Here's a real doozy for you engineer types...

[laserhawk64]

H'okay. Pardon the long post here... it's a doozy, as labeled, and thusly has a doozy of a post to go with it. Also, long explanation of what I'm trying to talk about (laying the groundwork before the question, if you will) is NOT an excuse for tl;dr. Further, I've spent *at least* the past four hours trying to find the answer to this on Google, and my efforts so far have been fruitless. (That said, I readily admit that my Google-Fu is next-to-worthless.) All that said...

If you take apart a PC fan motor, you will find four coils, a rotating magnet ring, and a bit of utterly worthless (for my purposes) circuitry. The research I've done on the World Wide Waste O' My Time seems to suggest that this is actually a "two phase alternating current" brushless motor, with the otherwise-useless circuitry doing translation between the motor proper and the rest of the computer.

More info about brushless DC motors and how I came to that conclusion here. (Warning: this stuff is a mess. It's *very* easy to get a severe case of tl;dr from that page.)

So crazy ol' me starts a-thinkin'... those motors are cheap as cr*p and easy to build, and if you spun them around, you could probably generate a lot of electricity [small-scale proof that this idea actually works] -- or at least enough to do *something* useful, like run my little ASUS netbook for a couple hours a day...

Here's the doozy. That motor, working as an alternator (generator = dc, alternator = ac), has an output of two-phase alternating current. Standard wall current here in the United States is what's properly called split-phase alternating current.

Hence, my question: how the @%$#$!! does one convert two-phase alternating current to split-phase alternating current, in order to make it useful?

 

[PottedMeat]

Hence, my question: how the @%$#$!! does one convert two-phase alternating current to split-phase alternating current, in order to make it useful?

Generally n-phase at x frequency is converted to m-phase at y frequency with an inverter.

Basically you turn AC to DC to AC. In the DC portion of the circuit you can switch the direction of the current at specific times to output AC at whatever phases and frequency you desire.

I believe there are mechanical ways of doing this too, like a rotary phase converter ( converting single to three phase ).

 

[laserhawk64]

That's a good start, but I'm more looking for a method than just a mathematical formula (math isn't my strong point anyways). There's got to be a way to build one of these things, and there's also got to be a project page explaining how it works, and how the person who built it made it work.

I can understand build write-ups. Most of math is well beyond me

 

[ModestGamer]

You could buy a automotive alternator from a motor vehicle and rip the diode bridge out of it. Whala 3 phase. down converting is easy from there and getting the proper HZ is as easy as setting the speed correctly. You must however energize the field to make current and this could be done with a 9 volt battery. Once the field lights just rectify the output down with a few diodes back to the field coil and put a few resistors inline and you can also control the output if you clever enough to get 120v 60hz power. Makes a great back up generator for a refrigerator if your smart enough to get it working. You could conversly change the winding to hit your targets as well. .

BTW very easy to get from 3 phase to what your looking for.

 

[laserhawk64]

:sneaky: You're getting closer to my overall idea... but not to sound TOO much like a mad scientist... I'm not quite ready to say what my final goal is here. Suffice to say that an automotive alternator won't do what I'm looking for in this particular case Needs to move too fast, and it's too much of a jury-rig.

That said, given that three-phase to split-phase is indeed an easy conversion... is there a way to make a similar motor-like generator that works as three phase? Perhaps by using six coils instead of four...?

 

[PowerEngineer]

Well, it's true that pretty much any motor can be converted into a generator by spinning its shaft using some "prime mover". How much electricity you get out depends on the power of the "prime mover" and the efficency the converted motor.

My understanding is that the fan motor uses an IC that includes a hall effect sensor to determine the rotor position and some sort of switching transistors to apply the DC voltage to the coils in a way applies torque to the rotor (through the interaction of the coils' magnetic fields with the magnetic field of the permanent magnet in the rotor). You're right that this circuitry needs to be removed for a conversion to a generator.

What you're left with are just four coils across which voltage will be induced by the changing magnetic fields you cause by spinning the rotor. The amount of voltage (and current) that each coil can produce will be a function of the speed of the rotor, the stregth of the rotor magnet, and the "leakage" of magnetic flux (i.e. field from the magnet that do not link with the coils). Also note that the voltages induced in the coils will be alternating (i.e. AC) and that the frequency will be dependent on the rotor speed, and (since the four coils are all 90 degrees apart) the voltages on the four coils will be out of phase by 90 degrees with each other.

In the "small-scale proof" you cited, he uses four diodes to rectify the AC outputs from the coils into a (roughly) DC output. As PottedMeat suggested, this is usually the best way to convert the power from a variable frequency source into something usable.

In fact, the newer designs for those big wind turbines do the same thing. Their variable frequency output is rectified to DC and then inverted back to AC at 60 Hz. Not being locked into an AC-to-AC link to the 60 Hz power system gives the wind turbines a much more flexible range of operation.

Note that the output of the "small-scale proof" was only 20 milliamps at 2 volts (0.04 watts!). You'd need at least 50 of these to light a flashlight.

 

[laserhawk64]

OK... based on what I'm hearing (please PLEASE correct me if I'm wrong) I need a 6-coil version of the fan motor to make 3phase AC. Then, my device works like this:

The alternator spins, and (duh) converts mechanical energy to electrical energy.This electrical energy (3-Phase AC) gets converted to split-phase AC. It then passes through a full-wave rectifier and then a voltage regulator, to a charge controller. The charge controller runs a battery bank AND has a built-in inverter (this is a common combination, I believe) which provides power out to... well, honestly, whatever I want :twisted:

Mechanical Energy Input --> 3ph AC --> Split-Phase AC --> DC --> Storage --> Split-Phase AC Output

EDIT: it'd probably be good to note that unlike the small-scale proof that I mentioned, I'm going to be looking at a somewhat larger contraption that's a LOT more capable.

 

[PowerEngineer]

The number of phases on the generator side will make no difference at all (other than to the choppiness of the voltage) if you plan to rectify the output to DC. As I stated above, the four coils in the fan motor can be split out so that each provides an AC wave -- effectively four-phase power. By connecting them through rectifying diodes, you get a DC-ish output that is the highest voltage on any of the four coils at that moment in time.

"Split-phase" AC is just single-phase AC. Center tapping that phase to give you both 120 volt and 240 volt service is not something to get hung up on.

Not sure what scale of power production you are envisioning, but I'd try to design it to provide DC at 12-14 VDC so that I could feed the power into a standard car battery and then use automobile DC-to-AC converters to drive your AC devices.

Prime Mover --> Multi-Phase AC --> 14 VDC --> Car Battery --> 120 VAC

Proceed carefully...

 

[laserhawk64]

Thank you, PowerEngineer... I think I understand this well enough to get going. You're a really helpful and smart person, and I appreciate your assistance very much.

Now if someone could kindly call the local thread-closing service...