Electronic – Calculating the capacitor values to control ceiling fan speed

A variac works just fine. I have a medium sized AC fan hooked directly up to a small, 1A variac which I made years ago as a quick and dirty solder fume extractor, but it worked so well that quick and dirty turned into 'what I've used for years'. The speed control works exactly like you'd expect.

What the amazon seller says is correct. You're also correct. The key difference is that virtually all AC fan motors are a special kind of induction motor called a shaded pole induction motor. These motors do not need any external capacitors to introduce phase lag, and instead rely on copper rings around each of the poles that act as single turn transformers, introducing a delay in the magnetic flux for that pole, and the net effect is a second phase.

This makes for a fairly inefficient, but extremely reliable motor. Due to their construction, they generate torque over their entire RPM range, including stopped, and their torque is mostly dependent on voltage, except for some nonlinearities at very slow speeds where reduced torque is generated.

You can see this in most fans - you turn them on but they take a while to get up to speed. The load from the fan blades and the air they move usually determines the final speed, which is whatever speed where the generated torque and the load torque match.

Since the load on a fan increases with rotation speed, this means that simply by varying the voltage, and thus the torque, this will also vary the speed the fan will spin, as less torque generated means the fan will reach that matching load torque at a lower speed.

Also, just to be clear, there is no magic going on with the TRIAC driver, it's nothing more than a dimmer circuit. It's chopping up the AC waveform to generate a lower Vrms. It's just a very noisy and harmonic filled way of reducing the average voltage, so using a TRIAC driver isn't any different from using a variac.

The variac would be much less noisy, and not add any harmonics, and ignoring cost, size, and weight, would be the preferred solution. The primary advantage of TRIAC drivers are cost. Either will work, but a small variac would be the more elegant and low-noise solution.

Have a look at this article.

^{simulate this circuit – Schematic created using CircuitLab}

Figure 1. 3-speed fan control.

It appears that the standard speed control uses additional capacitors in series with the windings. By switching in one or two capacitors three speeds can be obtained.

Table 1. Speed selection.

| SW1 | SW2 | Speed |
+-----+-----+-------+
| Off | Off | Off   |
| On  | On  | High  |
| On  | Off | Med   |
| Off | On  | Low   |
+-----+-----+-------+

Table 1 shows that switching in all the capacitance gives highest speed. What's not so obvious is that it also gives highest starting torque. Most ceiling fans use a pull-cord switch which cycles through the sequence above. The switch deliberately goes from off to max so that the fan will start reliably. The user can then slow down the fan by one or two further pulls.

The capacitor values will depend on your supply voltage and the rating of the fan motor. They need to be mains voltage AC rated.