Homebuilt handmade "3 phase" electric motor

[skimask]

This is something I've been wanting to build since I was a kid...(and I really wish I had some pictures to post)...need some help figuring out where I went wrong. I got most of my info and research from Wikipedia, so we'll see how far that gets me...

I skipped right over the regular science fair project DC motor with a couple of nails for electromagnets and went for a '3 phase' (PIC controlled of course) synchronous motor...that's probably problem number one right there...

The rotor is an 10 inch piece of carbon-fiber/kevlar arrow shaft with a piece of 6 inch long, 3 inch diameter white PVC pipe with center drilled end caps mounted on it (the arrow shaft goes thru the middle of it and it's centered very well). I've got 2 pieces of 8 ga. solid copper wire around both end caps (a bit tough to bend) with 36 pieces of 8 ga. solid copper wire connecting both ends, equally spaced (or at least as close as I could get it), and each piece of wire connecting each end is soldered to the rings at both ends. And I've got all of that secured with a good coating of clear silicone and a few zip ties.

9 total field coil windings (3 each for each of the 3 phases), each one is a steel bar (6 inch long, 1/4" x 1/2" wide/thick), with 36 turns (12 turns, 3 layers deep) of 20 ga wire, equally spaced at 40 degree intervals around the rotor. The field coil windings are oriented so the thin edge of the bar (without the wires) is facing the rotor.

The main driver is a PIC running multiple H-bridge type setups using MOSFETs and driving the field coils with +/- 24v, current limited to about 3 amps per phase. Basically, each phase is being controlled like it's driving a reversible DC motor. Each phase steps in a sine wave like fashion, +24v, 0v, -24v, 0v, +24v, etc.

The controller part works just fine. I've got it set up so I can control and monitor RPM/frequency with my 'scope and I've got red/green LEDs on each phase and coil and I can watch them sequence correctly. Everything is isolated from everything else. The PIC controller runs off a NiMH pack, feeding optoisolators, which trigger the power MOSFETs, which run the coils from their own separate lead acid battery pack. I don't see any spikes at the PIC on the 'scope when the system is running.

But the motor doesn't spin!
I've got as tight of a gap as I can get between the rotor wires and the field coils, roughly less than 1/16".
The bearings that the arrow shaft rides on are nice and smooth and practically frictionless.
The rotor seems well balanced, I can't get it to stop at any single spot repeatedly during spin tests (both horizontally and vertically (even though I don't have any sort of thrust bearing on it).
As 'wikipedia' mentioned, I offset the ends of the rings on the rotor by about 10 degrees (supposed to help start it) making it look like the interconnecting wires are on at an angle.
I even tried turning the field coil windings/bars 90 degrees so that the wires were facing the rotor instead of the blank steel end, had to increase the air gap a bit, about 1/4", so that could've killed that idea.

The idea, according to wikipedia, is that the field coils induce current into the copper wiring on the rotor to set up a counterbalancing magnetic force which either attracts or repels the field coils on either side of it.

Any ideas where I went wrong?
( I mean besides the obvious skipping over the beginner's DC motor part... :D )

[skimask]

This is something I've been wanting to build since I was a kid...(and I really wish I had some pictures to post)...need some help figuring out where I went wrong.

Tried a couple of things today...
-Painted a black stripe on one end of the pipe, glued a piece of foil for a reflector, added a light source and a phototransistor, wrote some code, tried to get rotor to sync up with the software/MOSFETs. The motor didn't work, but the code works, I get a good rpm reading on the LCD.
-Replaced the resistors on the regulators, upped the current limit to about 5 amps. I'm afraid if I go much higher, I'll fry up some wiring.

Any other ideas?

[mackrackit]

You need something magnetic between the copper on the rotor. Think of the rotor as a shorted transformer. The copper rings and bars are the short.

Wiki should have said the idea is to induce a magnetic field, not current.

Think of putting a nail on a magnet and the nail will then be able once the magnet is removed, to pick up a paper clip. The nail will also have a N and S pole. Rotate the magnet and the nail will try to move (compass). The rotating field of the stator is doing the same thing to the laminates on the rotor. All the copper is doing is causing a short.

There are several theories for the shorting, The one I feel is correct is to create an insulation between the magnetic parts of the rotor.

Having the electro/mechanical part of the rotor on a twist does help starting as the magnetic parts will cross phase ( in two fields at the same time ) .

Over lapping the stator coils will also make a difference but what you have should move. (spin start? maybe) The bare metal in the stator coils should be facing the rotor, like you had in the first place.

Look at this http://ww1.microchip.com/downloads/en/AppNotes/00887a.pdf

If this does not help, I have a friend who owns a re-wind shop. I will get his input.

[skimask]

You need something magnetic between the copper on the rotor. Think of the rotor as a shorted transformer. The copper rings and bars are the short.
So, you're saying, if I were to lay the rotor out flat, it would go: copper wire (shorted to every other piece of copper wire), insulator, slab of something magnetic-able, insulator, another copper wire, etc.etc. I'll give that a shot...


Wiki should have said the idea is to induce a magnetic field, not current.
It did...but it said inducing a current is basically a side effect of the magnetic field. Same thing but different...


Having the electro/mechanical part of the rotor on a twist does help starting as the magnetic parts will cross phase ( in two fields at the same time ) . Over lapping the stator coils will also make a difference but what you have should move. (spin start? maybe) The bare metal in the stator coils should be facing the rotor, like you had in the first place.
Overlap - I've tried a few different angles so far, even angling the stators the opposite way to the rotor angles...
Spin start - no joy...
Bare metal - that's what I thought...and I've put it back that way.

I just wanna see this thing spin so I can giggle like a little 12 year old...

[mackrackit]

So, you're saying, if I were to lay the rotor out flat, it would go: copper wire (shorted to every other piece of copper wire), insulator, slab of something magnetic-able, insulator, another copper wire, etc.etc. I'll give that a shot...
Yep, but you will not need the insulator. The current created and suppressed in the copper is the magnetic insulator.



It did...but it said inducing a current is basically a side effect of the magnetic field. Same thing but different...
OK, must have been a different article, looked it up on wiki to see where you were coming from. Chicken or egg, which came first?



Overlap - I've tried a few different angles so far, even angling the stators the opposite way to the rotor angles...
Spin start - no joy...
Bare metal - that's what I thought...and I've put it back that way.
Not sure about the spin start, never seen one with straight coils. You could add a start sequence to the code like a stepper if needed.
Attached is a bad drawing of what I mean.



I just wanna see this thing spin so I can giggle like a little 12 year old...
Always nice to do something for fun, and not have a dead line.

[mackrackit]

Might be able to use welding rod between the copper on the rotor. Flux not included:)

[skimask]

Yep, but you will not need the insulator. The current created and suppressed in the copper is the magnetic insulator.
I meant that in terms of electrical insulator, like a thin piece of plastic or something.


Attached is a bad drawing of what I mean.
Yep, that's what it looks like, roughly...same idea all around.

I wonder, if and when I get this thing going, how much power I'll be able to pull out of it.
Rebuild it, put a bigger shaft in it, beefier wiring, bigger copper, and so on and son on...
I wonder what the limit would be...on a beer budget anyways...

[mackrackit]

Well ,on a beer budget the aluminum could be used in place of the copper:)

Really, I do not think there is a limit. Time maybe. I have been working on a 1KW wind machine for awhile now. Permanent neo magnets, quad disk rotors 3 phase each, 6 foot turbine.

Get a little money, get a little time...

[skimask]

Do I even have my nomenclature correct for this type of motor?
Is what I'm trying to build really a 3 phase AC synchronous motor?
Or is it really an induction motor with 3 driving phases?
I did some more reading and it almost seems like an actual 3 phase AC synchronous motor has a powered rotor, with the 3 phases on the rotor and the 3 phases on the stator out of phase, and it's that out of phase that makes it turn.
And I'm still trying to get my head around how that 'shorted rotor' with all of its 'shorted' copper bars and wires actually generate that 'counter balancing magnetic force'... Got my left hand out with the 3 fingers pointing every which way...maybe I should use my right hand?

[mackrackit]

Do I even have my nomenclature correct for this type of motor?
Yes and No

Is what I'm trying to build really a 3 phase AC synchronous motor?
No, You are building an Asynchronous motor. Synchronous motors need to be spin started to bring then up to sync speed. The starting motor is called a "pony" motor. Or a wound rotor can be used. These are not common because of the cost to build and the controls needed.

Or is it really an induction motor with 3 driving phases?
Yes, Because a EMF in induced at the rotor. In the US we speak of single and three phase, but what about two? This is what adds to the confusion. I would guess your water heater runs with two 120 volt legs, ths is really two phase. Look at vector diagrams for "delta" and "star" sometimes called "Y" to get a visual on it.

I did some more reading and it almost seems like an actual 3 phase AC synchronous motor has a powered rotor, with the 3 phases on the rotor and the 3 phases on the stator out of phase, and it's that out of phase that makes it turn.
Yes, see above.

And I'm still trying to get my head around how that 'shorted rotor' with all of its 'shorted' copper bars and wires actually generate that 'counter balancing magnetic force'... Got my left hand out with the 3 fingers pointing every which way...maybe I should use my right hand?
You may have to take your shoes off for this one:) one hand will not work

I scaned a few pages from one of my books "Electrical Circuits and Machinery Volume 2 Alternating Currents" Copyright 1942.

I hope you will be able to read it as the pages are getting somewhat on the yellow side.

Again there are two ways of looking at this. Both using the "shorted" idea.

1- Because of the multiple shorted circuits a large current is produced on the rotor, giving a large magnetic flux. This causing the iron between the copper to become a magnet.

2- If an iron bar is placed in a magnetic field it will take on the prperties of the magnet with one end being "N" and the other being "S". If the maget is rotated the iron bar will rotate too. The large current between the iron bars on the rotor act as a shield so each piece of iron has properties independant of the other.

Visualize the setup from the end with four poles . Every 90 degrees around the stator at any point in time there will be N,S,N,S. Looking at the rotor as a whole, an inductor with multiple high current shorts.

The thing with 3 phase and 120 degress between each phase mahes a rotating field with lets say very small gaps. You might think of a stepper motor setup with very small steps.

Single and two phase motors need an extra winding to start to "throw things out of phase" and some times to keep things out of phase. ( built in phase converters ) Start caps and run caps are sometime hooked here.

I hope I am not making this more confusing.

[mackrackit]

More from the book.

[skimask]

Synchronous motors need to be spin started to bring then up to sync speed. The starting motor is called a "pony" motor.
A couple of extra windings or something on the end, just enough to give it a bit of a kick?


Because a EMF in induced at the rotor. In the US we speak of single and three phase, but what about two? This is what adds to the confusion. I would guess your water heater runs with two 120 volt legs, ths is really two phase. Look at vector diagrams for "delta" and "star" sometimes called "Y" to get a visual on it.
Have basically the same types of things on the aircraft. Gyro motors are 3 phase motors, but they'll run on 2 phases as long as you don't lose the common ground. Got an emergency inverter for a backup system. Takes DC and outputs 2 phases and a ground. I have a heck of a time trying to explain to the younger guys why the emergency inverter isn't designed to start the gyro motors (even though that's how the pilots check out the inverter and end up burning them up occassionally), it's only designed to keep them running in the event of a power failure.
About the 1 vs. 2 vs. 3 phase thing...I get that too. A lot of people seem to think they've got 3 phase power because they've got 220 hooked up to a dryer. I let them think that 'cause you just can't explain things to people who won't listen.


1- Because of the multiple shorted circuits a large current is produced on the rotor, giving a large magnetic flux. This causing the iron between the copper to become a magnet.
2- If an iron bar is placed in a magnetic field it will take on the prperties of the magnet with one end being "N" and the other being "S". If the maget is rotated the iron bar will rotate too. The large current between the iron bars on the rotor act as a shield so each piece of iron has properties independant of the other.
Ok, so it sounds like I really do need some magnetic material between my copper wire on the rotor. I've got an idea of stamping out some recetangles from an old car hood I've got laying around, epoxying a bunch of layers together, enough to make the thickness of the 20ga wire on the rotor, or more to make the air gap even tighter, and then mounting that mess to the rotor somehow. Maybe screw it down, or tape it up with some heavy duty plastic tape.


Single and two phase motors need an extra winding to start to "throw things out of phase" and some times to keep things out of phase. ( built in phase converters ) Start caps and run caps are sometime hooked here.
Shaded pole and that sort of thing?


I hope I am not making this more confusing.
I'm all about getting confused...as long as I get this thing spinning soon!

[mackrackit]

A couple of extra windings or something on the end, just enough to give it a bit of a kick?
Yep, that is why synchronous motors do not have good starting torque.



Have basically the same types of things on the aircraft. Gyro motors are 3 phase motors, but they'll run on 2 phases as long as you don't lose the common ground. Got an emergency inverter for a backup system. Takes DC and outputs 2 phases and a ground. I have a heck of a time trying to explain to the younger guys why the emergency inverter isn't designed to start the gyro motors (even though that's how the pilots check out the inverter and end up burning them up occassionally), it's only designed to keep them running in the event of a power failure.
About the 1 vs. 2 vs. 3 phase thing...I get that too. A lot of people seem to think they've got 3 phase power because they've got 220 hooked up to a dryer. I let them think that 'cause you just can't explain things to people who won't listen.
I do not bring the 2 phase thing up often. You should see the look on an inpectors face when I label a sub panel "2 PHASE".



Ok, so it sounds like I really do need some magnetic material between my copper wire on the rotor. I've got an idea of stamping out some recetangles from an old car hood I've got laying around, epoxying a bunch of layers together, enough to make the thickness of the 20ga wire on the rotor, or more to make the air gap even tighter, and then mounting that mess to the rotor somehow. Maybe screw it down, or tape it up with some heavy duty plastic tape.
Might be able to use welding rod. Laminates are used in transformers to suppress eddy currents, in small motors for cost. Large motors are cast iron/steel with a second pour for the aluminun. Air gap should be the same for magnetic and non magnetic parts. Sounds like you have punch press, How about a lathe? To attach things for testing, J B Weld.



Shaded pole and that sort of thing?
Yes, again that is why the rated HP is low on these motors. Back to top^.



I'm all about getting confused...as long as I get this thing spinning soon!
Can you imagine the spinning in Tesla's head. He reportedly modeled this all in his head.

[skimask]

Yep, that is why synchronous motors do not have good starting torque.
Then how do they get those AC types to work in electric cars (the ones with AC motors anyways) and the like?


I do not bring the 2 phase thing up often. You should see the look on an inpectors face when I label a sub panel "2 PHASE".
I checked the wiring at my house awhile back on a 'scope. I've got 2 legs of 120VAC, both in sync with each other, not 180 or 60 or 120 out of phase. What's up with that?


Might be able to use welding rod. Laminates are used in transformers to suppress eddy currents, in small motors for cost. Large motors are cast iron/steel with a second pour for the aluminun. Air gap should be the same for magnetic and non magnetic parts. Sounds like you have punch press, How about a lathe? To attach things for testing, J B Weld.
No lathe, no punch press, just access to a foot shear and a decent ruler, should be able to punch out a pile of decent rectangles in short amount of time. I wasn't going to use JBWeld, just whatever epoxy I could find that would work reasonably well on sheet metal. If that doesn't work, I'll pop-rivet the parts together. If I had a plasma cutter, I'd cut out chunks of frame rail and use that instead of the stacked up rectangles.
I don't have any welding rod handy, and I've got all this old scrap auto sheet metal laying around.


Can you imagine the spinning in Tesla's head. He reportedly modeled this all in his head.
I have trouble enough with 2 dimensions, let alone more than 3...by that I mean trying to imagine the interaction between voltage, current, and magnetic lines of flux...all at the same time...

Do ya think it matters much that I'm driving the thing using square waves vs. a halfway decent sine wave?

[mackrackit]

The AC electric cars I have looked at use Async motors like the one you are building. Spped control is easy with a varibl frequency drive like the one you are building. Depends on how the motor is built as to the low rpm torque developed.

Synchronous motors with rotor windings to a controler can give good low rpm torque but then the overhead will increase, and more trouble with slip rings and all of that.

I knew I was going to confuse something. Your house like most I think have one transformer connected to one of the phases provided by the power company. The transformer has two output taps, both 120 volt. So both legs are on the same phase.

A typical 240 three phase will have one 240 leg and two 120 legs. ( some are 208/277 but that is another story) All coming from different transformers. Look at a power utility pole. If I place a meter across any to legs or phases it will read 240 volt. The water heater thing I mentioned is what messed this up .

The attached pict is a schematic of a water chiller. The compressor is 3 phase and the circulation pump is a standard 240 volt setup. The pump motor runs from two of the 3 phases making it a two phase motor. The same motor will also run on the 240 two leg single phase at your house.

But you can not feed a transformer with one phase and have three taps and run a 3 phase motor.

[mackrackit]

Hi Skimask,

Did you get your motor working?

[skimask]

Hi Skimask,
Did you get your motor working?

Not yet, kinda slowed progress on the motor, holidays and stuff (birthdays, parties, etc).
I've gotten it to kick quite a bit, but not go 'round and 'round...so that's probably a software issue.

[mackrackit]

Understaand all of that (family first).

Might be a time problem. Could you post your algorithm. Might be able to help.

[skimask]

Might be a time problem. Could you post your algorithm. Might be able to help.

Don't have it here at work, but it's not as complicated as a person might think.
Just watch for the phototransistor input and switch to the next 'phase'.
I'm able to move the reflective strips to, in effect, change the timing of the 'phases' in relation to the coils.
The leds I have mounted to the base of the motor tell me everything is working as I've planned it to work.
It's got to be a problem with the air gap, or maybe I've got one of the coils wrapped backwards/sideways/upside-down, something simple.
It should work without a problem. I'll get 'er nailed down someday.