Ciao a tutti

Secondo questa e_mail ci sono i primi riscontri
positivi da parte di un meg-builder ...

Saluti. Stefano



Hi All,
after the new Bearden MEG-2, now also
Jon Flickinger has verified, that a MEG can be overunity.
His latest tests show COP=1.32.

I have uploaded his proof of concept measurement GIF pic
over here:

http://photos.groups.yahoo.com/group/free-...lst?&.dir=/MEG2

Here is, what he wrote.
These are several messages pasted together for clearer understanding.
======

All,

Attached is the file MEG_se1.gif showing a set of traces for a MEG
configuration utilizing the ideas presented earlier regarding fast switching
of the MEG primaries and their flux with a relative constant current supply.

The primaries are in series and are switched at a 50% duty cycle with one
(1) FET only. The output loads are purely resistive 45K resistors with the
output voltage taken across 1/3 or 15K of the total load. The frequency is
50KHz.

Input power is .0516 x 150 = 7.74 watts. Output power is (160 x 3)^2 / 45e3
x 2 = 10.24 watts.

Cop = 1.32

Regards,

Jon F


Stefan,

IMO, there are several factors for the modest OU in my latest MEG version.
First, the input and output coils are constructed of flat ribbon lead which
reduces the inter-winding capacitance considerably. Second, this low
self-capacitance of the windings allows the fast switching transitions of
the flux in the core. Although the device may also be utilizing the A-field,
I personally think at this time that the device is flux operated but more
testing is needed to confirm this one way or the other.

Regarding the series connected coils, if the coils are connected in
parallel, there is no OU but this may be specific to this configuration. If
the coils are moved toward the center PM assembly, the output increases.
This is contrary to a MEG operating as a normal transformer. The 50KHz
frequency gives the highest output at present. BTW, the PM assembly is a
very modest 1/2" dia NIB stack at present.

The power supplies are not running in a constant current mode for this MEG
test as they were in the earlier low repetition tests. The relatively
constant current drawn from the supply is a result of conduction during only
half of the complete cycle.

I have a busy day ahead of me but plan to get back to the lab this evening
for more testing. I can also post pix later as well.

Regards,

Jon F


Hi Stefan,

----- Original Message -----
From: "overunity2001"

[snip some stuff]

(Jon said-)


> > An interesting experiment for anyone to perform in this regard is to
power a
> > MEG like structure from a constant current source thru the primaries in
a
> > series connection and then open the primaries with a very low duty
cycle.
> > IOW, keep the current on the primaries for a long duration between
openings
> > and measure the resultant output in the loads and compare to the input
> > power. Now increase the "off" duty cycle and observe the results and
note
> > the differences between a low duty cycle and 50%.
> > IMO, these difference give hints as to how OU may be extracted from this
> > type of device.

(Stefan responded-)

> Good idea. Well if you still have your MEG, maybe you can do this in
> the future ?

I have already run these tests in the past but I've attached a file with
scope pix to show an example of what I mean. The traces are self explanatory
with the exceptions noted below. I should add that the primaries and
secondaries used on the AMCC-125 core are low self-capacitance windings and
are key to this operation.

Please note the supply voltage of 1 volt DC and that's not a typo. The
current is limited to 100 ma and the actual circuit operating voltage is
~.65 volts leaving a compliance of ~ .35 volts.

The gate drive signal shown is a complement of the actual gate drive signal
for the circuitry. IOW, the FET driving the primaries is actually "off"
during the CH1 gate drive time of 1.5us. The repetition rate is 20 Hz for
each cycle.

Please notice the fast rise time of ~400ns of the drain voltage at the
FET's turn off and also the peak ~ 460 volt peak level it reaches. Also
notice the corresponding increase in the output voltage waveform. There are
two identical outputs and the load resistors are 45K ohms on each side. The
output voltage shown is 1/3 that of the actual value across the load
resistors. IOW, each load resistor is made up of 3-15K resistors in series
and the output voltage is measured across 1-15K resistor.

So, doing some calculations, the input energy during the 1.5 us turn-off is
~ .15 uJ. The output energy for each side is ~ 4.7uJ or 9.4uJ total. This
equates to a COP of ~ 63.

Now it is true that this is only 1/2 cycle of a real operating MEG and it is
also operating at a low repetition rate plus there is some loss to the
primary fields as is indicated by the 20% reduction in current. However, the
idea is that if this level of flux field energy can be maintained on a
continual cycle by cycle basis as it is maintained in this example by the
constant current, OU is apparently possible.

Regards,
Jon