For Jorge's new clips, there simply is no logical reason for using toroidal coils that I can think of.  In theory, you are supposed to try to reason things like this out in your mind before even building something.  Just pure trial-and-error building is not where you want to be.  You want to apply your knowledge when you undertake to build something.  Except for the unremarkable Steorn "disappearing core" pulse motors that were demoed at the Waterways presentation, I can't think of a single logical application for using a toroidal coil in a pulse motor.

I made my Version 1 of the ring magnet spinner a bit more sturdy. The flimsy and rattling ball bearings were replace with Teflon slab which just have a hole for the axle.

I think I can still tweak it a bit and the Hall sensor has to go underneath the ring magnet because I want to place a "magic generator coil" on top.

The most useful setting seems to be 12V which results in about 3000 rpm for 0.5 Watt.

The most difficult part is the mechanical precision which is till bad in my Version 1 and still worse in my Version 2, see my Reply #1239 on: April 26, 2013, 09:46:35 PM-

Version 1 was better with the rattling ball bearings, see my Reply #1243 on: April 26, 2013, 10:25:47 PM.

Greetings, Conrad

Hi Conrad, Yes I agree with these setups we want very low mechanical drag as well as low windage. My rotor pushes a lot of air, in a similar way to a Tesla turbine, the disc throws off air all round and quite strong too. Other things that reduce the possible speed are rotor not balanced ( mine needs a re-balance ), misaligned bearings ( with my wooden frame I can manipulate it to align the bearings while running for least drag, but the next setup will be built more precisely and with some adjustment on one side for bearing alignment, also flexible coil mounts can "give" when the coil repulses the magnet and vibrate, in some cases that could actually aid in rotation of the rotor but would ultimately cause vibration in the whole setup and if a vibration can aid, it can hinder as well. Unwanted vibration is why in my opinion opposing forces are best and why I need to make a second set of coils for the opposite side of the rotor more coils will mean better efficiency and more torque as well.  Because of my larger rotor my magnet speed is about 15 meters a second at 2400 rpm if my calculation is correct. But the frequency is low 80 Hz. The magnets are at 120 mm diameter.

My motor coil is actually more like 12.8 mH so with the 330 uF it's resonant frequency is more like 77 Hz which is 2310 rpm, so with a second set of coils it should be most efficient
at 2310 rpm or so and develop the most torque at a speed just below 2300 rpm. I think if I was to run the motor at increasingly faster speeds over it's resonant speed, the input would rise a lot for little speed increase.

Conrad I see you are using square wave AC to drive the spinner ( H bridge ), that's then actually a "permanent magnet AC motor" not really a pulse motor if there is not an appreciable time break between input voltage alternations and no coil discharges or harnessing of the energy released when the magnetic field of the coils collapse. No big deal, just my opinion but to be a pulse motor the duty needs to be less than 100 % regardless if AC or DC is used. My duty is about 40 % or less. With lower resistance and slightly less inductance or higher voltage input it could be less.

Last night I re-familiarized myself with the picaxe program and it's input and output configuration so I can setup the boost converter for control of the input voltage I should be able to get up to 30 volts input so the voltage in the capacitor to dump through the drive coil could get to 50 or 60 volts then which will help to overcome resistance and shorten the pulse width.

Cheers

P.S. I quickly wound a test generator coil of about 70 mH and it produces about 20 volts.

Oh and I can run the setup with up to 0.8 or 0.9 A input current from the 12 volt battery and the mosfets don't even get warm.


..

 

Hi Mags, I think the first post in this thread linked below explains why a toroid can leak significant flux when used in a pulse motor. The reason toroids are considered efficient is because they tend not to "leak flux". If they are made to leak flux that goes out the window so a core with a break in it or a solenoid may work better, maybe not it depends on the situation and what is wanted.

http://www.overunityresearch.com/index.php?topic=2061.msg31233;topicseen#msg31233

When pulsing coils with not much inductance switch heating can be a problem, also the coil collapse should be directed to discharge into at least a few volts above the supply voltage, in my opinion it should discharge into double the supply voltage.

Cheers

 Was editing my previous post and a new one popped up all in grey quote box.  Must have hit quote instead of modify.  Never deleted a post. Can that be done in edit?

Mags

snip...
The oldschool Soundstreams used 'loosely' regulated supplies so the rail voltages were 37v +/- for 50w rating vs a fully regulated Precision Power for the same rated power the rails were 20v +/-.  The loosely regulated supply offers more headroom for those dynamic peaks and still able to hold 20v +/- during a continuous output. But music is not continuous power most of the time.

Some of the "real old school" audio "Amps" cranking out 500W RMS Rock N Roll, with a 50-75V  +/- supply rail, are still in hot demand, and many are still in use. Early Marshall and Vox Amps spring to my mind. For the Rock N Roll set, high powered valve amplifiers never died, they've just become the highly priced sought after jewels of the few.

Sorry  ..... off topic  ..... Cheers

Comparison H-Bridge driver and single transistor driver for Version 1 of my ring magnet spinner:

I went back to a single transistor circuit with a Hall sensor for spinning the ring magnet in my Version 1.

The single transistor circuit is slightly more efficient than the H-bridge, but can only spin the ring magnet up to 5200 rpm. See the circuit, scope shots and photo (with the measurements).

It is not worth while to go from a single transistor circuit to an H-Bridge for a simple ring magnet spinner. May be that is useful information for experimenters who want to build a ring magnet spinner.

Comparison tests in the making: trigger coil instead of the Hall sensor in a single transistor circuit, DadHav's circuit with two MOSFETs.

Greetings, Conrad

I've made an improvement as well, I've got the input power down for the the rpm as well as shorter pulse width. Pulse width is just under 3 mS. Also I setup the boost converter with a max voltage of 20 volts and three settings for current. Still improving things and I have a new idea to reuse all reclaimed inductive energy with only one battery.

With two magnets the 360 degrees is split into two 180 degree cycles, if the drive coil has a duty of 25% that is 45 degrees so that should be the difference in phase, the charging inductor should be at about 45 degrees phase angle to the drive coil to get two phases from one pulse. I need to design a new signal processing circuit that retains 25% duty at all speeds. Then it will have more torque. At present at the lower speeds the duty is less and less.

Acceleration Test 1
http://www.youtube.com/watch?v=Yukaj0QQzN0

Considering it is really a single coil motor made to use the charging inductor as a second phase it is fairly efficient and responsive. With a second Twin coil arrangement it will have a fair bit of torque, I think this would work well for a pushbike motor-generator, with a larger rotor mounted on the bike's rear hub which has about 18 magnets as closely spaced as possible. If the bike wheel did say 5 revolutions per second and the rotor had 18 magnets that would be a frequency of 90 Hz. And with a 27 inch wheel that has a diameter of about 2.2 meters that is 11 meters a second or 39 kph. 

Would be helpful for reducing effort and going faster.

And if the circuit was switched off, then the diode after the battery is bypassed it would generate and charge the battery when rolling down hill or braking. A "swing in" generator coil could also be used so that it is out of the way when not in use.

It's a case of "Have TIG welder and plasma cutter. Will build crazy bike". I've got a history with powered push bikes. some years ago I installed a 2.5 hp two stroke victa lawn mower engine on a mountain bike, it had a direct "cylinder" friction drive to the rear wheel, the drive roller was mounted directly on the crankshaft. It did over 70 kph, and accelerated very well.  The sound of a 70 kph lawn mover really freaked out the neighborhood dogs though. hehehe.

Cheers

P.S. I would like to plot the acceleration and torque curves of this setup. A fixed timing point can be used with good effect so accelerating it is just a matter of input voltage control. It does have a "power band" or "torque band" so to speak.

Prototypes are not meant to be pretty. I don't understand people spending big time and big bucks on unproven prototype setups, it's not smart in my opinion, "pretty is, as pretty does" so Mr Gump would say. (referring to "Skycollection's pretty setups")


EDIT: Timing effect video clip. http://www.youtube.com/watch?v=KtgPwqJOZJk

..