Quantum Energy Generator (QEG) Open Sourced (by HopeGirl)

[F_Brown]

I use Maxwell Ansys to simulate QEG, and to me you juste misinterpreted the the torque signal.
Here a torque and speed graphe from a simution.

My simulation is somewhat limited in a few ways.  One of those ways is that it presumes a massive flywheel.  One that is so massive that there is zero variation in the rotational speed of the rotor.  In other words in my sim the rotor maintains a constant angular velocity, which in the case of the image I posted is 6000 rpm +/- 0.000... 

Another way in which my sim is limited is that it fails to take into account frictional losses.

When the torque data from my sim is integrated over time, it clearly shows that significant input power is required to drive the system, and the input to output power or COP for this case is about 0.756.  When my sim is set up, meaning the value of the resistor in series with the primary is adjusted, to match the parameters of replication systems, the COP drops roughly to about 0.35 +/- about 0.1, giving a crude agreement with replication results.

Just how is it you think I am misinterpreting the torque data?

Notice the 0 Speed is upside left and scale for torque is right.
Note the correlation between negative acceleration and torque...
The load resistor is on the 2 "secondary's"  coils.
For me, the QEG is a simple reluctance machine, without mistery....
And yes, this machine needs something to start, noise, little voltage , little current or earth magnetic field, to start.

I agree the QEG is a simple variable reluctance machine.

I inject a little voltage noise into the primary circuit to give the parametric modulation something with which to start.

Also I noticed that when the value of the load resistor is made zero or very small.  Strange things start to happen like large and on going surging oscillations in the primary current.  With some values of low load resistance I have seen the efficiency of the system slightly exceed 1.  This I believe is due to discretation of time and error margins in the math used in the SPICE simulation engine.

In your sim what kind of a load have you placed on the system?

***

Come to think of it you mentioned you placed the load on the secondary coil. 

Just what value did you use for that load, and how did you determine the variable coupling coefficient between the primary to secondary?

***

Thinking more about this, you are graphing angular velocity rather than angular acceleration.  Why you mention the term acceleration is making me wonder what you mean by using that term.

Presuming negative torque in you graph means positive load on the drive motor, looking at your speed vs torque graph a drop in speed correlates to an increase in load upon the drive motor.  You neglected to show the primary current trace, although this increase in load upon the drive motor generally correlates with the time of high current flow in the primary.  Thus the loading pulse on the drive motor relate to the power required to pump AC current through the primary on each half of its cycle. 

This is what is expected.

Since the torque is negative, this action might be more intuitively interpreted as pulling current through the primary rather than pushing it, since the rotor at this time is being pulled or attracted back toward a pole by cogging force as the drive motor is moving the rotor away from the pole.

The positive torque peaks correlate to when the rotor is approaching a pole while current is still flowing in the primary due to its resonant nature and causing the rotor to be attracted toward the pole by cogging force.   

[isim]

It's me who misinterpreted yours graphic, excuse-me...
English is not my native language!


"Also I noticed that when the value of the load resistor is made zero or very small.  Strange things start to happen like large and on going surging oscillations in the primary current.  With some values of low load resistance I have seen the efficiency of the system slightly exceed 1.  This I believe is due to discretation of time and error margins in the math used in the SPICE simulation engine. "
Ok, to get a good and reproductive sim, I need to put a 0.02 µs step for a 12000RPM rotating stator(5ms)(250 steps by revolution) . And as you,  i think yours "Stranges things" are "due  to discretization of time". With 0.5µs steps the sim is not correct.
I put my the Spice part of my sim in attachement. The 4 coils are used in the main Maxwell Ansys program (see second attachement).
I can put a constante speed or a torque.

Here is some info on this sim: QEG size and Coils as in the "official pdf" with
Primary coils = 2*3500T; R=2*35ohms
Secondary coils = 2*350T; R=2*0.25 ohms
Load= 250 ohms

Torque= 0.4Kg m
Damping= 2e-005  N m s / rad
Stator inertia=0.000173 kg m²

Initial speed = 11045 RPM (in this sim)
Initial Vco=-15000V
Initial Current in "primary" Coils=-0.75 A

The switch on the schema is there to simulate (very badly, i must said) one spark plug if Vco>25kV...
And no suprise: no OV  Efficience< 35%

I will try to with a 6000RPM speed but with a load on the two "secondary" coils.

Is you send me yours schema, i can test it, to make a comparaison...
@+

[F_Brown]

Maxwell 3D huh...

I wish I had a copy of that. 

Anyway, here is a copy of my version 2.1 QEG SPICE model.  It's for LTSpice.  If you unzip the file into a folder and open QEG_V2.1.asc with LTSpice it should run. 

I've got easily over a couple hundred hours invested into research and development for this model.  Any and all gratuities would be appreciated.

[isim]

@FBrown
Thanks, i will try your setup!
No problem to get V14, but i don't now if i may put à torrent link on this threads!
But i will send you a link, if you send me your address by private mail
@+

[F_Brown]

Thanks, my email is on the SPICE model schematic.