The Ossie motor

[futuristic]

Arrhhh that was the trace when I had probe connected wrong way so it should be flipped.
Actual ratio is then 97,46%. Well not bad anyway...

Frenky

[futuristic]

Video of motor with optocoupler CNY 37:
http://www.youtube.com/watch?v=CAPKqkrDx14

[woopy]

Hi frenky and Gyula and all

i have reread all what is at disposal about the 2 thin coils (from Garry Stanley) and is seems that there is anyway Lenz law but weak, (i would say about 1/2 Lenz).

That is the rotor is attracted with CEMF untill TDC and than it is liberated witout CEMF. So for each pulse there is half the CEMF (1/2 Lenz) is it somehow correct ?

So i made a test today  See pix

I tried to get attraction and repulsion of the rotor on a single pulse. And i placed the magnets to be slighty wider than the diameter coil  and the sandwitch magnets are inverted (one for attraction and one for repulsion) There is now 4 magnets acting together. And it works very well. I have to hold the coil with my hand or it slide away at start.

So in my previous config  (with only 1 active coil and sandwith rotor with each 4 magnets at 90 degrees)  i got  4 pulses per RPM per coil and 4 time the half CEMF = 2 full CEMF  (is it correct ?) per coil.

And now i get 2 pulses per RPM and also 4 half CEMF,   so this config ought to be much better efficiency as it is not the double torque per pulse but perhaps 1.5 more I don't know.

I think this config will not work correctly with reed switch. The hall sensor swithes only on south pole so it does not pulse when the second magnet passes.

What do you think ?

Regards

Laurent

[Magluvin]

I have 3 electric bikes. 1 is a Wavecrest Tidalforce. I enclosed a picture of the motor opened. I have not opened my motor but I have replaced the batteries in the front wheel hub.

Imagine working with one of these babies already made. If you zoom in you can see the curved magnets on the drum and the coils, though you cannot see from this side of the motor, are pared as 2's with the core shared on the backside. So as 2 coils fire the pair combined are very strong. This motor will do 1000w up to a regulated 30mph, not street legal, and the same motor is used in the 749w version and regulated to 20 mph but has some go up to 20 and quiet. The difference is software.
But I was thinking it would make an excellent platform to work with and there may be more eff to get out of it capturing bemf, as I dont think this does. Unless that force is used to help phase reversal. From what I know when the throttle is on, any speed, the coils are on always, just they are pwm pulsed to adjust torque. And it is opticaly timed, on the T shaped board there are 7 sensors, must be 1 for speed? Dunno that much but I have found all I can. the black and white coding disk is in the cover that is not shown here. Each the thickness of 1 core.
I just though I would share that. It can produce ideas. I would like to come across a broken one to just have the core elements, but these bikes are collectible and rare. Hard find. But there are many good chinese motors like on my Trek bike that are cheap enough to futz with.
Mags

[gyulasun]

Hi Laurent,

...
That is the rotor is attracted with CEMF untill TDC and than it is liberated witout CEMF. So for each pulse there is half the CEMF (1/2 Lenz) is it somehow correct ?   

I do not know whether half Lenz effect exists but it must be true that the flux from the coming magnet pair surely penetrates into the stator coil and because this coil has a closed electric circuit (via the switch and the battery), motional induction must also happen that works against the attract flux, this sounds Lenz.

So i made a test today  See pix

I tried to get attraction and repulsion of the rotor on a single pulse. And i placed the magnets to be slighty wider than the diameter coil  and the sandwitch magnets are inverted (one for attraction and one for repulsion) There is now 4 magnets acting together. And it works very well. I have to hold the coil with my hand or it slide away at start.

So in my previous config  (with only 1 active coil and sandwith rotor with each 4 magnets at 90 degrees)  i got  4 pulses per RPM per coil and 4 time the half CEMF = 2 full CEMF  (is it correct ?) per coil.

And now i get 2 pulses per RPM and also 4 half CEMF,   so this config ought to be much better efficiency as it is not the double torque per pulse but perhaps 1.5 more I don't know.

I think this config will not work correctly with reed switch. The hall sensor swithes only on south pole so it does not pulse when the second magnet passes.

Very good idea to use a second rotor magnet pair near to the first one and with opposite poles wrt the first one, so that with one pulse at the correct time, you get attraction and repulsion at the same time. 
Let's consider this link where member DMMPOWER states that electromagnets facing in repulse mode do not have back emf (he surely means cemf that comes about the motional induction) because the flux lines of one coil cannot enter the other coil due to the repel mode: http://www.overunity.com/index.php?topic=5890.0
If this is true, it probably true for an electromagnet-permanent magnet interaction when they are also in repel mode: flux lines from the permanent magnets would not penetrate into the coil that is excited from the battery to also create a like pole.
This would mean that for the repel coil-second magnet interaction for your case here you would not have to consider any cemf due to Lenz, right?  I hope this is so,     Then this setup you devised should be even better than just one attract or just one repel magnet pair.  And considering Garry used 12 magnet pairs in attract mode to his coil pairs, just imagine the torque gained in your present attract-repel setup with ,say, 16 magnets, 8 for attract and 8 for repel, just quadruple your present 4 magnet case.

Thanks,   
Gyula

EDIT:  What is the distance between the facing magnets now and what is the coil's hight, I wonder. Also the coil data like DC res, maybe L?