Muller Dynamo for experimentalists

[i_ron]

Hi.

I would really like to understand what is so special about that scope shot?
I've build many pulse motors and mostly all had the same induced voltage trace on scope.

As J.L.Naudin demostrates here:
http://jnaudin.free.fr/html/mromexp.htm

futuristic,

The scope shots that have been posted such as mine, conradelektro's and JLN's are what you would expect.

What romero posted is quite different... the narrowness of the pulse and the  'porch' on each side

Ron

[teslaalset]

futuristic,

The scope shots that have been posted such as mine, conradelektro's and JLN's are what you would expect.

What romero posted is quite different... the narrowness of the pulse and the  'porch' on each side

Ron

Possibly due to different core materials (different B-H curves).
There are soft and hard ferrites, I am not sure what type RomeroUK uses.
Possibly caused by different distances of the rotor magnets to the stator core.

[yssuraxu_697]

Ok, I have more-less cracked the mystery.

In short: Only thing that can generate Rom's trace is phase shift. Now the reason for that phase shift is whole another matter. I only succeeded with physically offsetting the coilpair.

Maybe woopy's test explains it:
"What puizzled me is that if load completely (short) the outut after the bridge, i got very low current (about 4 ma) and when i begin to add the biasing magnets, the current climbs (first test up to 50 ma)"

I cannot retest it because my setup is low-speed and just wont hold up w/o biasing magnets (too much cogging - breaks it into pieces).

Scope shots are added:
coilpair_antiseries_unconnected.png - shows that only outcome of connecting them in series would be "flat line" which it was when I did it
coilpair_series.png - coils in series, classical stuff
coilpair_series_phaseshift.png - Rom's trace - coils still in series but physically phase shifted by quite a few degrees
coilpair_antiseries_phaseshift.png - Naudin's trace - coils in antiseries and phaseshifted, very interesting because only other way to achieve this is rotating pickup coil or rotor magnet by 90 degrees, as per Naudin's schematic

Main difference between my setup and Rom's setup (as publicly known) - I have long cores (2x the coil) and can physically phase-shift coilpairs.

So I repeat: reason for trace is phase shift. Question is - from where? Biasing magnets? Phase shift under load? Under load with help from caps? I dont think that it comes from physical offset like in my setup UNLESS Rom pictured not a COILPAIR but PAIR OF COILPAIRS offset exactly like his drive coils...

Keep up the good work guys! This is good stuff.

[LtBolo]

Opposed huh?

The discontinuity could be caused by mismatched coils, misaligned coils, the bias magnets, and a handful of other things.

The discontinuity could in fact be the interesting point, regardless of where it originates.

On the other hand...

...opposition is very interesting too. From a 300Hz perspective, the opposed coils are mostly canceling, allowing for some difference between the top and bottom coils. But what about a 10Khz perspective? 100Khz? 1Mhz? Higher?

Why does it matter? Well, Litz wire makes no difference to a system operating at 300Hz, and the 1N4007s wouldn't make a dramatic difference in the performance of the bridge...both things that RomeroUK stated as being of great significance. What both of those will make a big difference in is high frequency response.

I put a coil near a rotating drum with radially mounted magnets, and turned on the spectrum analyzer. From what I could tell there was a harmonically rich spectrum originating from the coil. The very sharp transition as the magnet crosses the center-line of the core would be quite fast, and naturally be harmonically rich.

For frequencies high enough, the opposed coils really aren't so opposed. Think hairpin. Below a certain frequency everything would cancel...all of the stuff that is loading the motor. Above a certain frequency, they would not cancel, nor would they load the motor. Not sure how long RomeroUK's wire was, but for reasonable wire lengths the coils would have a 1/4 wave resonance in the single digit megahertz range. Well into the range that the big bridge would struggle and Litz wire would become useful.

The whole thing might end up as a sort of coil shorting exercise where the coil simply remains shorted...and only the very high frequency energy is harvested. Just thinking out loud...

[starcruiser]

So if this is the case LtBolo, the secondary diodes may be faster than the stock bridge he was using and would capture the higher frequency components perhaps?