Controller for a Halbach DC 24-48V Motor

Thread Starter

MelMartinez

Joined Nov 28, 2011
17
Hi everybody !!!

I'm building a Halbach Array Motor/Generator and I need a controller to run it.

As a motor it would be DC 24-48 V; as a generator I have no idea how much power I can get out of it.

For some pictures please see my album:

https://picasaweb.google.com/117892529553203792785/ProtoWeb1

This is a plywood prototype. As soon as I get it working I would make it all aluminum.

Please help me with any advice or a diagram to build the controller.

The rotor has 12 permanent magnets and the stator has nine coils connected in three groups of three coils.

I'm glad to be in this very good forum and I thank you for any help or advice you may provide.

Good luck and kind regards.
 

strantor

Joined Oct 3, 2010
6,782
Welcome;
How much amperage do you need? when you say 24-48V, do you mean adjustable between those 2 numbers, or a fixed voltage anywhere in between?

What kind of lathe is that you got?
 

strantor

Joined Oct 3, 2010
6,782
Ah, sorry, I'd never heard of a hallbach array motor so I read up a bit on it. I take it you're looking for something like a BLDC motor control circuit for (by looks of the size of the thing) probably a couple hundred KW?
 

Thread Starter

MelMartinez

Joined Nov 28, 2011
17
Thanks for the reply strantor. I don't know yet how much amps I need. I guess I better check the coils first. I'll load them with DC 12V first and read the Amps.

As a motor I want to connect it fixed at either DC 24 or 48V. And you're right, I need a controller for a brush-less DC motor, but I don't have a diagram to build it.

The lathe is a South Bend model 1930. It turns 14.5", or up to 22" if removed the bed extension. It has 5'6" between points. I rebuild it three years ago and it runs like new.
 

strantor

Joined Oct 3, 2010
6,782
Ok, from now until somebody corrects me, I will assume that we are basically dealing with a BLDC controller.

I would read the resistance with a multimeter before I hook up 12VDC from a car battery and try to read the amps.

I expect you'll likely find that your windings read very low resistance, maybe less than 1ohm. if you do the math on that, (ohms law) you might find that your motor will try to draw thousands of amps if you just hook up DC to it. This is why most (advanced controllers, not like PC fan motor circuits) BLDC controllers have some kind of PWM current limiting.

I've also read (very limited reading) that hallbach are usually air-core motors, which doesn't seem to be the case in your motor, but nevertheless they were saying that the hallbach motor lacks torque at low speed, so generally needs to spin very fast. is that the case? How fast will your motor need to spin? This is important because the PWM frequency (for current limiting) needs to be a good deal higher than the actual commutation frequency, and all this times number of poles for mechanical rpm, etc. It can get crazy pretty fast. how many poles is the motor?

did you make this motor from internet plans? if so, do you have a link?
 

Thread Starter

MelMartinez

Joined Nov 28, 2011
17
I did check the resistance of the coils after soldering and it's near zero ohms.

A well made and balanced Halbach motor may spin from 60K to 84K RPM with no load.

The stator has nine coils in 3 groups of three coils, then it has three poles. The last picture in the album shows the stator diagram and connections. I made it the same.

I have no plans nor blueprints for this motor. I'm making it with scrap plywood ans some left overs from other projects. I make every part to fit the others as needed.

I thank you very much for your help and advice. I can't wait to see it running and make an electric bicycle with it.
 

Thread Starter

MelMartinez

Joined Nov 28, 2011
17
I don't understand the term "air-core" motor, besides that the coils will need some good air ventilation because they heat up really fast. I have that covered.

It's true Halbach motor have low torque at low rpm, but it increases at higher speed.

At this point I don´t know how fast I want it to run; I want it to run first with DC 12V and measure the rpm, torque, HP and overall performance. After that we can increase voltage to 24, 36 and 48V and do the same testing.

The purpose is to find the best configuration for this particular model, and improve it before I do the aluminum and bronze castings.


I've also read (very limited reading) that hallbach are usually air-core motors, which doesn't seem to be the case in your motor,

but nevertheless they were saying that the hallbach motor lacks torque at low speed, so generally needs to spin very fast. is that the case?

How fast will your motor need to spin?

how many poles is the motor?
 

strantor

Joined Oct 3, 2010
6,782
I did check the resistance of the coils after soldering and it's near zero ohms.
yup
A well made and balanced Halbach motor may spin from 60K to 84K RPM with no load.
Holy expletive mel! At least warn your neighbors before testing! all joking aside, I will trust you built it well and it won't self destruct; I'll just focus on the electrical stuff.

I'm a little confused by your design, having more magnetic poles than electrical poles. poles are usually referred to as pole pairs (per phase), indicating one tooth on one side of the motor, and one corresponding tooth on the opposite side. since you have 3 adjacent coils in series, there really are no pairs. you have 3 phases there I guess, but I believe you really only have a one pole per phase. Correct me if I'm wrong; I really don't understand the motor, but you made it, so I'm expecting that you do. please explain, for my own understanding, I am intrigued.

Ok, if you are spinning 60,000rpm and your motor is single pole, then your commutation frequency is S = f120/n = 500Hz. 500hz is the frequency you need to switch if you are not using PWM current limiting, rather just dumping pulses of full current into the phases. This method of not using PWM current limiting will probably result in a bunch of blown up semiconductors and heartache.
A thumbrule I have seen is that your PWM frequency should be at least 10X the commutation frequency, so 5khz minimum. for better control, and decreased noise, you may try to go above 18khz (outside the audible range) or above.

I suggest using MOSFETs for the switching. you will need 6 (big ones) to make a bridge. you will need a microcontroller for the switching logic. you will probably need hall sensors to control the commutation. outside of that, I don't know what to say at the moment. There really is a lot to this; its not as simple as just directing you to a link of a control circuit. there are many many factors at play and I don't want to send you off on a quest of buying parts that might not work for you. you really need to read up and realize what you are dealing with; I think you will find that all the hours and planning you put into the motor itself will have been the "easy part" when its all said and done.

here are some links to explain how the control circuit should work (assuming its just like a BLDC):
http://electrathonoftampabay.org/www/Documents/Motors/Brushless DC (BLDC) Motor Fundamentals.pdf
http://robots.freehostia.com/SpeedControl/SpeedControllers.html
http://www.freescale.com/files/product/doc/AN1916.pdf

these should get you started understanding how the controller should work, but after you understand those, then you have to get into the nitty-gritty of the actual components and layouts and such, lots to learn before you see that thing spinning.
I thank you very much for your help and advice. I can't wait to see it running and make an electric bicycle with it.
I was going to make you a recommendation before I knew this was for an ebike. Now that you've said it, its even more fitting: www.endless-sphere.com. Those guys know all about motor control and they are ebike freaks. I think you could get lots of help there.
 

strantor

Joined Oct 3, 2010
6,782
I don't understand the term "air-core" motor, besides that the coils will need some good air ventilation because they heat up really fast. I have that covered.
the core is whats inside the coil. you appear to have grade 5 galvanized bolt cores, not air cores.
 

shortbus

Joined Sep 30, 2009
10,045
This I think, is what is known as an 'out-runner' type motor. Your going to need to come up with a Hall sensor group, to trigger the coil firing.

Here are some links to a fairly easy BLDC motor control chip. All the necessary drive circuit are self contained within the chip. By adding the Hall sensor inputs, a capacitor and resister(for the oscillator) a pot for speed control and sense resistors for current control, the motor part should be good to go.

Chip one- http://www.onsemi.com/pub_link/Collateral/MC33035-D.PDF

Chip two for different voltage, other wise same as first one. http://www.onsemi.com/pub_link/Collateral/MC33033-D.PDF

If you want to run the motor in a closed-loop speed. http://www.onsemi.com/pub_link/Collateral/MC33039-D.PDF

And last but not least application notes. http://www.onsemi.com/pub_link/Collateral/AN1046-D.PDF

With these chips you can run at any voltage you want by just changing mosfets to suit the need.

Don't know every thing (some would say nothing), but haven't seen a motor or generator with a winding phased like you show. Usually the are every other pole a different phase. To have a "rotating" magnetic field.

And like Strantor said you need opposing coils, to make it work as either a motor or generator. The same goes for the magnets in the rotor, the one opposite has to have the opposite polarity.

Hope I'm wrong about this, and NO offense meant if I am, but did this design come from one of the free energy or alternate energy sites?
 

Thread Starter

MelMartinez

Joined Nov 28, 2011
17
I thank you very much for your advice, strantor and shortbus. You guys have opened my eyes about this project.

The only similar Halbach motor I've seen are those very tiny for model airplanes. They are made in China and use a very small BLDC controller.

Now I have a lot for reading from here to the weekend.

In the other hand, I've found this very interesting document that show a simple BLDC controller diagram based on six IGBT's. It looks quite simple, but it shows no details:

http://www.askmar.com/Magnets/Halbach Array Motor.pdf

Please see it and you'll find a very interesting Halbach array applications developed by the Lawrence Livermore National Laboratory.

I'll keep you posted with more progress.

Kind regards.
 

strantor

Joined Oct 3, 2010
6,782
here's a link for you.

I'm about 75% sure this halbach motor is not a crack free energy thing. However a lot of people are making crack overunity claims about them. They should be just another energy efficient motor, on par with BLDC.

Another thing I ran across when reading about the hallbach motor is that magnets should be touching (flush) on the sides to get the hallbach effect out of them. The pie-shaped sections between your magnets are going to be bad news.

In case you can't tell, you've sparked my interest with this hallbach motor thing. I hope you get it working. You're one of the few people I've found that's actually made one; everybody else is just talking about it. please keep us (or just for my sake) updated with what you find.
 

shortbus

Joined Sep 30, 2009
10,045
I have a few questions for you. Where did you find magnets like that? Where are you going to get the 'magnetic bearings'? How are you going to develop the vacuum needed? How deep/long is your final rotor and stator going to be? What power supply are you going to use to "spin-up" the motor for the 1-2 second burst of "stored energy"?
 

strantor

Joined Oct 3, 2010
6,782
I have a few questions for you. Where did you find magnets like that? Where are you going to get the 'magnetic bearings'? How are you going to develop the vacuum needed? How deep/long is your final rotor and stator going to be? What power supply are you going to use to "spin-up" the motor for the 1-2 second burst of "stored energy"?
it sounds like you've been reading the "crack free energy" version of the halbach motor design. The halbach motors that are actually out there and working right now don't employ magnetic bearings or spin in a vacuum or any of that other wackiness.
check out this csiro motor.
open the zip file and check out page 235 of SolarCarMotorPaperAsPublishedSml.PDF

And BTW Mel, there's a lot more helpful info in that zip file for you as well,.
 

Thread Starter

MelMartinez

Joined Nov 28, 2011
17
Another thing I ran across when reading about the hallbach motor is that magnets should be touching (flush) on the sides to get the hallbach effect out of them. The pie-shaped sections between your magnets are going to be bad news.

In case you can't tell, you've sparked my interest with this hallbach motor thing. I hope you get it working. You're one of the few people I've found that's actually made one; everybody else is just talking about it. please keep us (or just for my sake) updated with what you find.
One of the purposes of building this plywood prototype is to be able to change any major part easily. I can always turn an smaller diameter rotor and replace the magnets touching on the sides. I'd do that in one day at no expense. It would be a fun weekend job.

I'm just trying to build an efficient motor.

I'll keep you posted with some pictures and progress.

Kind regards.
 

Thread Starter

MelMartinez

Joined Nov 28, 2011
17
I have a few questions for you. Where did you find magnets like that?

Where are you going to get the 'magnetic bearings'? How are you going to develop the vacuum needed? How deep/long is your final rotor and stator going to be? What power supply are you going to use to "spin-up" the motor for the 1-2 second burst of "stored energy"?
The magnets are low cost low efficient made in Mexico. I have hundreds of them and also in different sizes. I use them for prototype purposes only. After I have this model running I'll use neodymium magnets.

Magnetic Bearings??? For very high rpm I use a set of slippery bearings. But that's a secret trade not for the public.

Vacuum, I don't know. Sorry I'm just a machinist trapped in moving sands up to the neck.

The final dimensions of rotor, shaft and stator are in progress. I'm still working in the shaft and I may have to make another rotor with an smaller diameter. But you'll see the final dimensions soon.

"Stored Energy" I don't know neither. All I know is that I want to connect from 1 to 4 car batteries to the controller and make it run with DC 12, 24, 36 and 48V... and see what happens.

Please excuse me if sometimes I don't follow you. I don't have your experience in electronics like you do. That's why I'm in this forums looking for some help.

I thank you again for your help and patience.
 

strantor

Joined Oct 3, 2010
6,782
Please excuse me if sometimes I don't follow you. I don't have your experience in electronics like you do. That's why I'm in this forums looking for some help.

I thank you again for your help and patience.
Nobody will get "frustrated" if you don't understand something, but be prepared (anywhere you go) to have this motor challenged and be ready to defend it as a real technology and not some wacko free energy thing. Not many people know about it.

Here in the forum we have a strict rule about "no HHO, no overunity, no perpetual motion" - basically the "free energy" type stuff. There are no shortage of people out there who think that the laws of physics don't apply to them, and that they can construct these marvelous devices that will conjure up kilowatts of energy out of some kind of never-ending chain reaction. A lot of these schemes are centered around ficticious types of motors (with "magnetic bearings" and/or "sealed in vacuum", etc), so any time somebody comes around asking about a "revolutionary" type of "new motor", suspicion abounds. I was fairly certain the halbach motor was a "free energy" scheme when I first started reading about it; after reading more, I'm fairly certain its not. I believe it is a slightly more efficient brushless motor. I have seen 97.5% efficiency quoted, but I have seen the same numbers quoted for certain regular brushless types.
 

shortbus

Joined Sep 30, 2009
10,045
Mel and Strantor- all of the information I asked about was in the link Mel posted.

When you change the magnet arrangement from the Halbach to just standard segmented magnets, you just have a BLDC motor. The Halbach uses a 'quasi-dipole' magnet array. Each segment has the poles arranged in a specific way. They would be very expensive to reproduce.

The Halbach also needs to be both in a vacuum and have a precisely balanced rotor. Due to the high RPM and to reduce windage of the motor. Again due to the weight of Halbach's rotor it needed the magnetic bearings to allow it to spin at high RPM.

As Halbach designed it, it really wasn't a motor, but a "EMB" electro-mechanical battery. It would operate up to a certain speed as a motor, then coast down as a generator and speed up again. Even though it was developed by/at Lawrence-Livermore Laboratory, it still was a 'over-unity' attempt. Never figured in frictional loss and electrical loss in changing from motor to generator.

Don't know about what others are trying to do with it.
 

thatoneguy

Joined Feb 19, 2009
6,359
I've seen RC aircraft guys re-magnet their outrunner BLDC motors to the Halbach design for a more focused field. Speed was increased as well as overall power, but not a lot for battery runtime savings. Worse if anything. The BLDC controller needed some severe tweaking.

The hardest part is to actually get all the Neo magnets to touch for the entire circumference, they have a lot of repelling energy, so various epoxies and clamp methods were tried. I'll try to find the article on it, which also gave tips on what to do to make a standard BLDC controller work with it. IIRC, they used feedback from one of the windings during the PWM Off period to determine rotation and timing. i.e. Back EMF method rather than hall sensor method.

--ETA: Many Refrigerator magnets are of the flat Halbach type, though very thin, they stick well.
 

strantor

Joined Oct 3, 2010
6,782
Mel and Strantor- all of the information I asked about was in the link Mel posted.
DOH! I missed that part. I just skimmed down to the part where it talks vaguely about the control circuit. I see where you are coming from now, but if you look around, there are (very few) halbach motors out there and they don't use any type of vacuum or magnet bearings.
They would be very expensive to reproduce.
I hadn't given much thought to that, but you are absolutely right. Not only the magnets themselves, but machining the specific shapes, polarized in a specific direction. It would most likely have to be all done by hand. That's probably why there aren't many out there. From what I've seen, you might gain only 1 or 2% (or less) efficiency by using a hallbach design over a traditional BLDC, and pay twice the cost for it. That why it would explain only being used in solar racing as far as I have found.
The Halbach also needs to be both in a vacuum and have a precisely balanced rotor. Due to the high RPM and to reduce windage of the motor. Again due to the weight of Halbach's rotor it needed the magnetic bearings to allow it to spin at high RPM.
the real world examples I've found don't spin at 80-something thousand RPM, and frankly that sounds rather outlandish to me. I see no real advantage to spin that fast. outside a few thousand RPM you have to start getting super complicated in the engineering. The ones I've found are multi-pole and they spin normal RPMs, like <10K.
As Halbach designed it, it really wasn't a motor, but a "EMB" electro-mechanical battery. It would operate up to a certain speed as a motor, then coast down as a generator and speed up again. Even though it was developed by/at Lawrence-Livermore Laboratory, it still was a 'over-unity' attempt. Never figured in frictional loss and electrical loss in changing from motor to generator.

Don't know about what others are trying to do with it.
Ok, so the guy who invented it was a free energy nut. that fact is a big kick in the pants for anyone trying to sell the motor as a genuine technology, but shouldn't eclipse the fact that the motor is a genuine technology, and is not an overunity machine. As thatoneguy said, all it is, is a BLDC with several times more magnets, and with the magnets arranged differently. I wouldn't want to build or buy one because as mentioned, the gain will only justify the cost in probably less than 1% of applications, but I don't think there's any point to continue assosciating the halbach motor with "free energy".

There are already brushless motors out there that are >96% efficient. I really don't see any point of designing anything more efficient if there is any price hike at all to go along with it.
 
Top