The Ossie motor

[futuristic]

You can see from the bottom part of signal that it's size is 2.5 divs. 1div is 2V so AC is 5V.
But it could be that the Y scope axis is not calibrated to the 0 value correctly and that generator voltage is actually bigger than induced voltage.

The would also explain why there is no "bump" over the flat generator voltage.

[captainpecan]

A little update on my motor so far.  I have the dual pulse setup made and running now.  I've been banging my head against the wall tuning it, but I've finally got it running really well.  Only problem is that it's not as efficient as when I run it on one pulse per magnet pass.  Using 4 reeds is not easy to tune at all, and is next to impossible to pulse it correctly without causing a short circuit.  In other words, when the magnet is at TDC of the coil, it is trying to engage one reed from the first set, and one reed from the second set causing a direct short.  When the reeds are moved far enough apart to make sure there is no short circuit, then the pulse is not timed proper anymore.  Long story short, I'm working on biasing the reeds with a small permanent magnet to tune the pulses better.  Also I am working on different layouts for the reeds, to help avoid this tuning problem.  Including some idea's that were shared here earlier.

Overall, so far the double pulse setup does run much faster rpm, but it also loses more energy faster due to the slightly off time pulsing I have to be using to avoid a short circuit situation.  Good and bad results.  I'll share more when I get it running better.  Interesting enough, if I disconnect the bridge, the battery begins to charge up, without any recovery circuit connected.  Still weird, but of course murder on the reeds.

[captainpecan]

You can see from the bottom part of signal that it's size is 2.5 divs. 1div is 2V so AC is 5V.
But it could be that the Y scope axis is not calibrated to the 0 value correctly and that generator voltage is actually bigger than induced voltage.

The would also explain why there is no "bump" over the flat generator voltage.

Okay, so you see what I'm looking at too then.  I thought I was reading it right.  I was also seeing the wave drop deeper on the negative side than the positive and I figured that was where you were coming up with the higher induced voltage.  It just looked a bit funny.  What would help would be to let the motor get to the top rpms, then switch off the power and take a scope shot immediately if possible.  Then we could see the induced voltage wave to compare to.  Just a thought.

[captainpecan]

Okay, a little quick fun with photoshop and I think this is what is going on with that trace for sure.  It is generating a higher AC voltage than the DC pulse, but just barely.  It is the BEMF of the drive pulse itself that forces the voltage lower so we do not see the "bump" we would expect.  You can also see the small spike on switch opening.  That little spike seems to be close to the same voltage as the regular induced voltage if there was no BEMF at that moment of switch open.

A second thought from a bit deeper study.  When looking at how the coils were wound for this trace, leads me to believe the peak of the wave is more narrow than what JLNaudin and Ossie show because the inside core is wider, and the outside diameter of the coil turns is smaller than what Ossie and JLNaudin used.  Not that either is better or worse, just sharing what I am seeing by picking this trace apart.  In order to hide the pulse completely inside this AC wave it would require a fairly small pulse width.  This of course decreases RPM possibly to much.  Coil diameter may be factor to pay more close attention to, and not just inductance or number of turns.  A wider AC wave peak may be more beneficial for efficiency and higher rpms... At least that's what I see.

[gyulasun]

...
What would help would be to let the motor get to the top rpms, then switch off the power and take a scope shot immediately if possible.  Then we could see the induced voltage wave to compare to.  Just a thought.

Hi CP,

I agree and perhaps the best would be for Laurent to pull the MOSFET out of his board by hand, just after the power switchoff, to get rid of any influence on the waveshape the FET inner body diode may have.

Also, I wonder whether Laurent tried to find the best position for the Hall sensor, not only shifting it sideways alongside the circumference but radially moving it closer or away from the magnets rotational circumference.
That movements should change the length of the horizontal cut on his upper peak waveform.

rgds, Gyula