Bridge rectifier based mains (line) frequency doubler for an induction motor

[parsec]

I want to get a larger flow rate out of a common 3 speed induction motor desk fan.

Without forking out for an off the shelf VFD, I thought perhaps I could build a frequency doubler using a bridge rectifier.

Following the circuit from left to right, I'm thinking that I'll feed the 240V mains directly into a bridge rectifier, the full wave rectified output of which will go into a large valued capacitor to block the DC component. I thought that the inductance of the windings should smooth out the sharp bits at the bottom of the waveform, but otherwise I'm open to the idea of adding another capacitor to further filter the full wave rectified waveform into something more sinusoidal looking.

The circuit is complicated by the fact that (I think) the fan has an orthogonal/start shunt winding and capacitor.

Questions:

Can you think of any safety issues with not using an isolation transformer for this circuit, or otherwise?

I intend on measuring the inductance of the windings to choose the correct capacitor values. Will these values change much under load when the rotor is drawing its mutual inductance current?

Is this a viable approach given the additional complexity of the shunt winding? I am yet to fully probe my particular fan, but I've attached a schematic of another fan. I am unsure whether my fan has this many windings, but I know it has one cap.

http://www.electrical-forensics.com/LaskoFan/LaskoFanExemplar/LaskoFanExSchematic-LG.jpg

 

[vk6kro]

If you did put a capacitor in series with the output of a bridge rectifier, you would charge the capacitor up to the peak value of the bridge rectifier output and after that, the diodes in the rectifier would be reverse biased for almost all of the incoming waveform.

So, you wouldn't get anything out.

 

[parsec]

Ah yeah, I see now.

I guess I'd need a shunt resistor before the cap to discharge it on the falling part of each half sinusoid, which would look more like a standard RC high pass filter I guess.

A shunt inductor could be a nice solution, although given the frequencies I guess it would have to be impractically large.

 

[vk6kro]

A shunt inductor would be carrying DC to ground so it could easily saturate.

I can't really suggest any simple solution to this. There are variable frequency drives, but they get expensive for big motors.

 

[parsec]

single phase VFDs capable of handling about 1KW from china aren't so expensive really ($200 or thereabouts).

Certainly from a pragmatic point of view, buying one would be the logical thing to do given the amount of time I'll have to spend playing around to get such a scheme right. I just thought it would be an interesting learning experience trying to do things the hard way.

The other options I thought of were some kind of crude inverter constructed from a single transistor phase shift oscillator, but I suspect such a circuit would be extremely sensitive to an inductive load.

Perhaps I could drive a SCR or Triac with an oscillator?

It occurred to me that a cheap household VFD for fans and other lower power induction motors could have some amount of commercial demand. I have experimented with (presumably) triac based light dimmers for reducing fan speeds with mixed results.

 

[vk6kro]

That is cheaper than I thought it would be, but still pretty expensive for a desk fan.

I also tried a lamp dimmer and mostly it worked OK for fans and electric drills, but I did have one ceiling fan that started sparking with the lamp dimmer controlling it.

I have seen pedestal fans, like this:
https://encrypted-tbn0.google.com/images?q=tbn:ANd9GcTmUoYpqGACAVEpUF7mEZnwX0aU78l6VRJrUWCEn-tQCs72OaWs

that would blow you away if you got too close, so maybe this would be a cheaper solution.

Increasing the speed of a desk fan would probably not be a good idea anyway as this would cause it to overheat and possibly burn out.

 

[parsec]

It's for a box fan that fits neatly in a cavity next to a fireplace.

Yeah it would be a bit much to pay for a fan, but on the upside I'd have a VFD for other projects/uses.

From memory these triac light dimmers chop the waveform. Electric drills are universal motors from memory so I imagine a dimmer should work okay here. I experienced a reduction in speed but my induction motor fan got quite hot when I tried this with a small desk fan.

There's some point where the fan's speed becomes load limited as it slips too much relative to the higher speed magnetic field. Anecdotally, this seems to be before it burns out. I'm not particularly precious about cheap box fans anyway, I got this one at a garage sale for $4.

 

[parsec]

It's for a box fan that fits neatly in a cavity next to a fireplace.

Yeah it would be a bit much to pay for a fan, but on the upside I'd have a VFD for other projects/uses.

From memory these triac light dimmers chop the waveform. Electric drills are universal motors from memory so I imagine a dimmer should work okay here. I experienced a reduction in speed but my induction motor fan got quite hot when I tried this with a small desk fan.

There's some point where the fan's speed becomes load limited as it slips too much relative to the higher speed magnetic field. Anecdotally, this seems to be before it burns out. I'm not particularly precious about cheap box fans anyway, I got this one at a garage sale for $4.

 

[vk6kro]

Watched the video. Very funny. I was waiting for the smoke to start.

Fans are usually shaded pole induction motors and the control is either a series connected tapped inductor or switched series capacitors to reduce the voltage available. So a dimmer should work for reducing the torque.

A fan's load depends on the speed, so reducing the torque would reduce the speed.

A dimmer uses a triac which has a rapid switch-on and this is probably where my sparks were coming from. That experiment stopped at the spark stage, so I didn't persist with smoothing capacitors.