") and the zero torque for a period of time (motor is "off"). When the motor is on, it accelerates towards its full speed. When the motor is off, it decelerates ("coasts") towards zero speed. The rotating inertia of the motor and load provide the low pass filtering which results in an average speed. Triac-based motor controllers typically use this method: They turn the ac motor on for a number of line cycles and then allow it to be off for a number of line cycles.

Virtually all motors can be speed adjustable via pole changing switches. These switches effectively re-wire a specially designed motor to re-configure its rated or normal speed in discrete steps. This is typically only used for ac induction motors, but one could do this for most other motor types. The speed of the rotor depends on the number of electrical pole-pairs in a mechanical revolution. Increasing the number of poles results in a slower spinning motor. Multi-speed household fans are a common example of this control method.

A common property of ac motors is that they have a rotating stator magnetic field that the rotor follows. The faster the rotating field, the faster the no-load speed. Grossly over simplifying, this is how direct speed control of ac motors is achieved.

AC motors can be grouped into three general buckets: induction, synchronous and what I'll call "switched". Induction and synchronous motors both use sinusoidal currents to generate their rotating stator magnetic field. Switched motors use non-sinusoidal currents that more resemble square waves. Examples of switch motors include stepper motors, brushless-dc and switched reluctance machines.

Induction machines have an operating property that is called "slip." This means that if the machine's torque is non-zero, then there is a difference in speed between the rotor and the rotating stator magnetic speed. This normalized difference is called the slip. The slip and the torque are related such that increasing the load torque increases the slip. Depending on the load's torque-speed characteristic, this provides a basis for indirect speed control in an induction motor. Changes that affect the motor torque production will result in changes in slip based on the load's torque-speed characteristic. The change in slip means a change in speed. The torque can be adjusted by adding resistance in the rotor circuit if the motor is a wound-rotor. Changes in the stator circuit will also affect the torque production but require a specially designed motor to prevent the motor from being damaged. An ac motor has a current minimum when the supply voltage and frequency are a fixed ratio (V/Hz). Performing speed control by adjusting the stator circuit excitation will result in larger stator currents. This is acceptable if the motor is designed for this mode of operation; however, the needed margin to accommodate this results in a special purpose and typically more expensive motor. Motors not specifically designed for this operation will be damaged if operated in this mode for very long. Note that controlling speed via the stator circuit's resistance or voltage will also restrict the load's torque-speed characteristic to behave similar to a fan.

Synchronous and switched motors do not have slip and so if their rotor is not always aligned with the stator's magnetic field, then they are said to be "slipping poles" and will not rotate. This means that these motors can only have direct variable speed control if their stator excitation frequency varies. This usually means some form of electronic motor drive. (You can super old-school it for non-switched motor by using a cascade of induction and dc motors and dc and synchronous generators).

Because induction and synchronous machines have sinusoidal stator excitation, their electronic motor drives will be very similar. In low performance systems (e.g. open loop V/Hz), one could use the same drive & controls; however, for normal or higher performance, there will be differences in the control topology to address the differences between the two motors.

Electronic motor drives for switched motors are different since they produce currents that approach (or are) square waves. Often, the drive's output frequency is set by the speed command and the drive adjusts the torque (via current amplitude or phase angle) to keep the rotor aligned with rotating stator magnetic field.

The brush-less dc motor is so called because the stator excitation is based on the rotor position. The combination of the motor and inverter circuit and control operation results in a transfer function from converter dc bus to motor torque/speed that is identical in form to a dc motor.

  • \$\begingroup\$ Well, I may not have read and/or understood it all, but I shall declare this correct! :) And upvoted! \$\endgroup\$ – Stephen Apr 23 '12 at 11:20
  • \$\begingroup\$ Thanks. If you've got questions that are more application specific, I'll be happy to provide a more targeted (and shorter) answer... :) \$\endgroup\$ – madrivereric Apr 25 '12 at 20:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.