I played around a bit with backing magnets.

1)  I noticed that my RPM will increase as backing magnets are moved from far away towards the back of the ferrite all the way until the magnets are touching the ferrite.  Mag orientation is always opposing the rotor mags.  However, Vout increases for only the initial portion of the install travel distance, and then begins to decrease as the magnet is pulled by the ferrite and eventually is attached.  So Pout eventually DROPS even though RPM is still increasing.

2)  I changed Rload to get down to the same running "sound" where things get a bit noisy due to load, so keep that in mind if comparing these two sets of data.

3)  I then placed the backing mags on spacers where the voltage is approximately optimized (did not use the o-scope or cut exact thickness spacers: used what I had).  I was only able to place top mags in this experiment but it proves a point:  By tuning with the backing mags to max Vout, and not max RPM, the Pout increases, while RPM still drops.

4)  I played around with placing a vibration dampening weight on top of the unit.  Damping out the vibrations also increases RPM and therefor Pout.  So when this thing gets noisy, that is not good.  The more rigid your build, the better, imho.

Quote from: slapper on July 22, 2011, 10:14:35 PM
i'm curious as to what others have observed concerning the length of exposed
core toward the rotor. all i see is more lugging with exposed core and i don't
think i'm seeing anymore voltage with the extra exposed lengths.

FWIW, I have not done any exact testing, so this is just conjecture.  I have slightly different ferrite core lengths protruding from my coils.  This was done so that they read exactly the same inductance on a meter.  I also matched the two bifilar coils that I tested to the same inductance as my monofilar coils by the same method and this required that even more ferrite protruded.

I think that having any protrusion of ferrite core (or even the opposite) is another trade off situation.  Extending the core through the coil (assuming your core is longer than the coil) increases the impedance and therefor the Vout.  But extending the core also brings it closer to the rotor magnets which can cause drag in the form of cogging.

If I am to believe Dr. Turtur, we are simply trying to oscillate the impedance of the coil as the L component in an RCL circuit.  So oscillating the impedance by swinging the rotor magnets past the ferrite coil cores would achieve the best results if the impedance change was maximized while the rotor drag was minimized.

In my case, the slightly different ferrite protrusion lengths also causes slightly different cogging effects and therefor vibration, which converts your input power to acoustic vibrations, and not rotor RPM as intended.

M.

   This soap box mentality is why nothing gets done. If you know then show.  If not then dont tempt.

thay

Quote from: webby1 on July 22, 2011, 11:21:52 PM
This is very much tongue in cheek,  What is a virtual pole? It is another direction of flow, is that virtual???

Franken Motor uses these pole values, many others do as well,,,

Dammed I can't help myself sometimes.

The answer is in front of you, ask the correct question.

I will now stop posting, the new way of making a force is what I need to work on.

Tom Webb

P.S.  Romero was very deceitful, his system may of worked very well but he was not fully open,, or should I say not fully shorted.

Quote from: Thaelin on July 23, 2011, 11:13:37 AM
If you know then show.  If not then dont tempt.

Amen.

I've cleaned up the wiring on my motor circuit input power as much as I can and still allow for measuring Vin and Iin.  Minimizing the wire lengths resulted in the expected improvement in performance of the BLDC drive motor: RPM increased.

I also noticed an RPM increase as I cleaned up the Vout wires as much as possible.  So another area for builders to optimize.

But this problem still remains:  I cannot measure Vin or Iin with my DMM accurately.  Every time I touch either DMM lead to the circuit in order to set up for a reading I will immediately notice an RPM change (audibly).  I believe this is still due to the "antenna" effect from my DMM leads introducing noise to my BLDC motor drive circuit.

Question to the group:  Is there a simple method to eliminate any noise from my DMM probe lead "antennas" from affecting my circuit and therefor readings?

For now I am trying to "tap in" quickly and record the first reading I can get, but this is far less accurate than I would like.  And time consumming since I must allow for the system to stabilize once again between each reading on the Pin side.

Bring it.

M.