But when comparing it to an incandescent light dimmer, that's decades old, they wave form out of both devices looks identical. That's the part that does not make sense to me. I did not see any spikes or noise on the dimmer.

"they just don't make them like they used to" probably applies here. Your old school dimmer might have suppression components that are considered superfluous by today's industry standard. I would be curious to see how a 2023 cheapo dimmer from Home Depot stacks up against these two devices you're comparing.

 

I have a need to control the speed of a number of different fans, and find all solid state solutions quite noisy. I have several Variacs in locations where fan noise should be kept to a minimum, but in others I have solid state, as Variacs aren't that cheap.

I have this specialized motor speed control.
https://www.greenheck.com/shop/accessories/speed-controls/385031

It is much nosier compared to a Variac. So I decided to compare this Greenheck to a 1980's Lutron incandescent light dimmer that I had in the parts bin. Put a clamp on the output of both and looked at the output on an o-scope. Identical! There was no difference in the wave whatsoever.

So the question is, is there really such a thing as a purpose built solid state motor speed controller? Or are they basically just a light bulb dimmer in a different box?

Hello there,

I had designed several motor speed controllers and the simplest type is the one with the triac and phase angle control, which how most lamp dimmers are made. There are other types which are much, much better, and will act just like a variac without a variac.

There are differences in the designs of the triac types too. Some will go down to zero and up to nearly 100 percent, while others, the cheaper older types, will only go maybe 20 percent up to 100 percent and that is because they use a diac which does not allow control down to zero volts. The type that use an actual control circuit can go all the say down to zero because they can control the phase angle down that low. The control circuit is not all that complicated either, and today you can do it with a microcontroller and a triac. Before that it was not that much more complicated either using an LM358 for example and some resistors and capacitors, and the high current triac. One of the higher power ones i made controlled an angle grinder and used a 15 amp triac. The control circuit was the same for one using a 2 amp triac, which was mainly an LM358 used as a ramp generator and a comparator.

The better kinds will use a transistor bridge and output something like a sine wave or a pure sine wave. For important applications you might see a pure sine output, but for less than that important you might see a pseudo sine wave that is really just a bunch of pulses with average value of that of a sine wave. The interesting thing is that these do not usually need filtering for motor applications. The audio noise is less than with a triac because the switching frequency will be much higher so the current changes are smaller. With a triac, there is no good way to turn it off so once you turn it on you can't really turn it off. This means the frequency is limited to the line frequency which can be either 50Hz, 60Hz, or 400Hz, and most typically 50 or 60Hz. These are low frequencies and cause a lot of audio noise even in some regular transformers. Going to a transistor bridge circuit allows pulsing at 1kHz or even higher, which means the change in current is much less and the change in current is what causes the noise.

Did you happen to mention yet what the power level is for these fans? Is it 10 watts, 50 watts, 100 watts, 200 watts, etc. ? These power levels are not that high so i would think the circuit would be mediocre when compared to high power converter circuits. This would keep the price down. If you go to buy one though it may be more costly if there has not been many units sold over a year for example. You may be able to build one, but you'd have to have some experience building these things or at least some electronics experience.
The idea is kind of simple. You rectify the line voltage and filter it, then chop up the DC with a transistor bridge and that becomes the output. The chopping comes from a circuit that has a pattern stored inside that tells the transistors when to turn on and off. As you change the output level the pattern pulses get wider or narrower and that is what adjusts the speed.
You can add motional feedback if you like but that requires a sensor to sense the rotational speed, and is probably not needed for a fan speed control because the load friction is almost always constant.

 

The better kinds will use a transistor bridge and output something like a sine wave or a pure sine wave.

That's how a lot of modern (and cheap, too) electric bicycle motor controllers work. The 3 phases going to the motor are a sine wave simulation at around 16khz.

The A/C motors I have are 1/2hp, 3/4hp, 1hp, and I turn them down quite a bit, down to 20-30%, so the noise from solid state speed control is very pronounced.

Are you aware of any commercially available speed controllers that do sine wave simulation and are not very expensive? For expensive I can just get more variacs.

 

I haven't tried them on fans, but have you considered designing around a zero crossing driver such as the Fairchild MOC316x MOC308x series.?
See their AN-3006 app note.
Probably a simple PWM signal in to the controller.

 

That's how a lot of modern (and cheap, too) electric bicycle motor controllers work. The 3 phases going to the motor are a sine wave simulation at around 16khz.

The A/C motors I have are 1/2hp, 3/4hp, 1hp, and I turn them down quite a bit, down to 20-30%, so the noise from solid state speed control is very pronounced.

Are you aware of any commercially available speed controllers that do sine wave simulation and are not very expensive? For expensive I can just get more variacs.

Hi,

If i had the choice i would go with variacs as they are very reliable.

The other way is to use a variable frequency drive, but they run from about $60 and up to hundreds USD.

Since the main issue is the motor gets banged with a current, a cheap fix may be to try placing a small value resistor in series with the motor. This will limit the current rate of rise and therefore should limit the audio noise.
There will be a tradeoff in selecting a resistor value however. The smaller the resistor value the higher the efficiency stays, and the larger the resistor value the less audio noise. So the tradeoff is efficiency vs audio noise.
You would really have to try this though with some power resistors.
Since you only require about 30 percent of max speed you would be able to get away with a relatively large value resistor but it would eat up a lot of power unfortunately, and would have to be mounted in a box that allows a good air flow. The power rating of the resistor should be twice i^2*R to keep the temperature down, and even a higher power resistor you can. The larger the surface area the cooler it stays.
Keep in mind this would be an experiment, but should not cost much to try for one motor first.

The correct way to lower the speed of an AC motor is to lower the voltage and lower the frequency. There is a third power relationship if i remember right. As the frequency gets lower the motor speed decreases, and that allows a lower sine voltage to be applied. I can't remember if that is to keep the efficiency high or just keep the current lower. This however requires a VFD to get right and hopefully one that has the power relationship built in.

Since variacs seem to work for you though, you could look on Amazon for some cheaper variacs. They start at around $70 USD but not sure where you live.

It is also interesting that there are a lot of tutorials on the web for building a bridge type motor controller but i am not sure if that covers AC motors too or not.

 

If you do not need continuously variable fan speed then an auto-transformer with a few taps, and a tap switch, can provide control in a few steps. Many fan controls are HI-Medium-Low, which is adequate for most situations. Much cheaper than Variacs and more reliable and repeatable.

 

Very good information all around. Thank you.

On related subject, I understand that in series capacitance can slow down a PSC motor. I've seen some room fan, 30-40 watts or so, have a speed switch that selects between different capacitors.

What is a good reading to undestand the math behind this?

 

Look-up .........
Inductive-Reactance vs Capacitive-Reactance.
.
.
.

 

Very good information all around. Thank you.

On related subject, I understand that in series capacitance can slow down a PSC motor. I've seen some room fan, 30-40 watts or so, have a speed switch that selects between different capacitors.

What is a good reading to undestand the math behind this?

Are you sure that wasn't achieved by switching in different main windings?
The start winding and cap usually remain untouched.

 

Toshiba was an early leader in 3 phase VFD drive technology. They could do things such as maximum torque on startup with minimum RPMs. Chopping 3 triacs with their proprietary microcontroller algorithms they did far more than just variable speed control. And far more than any rheostat could do. Imagine stopping the motor of a fully loaded greatly inclined belt so that you had zero motor rpm but maximum torque preventing the load on the belt from reversing the belt direction and dumping its load all over the place. It's been too many years to remember all the details but it was impressive what they could do 40 years ago. Many startups tried to follow their lead and failed to perform as well.

 

Very good information all around. Thank you.

On related subject, I understand that in series capacitance can slow down a PSC motor. I've seen some room fan, 30-40 watts or so, have a speed switch that selects between different capacitors.

What is a good reading to undestand the math behind this?

Hi,

Capacitors are sometimes placed in series with some load, and that reduces the voltage that gets to the load and thus the current and so the speed would be reduced if it was a motor. However, due to the larger size of the motors you seem to be dealing with, the capacitors would have to be quite large, at least for AC capacitors. For example, if you wanted to reduce the voltage of a 1HP motor running at 120v to just 60v, you would need something like a 300uf AC capacitor rated for more than the full voltage. Not sure what that would cost, and it would not be variable unless you got some difference values like maybe 300uf, 200uf, 100uf, something like that, although we'd have to know more in order to calculate a closer value.
This means it is probably not a good idea, but it depends on what really want to do also.

The math behind this is not too difficult but it would require you to learn complex number math. Complex number math has two values rather than just one like some other common things. When you got to buy apples you think about maybe 50 cents each, or a dollar each, and those are just single numbers. Complex numbers are not that much different but you always have two parts: a 'real' part and an 'imaginary' part. They would be shown in different ways, as a couple like (1,2) or as an addition with an imaginary operator 'j' like: 4+7j.
For a capacitor, the real part is zero and the imaginary part is -1/(w*C) where w=2*pi*f where f is the frequency in Hertz.
To get the current now, you just add the load (assuming mostly resistive) to that and calculate the 'norm'. For example, if your load was 20 Ohms and the capacitor was 200uf, then the math would be just:
Zmag=sqrt(20^2+1/(w*0.000200)^2)
where Zmag is the magnitude of the impedance.
With f=60 Hertz, we would have w=377 so this would turn into:
sqrt(400+1/(377*0.000200)^2)=24
so the current would be:
120/24=5 amps.
Without the series cap the current would be 120/20=6 amps, so we reduced the current a little.
With a smaller cap value we would reduce the current more. With C=100uf we would get about 3.6 amps, which is now a lot less than 6.
With C=50uf we would see about 2.1 amps, which now is just one-third of the original current of 6 amps.

So you see how this works. Here is a simple formula you can use for now:
Zmag=sqrt(R^2+1/(w^2*C^2))
and the current is:
i=Vrms/Zmag
where Vrms is the RMS line voltage, typically 120vrms.

If you do keep the motors at a constant speed you can measure the current and use that to figure out the capacitor value.
Note however that if the line voltage sags for a while then the motor speed will decrease, and if it goes up then the speed will go up. There is no real regulation.

 

Are you sure that wasn't achieved by switching in different main windings?
The start winding and cap usually remain untouched.

I have repaired several cheap pedestal fans, and I have worked on a lot of them. Those fans do use a series capacitor and it is switched around.
These fan motor systems are not to be compared to better quality capacitor motors that have a capacitor as part of the system. Cheap junk is totally different, Max.

 

I have repaired several cheap pedestal fans, and I have worked on a lot of them. Those fans do use a series capacitor and it is switched around.
These fan motor systems are not to be compared to better quality capacitor motors that have a capacitor as part of the system. Cheap junk is totally different, Max.

Hi,

Oh that's interesting, the ones I worked on had a motor start capacitor and used different windings to get different speeds.

 

Hi,

Oh that's interesting, the ones I worked on had a motor start capacitor and used different windings to get different speeds.

Some were made that way, with others it is far less obvious. But in every case it is clear that making the cost to produce the item is by far the major goal.

 

Hi,

Oh that's interesting, the ones I worked on had a motor start capacitor and used different windings to get different speeds.

The ones I have mainly come across do not touch the start winding, which makes sense, as that should not change.
Just different main windings are selected.

 

Some were made that way, with others it is far less obvious. But in every case it is clear that making the cost to produce the item is by far the major goal.

Hi,

Yes i have to agree with that, and i have seen some really cheap ones. I think they all have to use thermal fuses now though.

 

The ones I have mainly come across do not touch the start winding, which makes sense, as that should not change.
Just different main windings are selected.

Hi,

Yes, that is what i usually see and can't remember seeing one any different now that i think about it.
I've had to replace thermal fuses and start capacitors, and one had a bad winding and i fixed it and then it went bad again and the surface winding wasn't the problem that time, it must have been a place in the winding deeper in.

 

NONE of the cheap fans that I have repaired or tried to repair had what I would call a start winding, although they may have had separate phased windings. They were all rather "high slip" induction motors.

 

Most likely a shaded pole fan motor.
VERY inefficient motor.

 

Most likely a shaded pole fan motor.
VERY inefficient motor.

Certainly efficiency is not one of the top 20 requirements given to those who design these motors for this type of product. And with plenty of airflow heating is seldom an issue, until the inferior grade lubricant fails to lubricate after the warranty runs out.
Many of these fan motors are returned to operation just by adding a few drops of motor oil.