Hi Stargate
The harbor freight 900w generator scope pboto looks like worn out brushes and bad connection througout!
The drawing by Pierre looks too smooth of peaks in middle.of wave to be what happens with peak coil shorting  but maybe his little scope displays it like that.


Could be like dog-one describes and and is parametric resonance (whatever that is exactly)

Quote from: pmgr on June 22, 2018, 06:32:42 PM
I have pointed out many times that the inductance changes as you go around the stator.

That is true. Except if the point is to create a minimum inductance when in full registration and maximum inductance when we are far from full reg. This would justify Jerdee's opinion, and also would minimize the losses.

Not to speak about the coil shorting technique. Dog-one said everything in one line.

une petite question pour Pierre ,
Comment étes-vous arrivée à stabiliser la tension de sortie à 139V malgré une fluctuation de 5V sur l'entrée?merci d'avance .
cordialement Mosha

Let's talk about a piston in an engine for a moment.  Suppose when the piston is accelerating downward we decrease the mass by 90%.  Then when the piston is nearly bottomed out, we return the mass to it's original amount.  Do that on every stroke, both upwards and downwards, twice per cycle.  Pretty soon that motor will accelerate into oblivion.  That's your parametric resonance.  A tough act to do mechanically in an engine, but...

We can do a very similar thing with a pair of inductors.  Short one of them to decrease the inductance, then open the short when we have reached the peak.

Now let's think a minute.  When we short, we create a current surge in one of the inductors.  What does this do?  It juices up the magnetic field in a particular polarity.  So what happens to a neighboring coil?  Inductance on the output side is the production of an EMF based on the change (derivative) of the current on the input side.  So in the DZ Gen, we need to think about rotating transformers circulating the stator core.  You have big current in one winding and big voltage in the other.  Every cycle we step around the core, current pushing/lagging voltage (don't get that backwards, because we want the EMF to add to the CEMF), building on top of the previous cycle.  If your rotation is such that the induced voltage is opposing the change in current, you're sunk.  Go the other direction.  Move those little transformers in a manner that is productive.

Just out of interest, I took this shot of the input current while I still had the relays working.
It used a set of relays to reverse the DC rail, so every other pulse changed polarity.
Output was insignificant,as you might expect.
L192