The reason for the capacitor is that it provides a phase shift on one winding with respect to the run winding in order for the motor to start rotating, just placing a single phase across a 1ph ac motor can cause it to just remain stationary due to the field oscillating across 180° , the start winding provides a phase shift WRT the start winding.
The ideal phase shift angle is 90°, but changing the winding can move the angle to some other value, not quite so efficient.
But starts, non the less.!

 

The ideal phase shift angle is 90°, but changing the winding can move the angle to some other value, not quite so efficient.

So the specified run capacitor size should never be messed with regardless of how fast the motor is spinning? If efficiency is a major goal.

 

Aye, The value is selected for a reason

 

	No Capacitor		40uF Capacitor
Tap	Watts	Power Factor	Watts	Power Factor	Watts @ 1.0 PF
Low	328	0.89	74	0.30	247
Med-Low	362	0.87	56	0.23	243
Med-High	398	0.81	44	0.18	244
High	461	0.75	34	0.14	243

Ran some more tests with a 40uF capacitor in series, and unexpectedly (to me) the fan output levels are inverse of what they were with no capacitor. The lowest tap is where the fan is blowing the most.

I noticed that if the watts figures are converted to 1.0 power factor, they all equal to 243-247 watts on all motor windings.

I'm probably observing some basic electrical engineering concepts here, but can anyone explain what I am observing without me making stupid guesses?

 

	No Capacitor		40uF Capacitor
Tap	Watts	Power Factor	Watts	Power Factor	Watts @ 1.0 PF
Low	328	0.89	74	0.30	247
Med-Low	362	0.87	56	0.23	243
Med-High	398	0.81	44	0.18	244
High	461	0.75	34	0.14	243

Ran some more tests with a 40uF capacitor in series, and unexpectedly (to me) the fan output levels are inverse of what they were with no capacitor. The lowest tap is where the fan is blowing the most.

I noticed that if the watts figures are converted to 1.0 power factor, they all equal to 243-247 watts on all motor windings.

I'm probably observing some basic electrical engineering concepts here, but can anyone explain what I am observing without me making stupid guesses?

Hi,

When you say "Low, Med-Low, Med-High, High" what are you referring to exactly? By exactly I mean the drive voltage or a position on some dial like "50 percent".
It's better to have actual numbers that indicate something more succinctly, than Low, Med, High, etc.

 

When you say "Low, Med-Low, Med-High, High" what are you referring to exactly? By exactly I mean the drive voltage or a position on some dial like "50 percent".

They are windings on a 4 speed PSC motor. Different taps. I'm using the names from the diagram on the side of the motor.

 

They are windings on a 4 speed PSC motor. Different taps. I'm using the names from the diagram on the side of the motor.

Oh I see, well then it may be hard to correlate what you are seeing to some electrical theory.

 

I decided to run some empirical tests on the difference between slowing down a motor with a Variac vs. capacitors. I set up a fan with an RPM sensor and a microphone, powered from one of those smart outlets that monitor power usage.

Nothing changed in the set up except the Variac being swapped for capacitors. I dialed in the Variac to get the same RPM as the capacitors.

Surprisingly the efficiency of both methods is identical. The Variac does show some losses in IR, the capacitors show no heating at all. I thought the capacitors would be more efficient by a few percent.

Noise level between the Variac and the capacitors is the same also. Spectrum plots are identical.

 

I decided to run some empirical tests on the difference between slowing down a motor with a Variac vs. capacitors. I set up a fan with an RPM sensor and a microphone, powered from one of those smart outlets that monitor power usage.

Nothing changed in the set up except the Variac being swapped for capacitors. I dialed in the Variac to get the same RPM as the capacitors.

Surprisingly the efficiency of both methods is identical. The Variac does show some losses in IR, the capacitors show no heating at all. I thought the capacitors would be more efficient by a few percent.

Noise level between the Variac and the capacitors is the same also. Spectrum plots are identical.



View attachment 325191

Hi,

And that only took a year to do

I guess these results are not that surprising because the device being powered should get the same waveform from both methods. If there was any significant nonlinearity present though that could change things.

 

Usually, on a fan motor that actually has taps for the different speeds, the motor is an induction motor and the different taps increase the resistance and the inductance, providing an effect similar to using the Variac to control the speed, because the slip increases as the current drops. Then eventually it was discovered that capacitors were much cheaper, and so it only took two capacitors to deliver three speeds, HI (direct), medium (two series caps in parallel), and LOW, (only one series cap.) and the caps were much cheaper, although a whole lot less dependable. Possibly acceptable in a cheap portable fan, but really shabby in an expensive ceiling fan.

 

Hello again,

One thing we didn't mention yet and that has to do with how the word "controller" is interpreted.

The title of this thread presently is, and I quote:
"What is the difference between a motor speed controller and a light bulb dimmer?"

The word "controller" has different meanings which could be just an 'adjuster' or an actual 'controller' that uses feedback from measurements.

An 'adjuster' under the context of control theory would just be a feedforward type of controller that has no feedback at all, while a true 'controller' would have part of the circuit measure the parameters to be controlled and provide feedback to an error amplifier so that the controlled parameter(s) could be adjusted automatically to meet the requirements of the adjustment.

Triac bulb dimmers almost never have any feedback. If they did, they would have to actually measure the light output and provide that measurement to the control circuit so it can make automatic adjustments.
Motor speed 'controllers' can be found that have either feedforward control or control with actual feedback, and that feedback is usually at least the speed but it could also involve rotational acceleration or even other parameters that help to maintain the speed at a constant level regardless what the load is. An example of this is with the more modern Dremel tools. The early ones had NO speed feedback and so when you dig into the work with the bit the speed slows down a lot. The more modern ones have feedback so that as the speed starts to decrease the control circuit provides more power to the motor so it does not slow down as much. That makes a big difference in the operation.

Both types of controllers could have a potentiometer for adjustment, but only the true controller can maintain the speed at a constant level as the load is changed. This is similar to a voltage regulator that keeps the output voltage constant regardless what load we apply. I guess that means we can also call the type of speed controller that uses feedback a motor speed 'regulator'. It attempts to keep the speed constant. The other simpler type just sets the voltage of the motor and allows the speed to reduce without concern with increased load.

 

Actually, the feedback in a triac light dimmer is all internal, in that when the voltage reaches the trigger point, the triac witches. And that voltage depends a bit on the load resistance at the moment. So the feedback does not relate to the light output but rather to an internal circuit condition.
I also have a number of three-terminal motor speed controllers, (that is the label on the packages), where the feedback is the user adjusting the speed control knob to achieve the desired speed. The obvious difference from a light dimmer is in the construction, as the power required by an electric motor is dependent not only by the applied voltage but also by the mechanical load applied that affects the electrical power required. THAT IS TOTALLY DIFFERENT from controlling an electrical light, where the light output is not one bit affected by the light intensity surrounding the controlled light.
Of course, a final difference between a light dimmer and a motor speed controller is the price. Almost without exception a motor speed controller will cost more.

 

Actually, the feedback in a triac light dimmer is all internal, in that when the voltage reaches the trigger point, the triac witches. And that voltage depends a bit on the load resistance at the moment. So the feedback does not relate to the light output but rather to an internal circuit condition.
I also have a number of three-terminal motor speed controllers, (that is the label on the packages), where the feedback is the user adjusting the speed control knob to achieve the desired speed. The obvious difference from a light dimmer is in the construction, as the power required by an electric motor is dependent not only by the applied voltage but also by the mechanical load applied that affects the electrical power required. THAT IS TOTALLY DIFFERENT from controlling an electrical light, where the light output is not one bit affected by the light intensity surrounding the controlled light.
Of course, a final difference between a light dimmer and a motor speed controller is the price. Almost without exception a motor speed controller will cost more.

Hi,

I agree in part, but a fully controllable system has feedback that measures the actual thing being controlled, and in this case, the light output. We can look at examples that are already in place as we speak.

Most notably maybe, some of the Samsung line of smart phones but probably others too that I just have not looked at yet. They have adaptive screen lighting that measure the ambient light and adjust the screen brightness accordingly.
This could be true for any lighting system though where you want to conserve energy. For example, a sunroof or window that lights the room partly by the sun, but as the sun light diminishes (clouds, nighttime, etc.) the electric light intensity would be increased in a linear fashion and automatically in order to keep the indoor light intensity constant over all environmental situations. This would be like outdoor lighting too that switches on or off depending on existing light intensity or a photo sensor night light.

In general though, when doing something like this the measured quantity is best taken as the thing actually being controlled. Just turning on a light bulb with a given voltage is just considered feed forward control, and that's not that great unless nothing but nothing ever changes. This includes an LED where the time of the half-life is getting close and the light output diminishes. If you want the original intensity, then you need a full feedback system that also measures the light intensity and provides that as feedback.

Another example is a motor where you want to keep the speed constant. It's not enough to simply apply say a 12vdc supply voltage to the motor because there are various problems that can cause the speed to change, and that means the control system would want to be able to compensate for that, so it's not that much different from a good lighting system. Sure, you can turn an LED on with a battery and a resistor and hope you get enough light when needed, but that's not really a control system. Also, indirect measurements do not make as good a control system as direct measurements of the controlled parameter(s). Some indirect measurements however when coupled with a direct measurement can make an even better control system (state variable control).

 

MrAL is certainly correct. Feedback is rather critical to providing accurate control of "whatever" in the vast majority of systems. So perhaps the question could have been about the difference between a light dimmed and a motor "speed adjuster", which is possibly a better description of a package with no speed or torque feedback.

As for the effect of cheap motor FAN speed controls, they do tend to reduce the power fed to the fan motor and reduce the speed rather well. One quite rainy day I was asked to "quickly fix" the rooftop exhaust blower atop the friend's building. That fan had a "controller" installed in series to adjust the speed and thus the air delivery to the required value. That module was about three times the size of the average light dimmer module, although it still fit inside a single-device electrical box.
And the device survived a whole lot of attempts to start the blower while the half horsepower 120 volt motor was stalled, making a very loud buzz each attempt.

 

And that only took a year to do

I was pretty content with how it worked. So much so that I'm thinking of building a capacitor decade box to get back ability to change speeds.

I guess these results are not that surprising because the device being powered should get the same waveform from both methods.

So I put a current clamp on the capacitor set up and hooked it up to an o-scope. What I saw was an almost triangular wave. Not a sine wave one would expect out of a Variac. I thought this would be more noisy vs. a pure sine wave, and I thought it sounded a little more noisy to my ears. But that probably was a psychological bias. The microphone disproved that theory.

 

I was pretty content with how it worked. So much so that I'm thinking of building a capacitor decade box to get back ability to change speeds.



So I put a current clamp on the capacitor set up and hooked it up to an o-scope. What I saw was an almost triangular wave. Not a sine wave one would expect out of a Variac. I thought this would be more noisy vs. a pure sine wave, and I thought it sounded a little more noisy to my ears. But that probably was a psychological bias. The microphone disproved that theory.

Hi,

I was just kidding about the "only took a year to do" it is interesting that after that long you are still working with it. That's what I call persistence and determinism. That's actually a good quality

The triangle-like wave could be because of the way transformers draw current. The excitation current looks like a sort of triangle, with the center is pointed and the other parts a little rounded.