Creating a 3 phase output from a single 12v AC or DC rail.

[Life4Never]

Hey guys, I'm new here but considering that I watch Dave with his videos, I have decided to ask the community for help.

I am currently undertaking a little side project for school. A friend and I have decided to try and create a squirrel cage induction motor from easy to find parts as a demonstration of induction motors to the Physics department as a year 12 parting gift. The whole mechanical design of the project is simple and pretty much set out already however the electronics is a bit more difficult, at least for us who have basic electronic knowledge. The school has power supplies which can supply a voltage of 12 volts in either DC or AC at 50 hertz. I planned to create the 3 phase output from this 12 volt source. I have looked on the web about phase shifting the wave by using RC circuits however this was a stretch with my knowlege in DC only and pretty basic, as well as that  the electronics was not really the point of this project.

The basic setup is 6 electromagnets which are paired up so there are 3 circuits. Each individual circuit must receive a separate AC source which are out of phase by 120 degrees. While I could use a 3 phase power input, I was also considering using a timer of some sort to time a burst of current at a set interval to try and create the illusion of a moving magnetic field however I have no idea if this would work or not.

So this is my problem, somehow creating a 3 phase power output from the power supply provided in order to make this induction motor work. As this is a demonstration project I hope it could be easily made and with  easily sourced parts from the interwebs.

I have considered whether the voltage was too low however I could always step it up using a transformer or I might as well try with 12 volts and if its too weak I could cheat a bit by placing magnets in the actual rotating section. I could also go with a 2 phase design as that would work as well but that would only be if the original plan is too out of reach.

Thanks for reading and I hope you can help me with a schematic of a possible circuit

Robert

[Psi]

Radio controlled planes use a 3phase controller to run the motors, which can be up to a 3kW, from a ~11V DC lithium battery pack.
They're basically a micro-controller and 6x N-channel mosfets to form a 3 output bridge circuit.

You can pick them up from hobby king for $30 or so
http://www.hobbyking.com/hobbyking/store/__24564__HobbyKing_70A_ESC_4A_UBEC.html
You would need something to simulate the servo feed (which controls the speed of the motor)
A servo tester should work
http://www.hobbyking.com/hobbyking/store/__8296__Turnigy_Servo_Tester.html

The only issue you might have is speed, the 3 phase maybe too fast for your motor.

The driver works something like this

[Rerouter]

if you want to make it nice and simple, get a quad inverter, with schmitt trigger inputs prefered, if not a seperate schmitt chip,

you hook up 3 inverters each with a schmitt on there input, in a loop, and use a resistor and capacitor on the output of each to make a time constant, ends up giving out a perfect if square three phase waveform to drive an induction motor (via some power element)

here is an example circuit, http://tinyurl.com/96zpgzd

other than the inverters and 10K/1uF time constant is what is effectivly a schmitt trigger just to try and not confuse you too far,

here is an ideal and very common chip to do just that http://web.mit.edu/6.115/www/datasheets/DM74LS14.pdf

7414 is a very very common chip which every supplier you look at should have

[Life4Never]

Radio controlled planes use a 3phase controller to run the motors, which can be up to a 3kW, from a ~11V DC lithium battery pack.
They're basically a micro-controller and 6x N-channel mosfets to form a 3 output bridge circuit.

You can pick them up from hobby king for $20 or so

The driver section is something like this

Interesting. I thought most RC aircraft have moved to brushless DC. Probably the larger planes then. I'm having trouble finding it on the Hobby King website. Could you provide a link? EDIT: DW Seen that you have linked it.

if you want to make it nice and simple, get a quad inverter, with schmitt trigger inputs prefered, if not a seperate schmitt chip,

you hook up 3 inverters each with a schmitt on there input, in a loop, and use a resistor and capacitor on the output of each to make a time constant, ends up giving out a perfect if square three phase waveform to drive an induction motor (via some power element)

here is an example circuit, http://tinyurl.com/96zpgzd

other than the inverters and 10K/1uF time constant is what is effectivly a schmitt trigger just to try and not confuse you too far,

here is an ideal and very common chip to do just that http://web.mit.edu/6.115/www/datasheets/DM74LS14.pdf

7414 is a very very common chip which every supplier you look at should have

Hmm. I like that example circuit. I'll look into this chip and see if it would work. Seeing as it switches from high to low, I require the circuit to go into negative voltage to reverse the current to switch the magnetic field. Wondering if it could do this.

[Psi]

na, most RC planes/helis/quadcopters are brushless 3phase AC motors.
Well. actually, it's not strictly 3phase. It has 3 phases but only drives 2 coils at any given time.
Using the 3rd for feedback.

Come to think of it that may cause issues if you actually tried to use one for your motor, but from what you've said you want to build a controller anyway.

Check out this tiny one.


And others..

[Life4Never]

How would these be wired internally?

The simple design would be

But seeing as there is only 3 wires going into the motor than where would the ground go?

would it be wired more like this?

Considering that red and black be the positve and negative and the blue wire be the control lead?

Actually I would prefer not to built the controller lol. This stuff would actually save a lot of time with the little time we have left before the HSC.

Well we can open one up and see the internals to work out how they work as a friend just happens to have one.

[Psi]

Yeah, they're pretty cheap. You could probably rip it apart and use just the fet driver section with your own controller.

However if its for a demo unit maybe it should be a proper full on AC 3phase sine wave driver rather than simplified DC square wave.

[Life4Never]

The electronics doesn't really matter. As long as it works. We are really just trying to demonstrate the changing magnetic field and how it causes the rotor to spin. Since basically all DC brushless motors are really just electronically controlled AC motors, I'm really wondering if the actual controller you mentioned above would work. It's not like we are attaching a scope to it.

[Rerouter]

Yes you can use both my control circuitry and his push pull driver to run the thing, though i found a better chip, a 40104, its the same thing but runs at up to 15V,

take a look at the schematic, that should run without an issue for you if you need simple,

edit: got my components upside down
also you shouldnt need to terminate the center it can be left floating

[caroper]

This may be overkill for your application but it does directly address what you are after.
http://www.doiserbia.nb.rs/img/doi/1451-4869/2010/1451-48691002195M.pdf


Cheers
Chris

[T4P]

Radio controlled planes use a 3phase controller to run the motors, which can be up to a 3kW, from a ~11V DC lithium battery pack.
They're basically a micro-controller and 6x N-channel mosfets to form a 3 output bridge circuit.

You can pick them up from hobby king for $20 or so

The driver section is something like this

Interesting. I thought most RC aircraft have moved to brushless DC. Probably the larger planes then. I'm having trouble finding it on the Hobby King website. Could you provide a link? EDIT: DW Seen that you have linked it.

All of them are brushless ... if it was a brushed ESC it would have only 2 wires not 3 wires
http://www.dealextreme.com/p/mystery-26500rpm-3800kv-brushless-motor-for-model-airplanes-and-diy-projects-11-1v-dc-18874?item=28
I got this the other day but it's insanely FAST

[Life4Never]

Yes you can use both my control circuitry and his push pull driver to run the thing, though i found a better chip, a 40104, its the same thing but runs at up to 15V,

take a look at the schematic, that should run without an issue for you if you need simple,

edit: got my components upside down
also you shouldnt need to terminate the center it can be left floating

Sorry still new to this stuff but how does the chip relate to the circuit diagram? Is it a simplified depiction of what's in the chip? Or are those triangles (ICs) placeholders for the chip but its just one imput and one out?

And on the RC controller, does it produce "AC" in square waves, really doesnt matter? If so does it contrinually shift the position high voltage from one to the next then to the next in a repetitive pattern because that is all I want.

Also

[Psi]

i think some of those model aircraft motor controllers are programmable and let you set the pulse speed.

However I dunno if they can be set slow enough for your demo motor though, I imaging it would needs several orders of magnitude less rotations per second than a typical RC motor.

[T4P]

Usually PWM with low RDS(on) MOSFETS
Only limited by the mosfet's Max (A)
A cheap ESC can be this : http://www.dealextreme.com/p/mystery-speed-controller-pentium-30a-for-brushless-motors-300-450-r-c-helicopters-34338?item=18

[Rerouter]

its a functional depiction, there are 6 inverters in the package, so the chip is broken up into those 6 function blocks to make the schematic more clear,


the RC controller produces a three phase square wave, as a sine would involve a more complex controller,

[Uncle Vernon]

Hey guys, I'm new here but considering that I watch Dave with his videos, I have decided to ask the community for help.

I am currently undertaking a little side project for school. A friend and I have decided to try and create a squirrel cage induction motor from easy to find parts as a demonstration of induction motors to the Physics department as a year 12 parting gift. The whole mechanical design of the project is simple and pretty much set out already however the electronics is a bit more difficult, at least for us who have basic electronic knowledge. The school has power supplies which can supply a voltage of 12 volts in either DC or AC at 50 hertz. I planned to create the 3 phase output from this 12 volt source.

.....................................

So this is my problem, somehow creating a 3 phase power output from the power supply provided in order to make this induction motor work. As this is a demonstration project I hope it could be easily made and with  easily sourced parts from the interwebs.

Given this is an educational exercise, why not apply the kiss theory and construct a rudimentary MG set.

[lab supply]-----[DC motor]==mechanical linkage==[3 phase motorgenerator]~~~~~3 phases~~~~~~~[3 phase demo motor]

You could add a direction switch to the motor, speed control of the DC etc.


What supply voltage does your 3 phase creation require?

[Life4Never]

Given this is an educational exercise, why not apply the kiss theory and construct a rudimentary MG set.

[lab supply]-----[DC motor]==mechanical linkage==[3 phase motorgenerator]~~~~~3 phases~~~~~~~[3 phase demo motor]

You could add a direction switch to the motor, speed control of the DC etc.


What supply voltage does your 3 phase creation require?

I have actually thought of this. This is probably the best way to create real 3 phase AC. Are there cheap 3 phase generators on the market? I guess I'll search ebay. Because if we created our own generator I doubt it would be very efficient at all. Maybe we could find an existing AC motor and turn it into a generator.

[Uncle Vernon]

Maybe we could find an existing AC motor and turn it into a generator.

You could be onto something.

[Life4Never]

Hmm been looking for compact size ac motor (tiny sized) and have not really found anything.

Do those RC motors work on real 3 phase power? If so, that would mean I could create a 3 phase output from it correct?

Well I think I'll focus on the mechanical aspects for now and get the thing together first.

[ejeffrey]

Keep in mind that a 3-phase induction motor will not work as a generator, at least not without modification.  Without current flowing through the windings there is no way to set up the rotating magnetic field.  You need a synchronous motor -- a small BLDC motor can probably be made to work, although it may not give a nice sine wave.

[Psi]

Brushless RC motors make terrible generators unfortunately.


You could generate your 3phase signal without too much work using..
- An Arduino with DAC module (3 channels required)
- 3x simple mono audio amplifiers (or 2 stereo) 

Just program the Arduino to generate 3 sinewaves 120deg apart and use the audio amp to get some power behind it.

The code could just be a 3x64 array of hardcoded DAC values.
Then loop through the array setting the 3 values and doing a simple delay.
You could add more elements but 64 is probably enough to get a reasonable sinewave


Or you could forget the DAC module and just find an Arduino library that can do PWM analog output and feed that into the audio amps.

[T4P]

Computer BLDC fans can be hacked for 3 phase output 

they are 3 phase but to shift the phases you need 0.1uF caps

[Psi]

hehe, if that's true then two PC fans back to back plus a few audio amps maybe all you need.

One fan running on 12V to spin the other one that's modded to generate 3phase.

[Uncle Vernon]

You need a synchronous motor -- a small BLDC motor can probably be made to work, although it may not give a nice sine wave.

Yep you've got it. Given the intent of the project I'd doubt a perfectly formed sine wave matters too much, the expected modified square wave would be quite sufficient to drive an unloaded motor.

[IanB]

Keep in mind that a 3-phase induction motor will not work as a generator, at least not without modification.  Without current flowing through the windings there is no way to set up the rotating magnetic field.  You need a synchronous motor -- a small BLDC motor can probably be made to work, although it may not give a nice sine wave.

It seems it can be done if you use the three phase motor in a combined motor-generator configuration. Here's a demo of such a setup:



Beware, there are a few advanced hacker skills on display here, not to mention a cheap throw-away oscilloscope 

[Life4Never]

Computer BLDC fans can be hacked for 3 phase output 

they are 3 phase but to shift the phases you need 0.1uF caps

I'm pretty sure that a BLDC fan is a 2 phase system since I opened one up a few days ago.

So I'm kinda wondering how to get a 3 phase out, there is a way to manipulate this or are there fans which have 3 pole system?

Or is it 4 and we just get rid of a phase and so some shifting of the waveform?

Also I need some information on how to wire up the 3 phase power. Would it connected like so:


Because right now it is like:

Where there are 6 wires comming out from our design.

[T4P]

It's exactly like that, you just have to break open many fans to look for a 3 phase ... you'll never be sure
But the PWM ones might because 3 phase is easier to control precisely the RPM

[ivan747]

Brushless RC motors make terrible generators unfortunately.


You could generate your 3phase signal without too much work using..
- An Arduino with DAC module (3 channels required)
- 3x simple mono audio amplifiers (or 2 stereo) 

Just program the Arduino to generate 3 sinewaves 120deg apart and use the audio amp to get some power behind it.

The code could just be a 3x64 array of hardcoded DAC values.
Then loop through the array setting the 3 values and doing a simple delay.
You could add more elements but 64 is probably enough to get a reasonable sinewave


Or you could forget the DAC module and just find an Arduino library that can do PWM analog output and feed that into the audio amps.

No need for a DAC, just use the "Analog output" (PWM) of the Arduino and filter it using a low pass filter. PWM is a technique used to get kind of an analog output. It switches a digital output on and off very fast. Here is a tutorial:




If you filter this PWM value you can end with a basic digital to analog converter. You can generate a sine wave with software with this "analog" output. But, to make it a real sine wave you will have to filter the fast on and off switching so you can get a real clean sine wave. You can do that with a low pass filter




Then you can that signal though the audio amplifier. Maybe if the PWM frequency is high enough you won't even need a filter because the audio amplifier will take care of that by itself.

An amplifier like this should work a treat: https://www.quasarelectronics.com/3097-2w-mono-amplifier-module-lm380.htm




Have fun!

[ivan747]

P.S. another way to move a 3 phase motor would be using RC equipment like many people has suggested before. Get a cheap ESC (electronic speed control) for a 3 phase motor, a compatible motor and a servo tester. Servos receive signals from the RC receiver the same way an ESC does, so by getting a servo tester, the ESC will "see" that it was pugged int to a receiver. You can even control the speed with the tester. You can also make your own if you want to have some fun.

[Life4Never]

P.S. another way to move a 3 phase motor would be using RC equipment like many people has suggested before. Get a cheap ESC (electronic speed control) for a 3 phase motor, a compatible motor and a servo tester. Servos receive signals from the RC receiver the same way an ESC does, so by getting a servo tester, the ESC will "see" that it was pugged int to a receiver. You can even control the speed with the tester. You can also make your own if you want to have some fun.

This seems to be the most simple way to get this project working. Since these have be obviously designed for small motors, i wonder how well, if at all, they will work for one with an approximate diameter of 15 centimetres. Though looking at the specs of these ESCs, they should output a decent enough current though I am worry if this would trip the power supply's safety to cut at 5A. I can see a problem when it is furst turnt on but when the motor is actually moving this should more be an issue.
Plus this is an induction motor and not the standard found in RC motors which contain magnets.

Time wise, we have about a month to do this, so I might have to go on a whim and just buy it and see if it works or not. If not, then onto the next idea.

[T4P]

"Small" motors
Indeed. 30A @ 11.1v ... not small
The ESC's themselves are usually limited by the mosfets, like the cheap one i know was 50A@5-26v

[Life4Never]

Computer BLDC fans can be hacked for 3 phase output 

they are 3 phase but to shift the phases you need 0.1uF caps

As a backup to the RC motor controller idea, how would you hack the actual fan? Snip off the wires and then connect add some capacitors in parallel and run the wiring off?

It would be an interesting project. I might just do it on the side as a little side project.

[T4P]

Computer BLDC fans can be hacked for 3 phase output 

they are 3 phase but to shift the phases you need 0.1uF caps

As a backup to the RC motor controller idea, how would you hack the actual fan? Snip off the wires and then connect add some capacitors in parallel and run the wiring off?

It would be an interesting project. I might just do it on the side as a little side project.

Should be it. Remember to drive with AC

[SeanB]

Strictly speaking most of the cheap fans i have looked at are 2 phase motors, with a skewing that makes them run in one direction only. The common drivers I see have a 4 wire hall and power driver integrated circuit, with only 2 phases driven, and a open collector on a transistor to give an output for RPM sensing if used.

A drive motor out of a floppy drive or a cdrom is 3 phase though, but you will need to either drive the second coil in parallel or figure out how to fake the hall feedback to it.

[Life4Never]

Okay, after some research on how ESC works ie (3 phase square wave generator) I have some more questions on the effect of it on the actual motor.

Firstly: Will an induction motor be able on this square wave? I am wondering if current would be induced in the rotor using square waves at all. Has anyone run a inductor motor such as a fan off a UPS? The cheaper ones produce square waves.

Second: How much current would the ESC draw? I am thinking of 2 solutions for power. The first is straight off a power supply, either the ones supplied or a PC ATX PSU, or running it off a LiPo and charging it before use.

Thirdly: Does the cheap ESCs have any feedback to determine how to deliver current to each of the 3 leads? Or are they just delivering square waves through 2 of the 3 wires to create the illusion of a moving field without any feedback.
 

[Psi]

My 230v desk fan runs on my square wave inverter, however it doesn't sound very happy about it.
Speed is normal though.

But i dunno what type of motor's inside it.

[T4P]

They use back EMF as their sensor on the third wire the other two wires are square waves

[Life4Never]

They use back EMF as their sensor on the third wire the other two wires are square waves

Hmm that might be a problem. Theres no magnetic field in the rotor to actually cause any back emf if the third wire is not producing a square wave to at least induce a current.

What will happen if it detects nothing on the sense?

[T4P]

the ESC stops.

[David_AVD]

I was pretty sure those brushless RC motors are 3-phase.  If they were 2-phase, how could they run backwards when required?

[T4P]

I was pretty sure those brushless RC motors are 3-phase.  If they were 2-phase, how could they run backwards when required?

Oh okay  Cheers.

[Psi]

Pretty sure the motors are 3 phase, but the ESC only drives them with 2 phases while using the 3rd for monitoring.

It's not static though, it switches around in all combinations.
(If you looked at it over many cycles you would see all motor wires used for both power and monitoring at different times)

This is how it can control direction as well as speed.

[Life4Never]

So judging by the information so far, an ESC won't work on an INDUCTION motor. It might work for BLDC motors but thats because it has magnets inside which can set up back emf.
There will be no back emf as it is an induction motor (no magnets). <--- There should be back emf on the 2 powered wires as we induce a current and then that can create back emf but then on the sense, I'm assuming its unpowered, there will be no back emf and therefore the ESC wont work.
Is this correct so far?

Or, does the "sense" wire generate an electrical pulse and measures back emf from that? If so then it might work.

[Life4Never]

Anyone?
Answer the above questions as I'm planning to buy the parts required this week.

If not then can anyone suggest another simple method with easily sourced parts in the local Sydney area? (I don't plan to do any programming. Out of the box, almost no setup required like the RC ESC and controller)

If all else fails, I will create a generator connected to an LED or bulb to demonstrate the generation of electricity by changing magnetic fields. I will need to buy some flat neodymium magnets though. A local source on the cheap?

[Panduit]

Computer BLDC fans can be hacked for 3 phase output 

they are 3 phase but to shift the phases you need 0.1uF caps

I'm pretty sure that a BLDC fan is a 2 phase system since I opened one up a few days ago.

Oh no


Ic driver NJM2624 3phase controller

[T4P]

Holy lord, hmm, i wonder what's inside a PFR1208XHE now