Just a few practical hints for winding toroids. Assuming you know how many turns you need . First take a short piece of wire , and wind 10 turns. Unwind it , and measure its length. You now can calculate total wire length needed . Add say 10% to be safe , more if the winding is to be multi layered. Find the middle of the wire . Wind the coil in 2 halves , start at say 6 o'clock and wind through 9 o'clock to 12 o'clock. Start again at 6 , wind through 3 to twelve o'clock . You may find that if the wire is long , a shuttle is useful . Take piece of 10 mm dowel about60 cms long , and cut a slot about 2 cms deep into each end [ like the slot in an arrow. Bind the dowel with thread to stop the slots splitting . wind wire end over end onto the shuttle , and keep threading it thru the toroid like a needle . Unwind wire from shuttle as and when necessary . Hope this helps someone.

Thanks for the info, it will most likely help save some alot of headaches.  I use pretty close to the same method, but there's a couple tips I can use.

Following everyones results, it seems that all that is required is to disrupt the domains in the core, to break the magnetic attraction.  So we should keep in mind that complete saturation may not truly be the goal here anyway. 

That is a matter of experimentation. If you can get the same torque while staying below saturation, its a bonus !
However, you don't want the rotor to get attracted in reverse direction, when it flies by the core. So you can compensate that by keeping the coil on for some more time, which means increased duty cycle and input power. Sure there is a trade-off.

Some more thoughts on input optimization:

Building some arrangement to vary the duty cycle manually will be best. Goal is to keep it minimum. When the rotor magnet is well past the core, there is no point in keeping it on.
Say if there are 4 magnets (or 4 pairs), its not good to keep it on after 45 deg. Seeing that the motor will work with as little as 25% duty cycle, its better to keep it as low as possible given that torque remains constant. If the core is driven into saturation, the viscosity (if there is such a thing ), will let the rotor not see it even if there is no current in it.

Collecting the induced emf (bemf or cemf or whatever) is a good thing, and it will lower the input even more, but I don't see how a battery would be able to keep up at high rpms, a cap will be needed if you plan to collect the emf.

That should cover the input side, IMO, and reduce some amount of hit-n-trial or rebuilds. Any more ideas are most welcome. Its very difficult to come up with design parameters at this time (and steorn is not telling any), but I will try to find something for output side also. And probably a decent method of measuring mechanical energy at the rotor, without which there is little point in building this stuff

captainpecan,

That's a great idea to cut round pieces of wood. I only have one size round cutter for the drill press, and it's too small though. What's nice about that idea is arrow dynamics. Unfortunetly I might end up cutting two pieces of wood, one for each side of the disc, and epoxying those to the disc and of course the magnets will be epoxied to the wood, but that will have some wind drag. BTW, does anyone think that epoxy can hold the wood & magnets at say 2000 rpm?

@PaulLowrance,

If you use Plywood and drill holes for the magnets AND then glue with Epoxy,
then the magnets will stay on the rotor at high rpm. It is no problem to
mount the coils above the magnets. If you have round magnets then the
drill method is the best way to do it.

Groundloop.

Hi to all,

I just watch at yours work since Steorn last presentation. I think you are doing grate job, and I hope that this project will succeed!
I want to point on Steorn patent:

http://www.overunity.com/index.php?topic=8411.msg218173#msg218173

I think it explains how orbo works, and how it should be optimized to work most efficient. Charts are self explanatory, and you can learn a lot. The main thing is that magnetic interactions in static system - due to the so called "magnetic viscosity" - are different than dynamic. According to patent when you increase relative speed between core and magnet, the core saturation is lagged in time and therefore its value decreased. While torque doesn't change as fast as saturation change (saturation decrease for about 100 times, but torque only 2). I wont explain every thing, you all have to read this. It is better explained there. Correct me if am wrong about this.

All the best!!!

marek

Hi marek,

Thanks for that important info! I just knew it, that it was due to magnetic viscosity. Since you studied the Steorn patent already, what type of magnetic core do you recommend?

Hi marek,

Thanks for that important info! I just knew it, that it was due to magnetic viscosity. Since you studied the Steorn patent already, what type of magnetic core do you recommend?

That I don't know, and patent says nothing about that. It just measure system. I think it must be set experimentally. It will depend on operating conditions (rpm, torque). When you look at the torque-flux chart I would try to run it at higher rpm, because you need less energy to desaturate core. But really, it has to be experimented with different cores materials. Worth thing to notice is that maximum torque is in quite narrow range before core, and drops to 0 when passes it. Negative torque too comes later, so maybe the electric pulse can also be lagged, and duty cycle can be reduced... Speculations;)

@PaulLowrance,

Yes, groundloop is right.  I always drill my magnets into the rotor whenever possible.  I use a Forester bit. Costs a couple extra dollars, but it makes perfect flat bottomed holes for round magnets.

This is what I was referring to that you can do with that hard drive platter.  This is not my Orbo replication, just a couple pictures of something else I'm working on also.  Hope it helps, it's pretty simple to do, and it lets you easily mount whatever you want to the rotor, while still using that awesome hard drive bearing!

Hi All,

Regarding Steorn's design and claim that the magnet's field on their motor does not induce an emf in the toroid coils, in an ideal world this should be correct. But in the real world nothing is perfect.

Using store bought chokes with few turns will typical keep this induced emf down to the 1-10mV range. But as you all start to wind your own toroid coils on ferrite cores with many more winding turns and winding imperfections then the store bought chokes, this non-ideal induced emf will get larger and typically it can get up to 100mV.

This induced emf due to the coil's imperfection can also make the coil slightly polarity sensitive. Normally, with perfect coils, the polarity of the coil connection does not matter because the direct link between the electrical energy in the coil and the rotor magnet's field is broken. But with imperfect coils where there is a slight induced emf in the coil, this induction creates a link to the rotor magnet's field and will effect the performance of the motor, slightly, depending on the amount of induced emf.

Undertanding this, the induced emf then become very important when using multiple coils around the rotor on your test motor. When you engineer and build your motors, magnet and coil placement as well as even magnet field strength becomes very important. This is because each imperfect coil's induced emf can become additive. But by specially checking for this with a scope and connecting the coils together with specific polarity you can get the induced emf to cancel or reduce to a minimal amount. Typically less then 100 mV for multiple coils.

I have found that my motor performance is optimal when this is done and the trigger is far more stable.

So seeking perfection and accuracy in the design and construction of you your motor is very important if you plan on using more then one toroid coil. I know this holds true for most pulse motors but I though it worthwhile to describe some of the reasons behind it for this motor.

Regards,

Ossie