Hello

I am glad to see that MEG builders groups still live (excuse my poor english because this is not my natal language but french). I signed up your group some days ago and I read a lot of your archives (not all yet!). So I used many advices to modify my MEG.

A friend has built a MEG (an electronic engineer did it for him) and he gave my his MEG on order to test it. So I uses your litterature because there was (and still is) many problems. The MEG is a copy of Jean Louis Naudin version 3.1 one. The driver circuit is exactly the same, magnet is not the same.

The Voltage and Intensity is not in phase at all in my mesurements. The circui used is very classic: One load is a resistive one (1 Mega ohm), and the other one is a serial assembly of two MOV rated 420V each and a 12 ohm (10W) resistor. The Intensity is measured throw the resistor (U = R*I) and the Voltage across the whole circuit, thanks to two probes connected to an oscilloscope and grounded on the side of resistor.

I experimented first a phase angle calculated to 87°, with approximate sinusoidal wave for output voltage, but an approximate tooth of saw wave for current, which led to the calculus: P = U * I * cos phi, where phi is phase angle. This led to a 75mW output measured with a 3,3W input (less than 2,5% COP) or with a 2,2W if you substract the driver power consumption (less than 3,5% COP), not good at all.

The frequency used was 19,2KHz. I decided to do another test with a slightly diffrent value of frequency. So I measured the capacity throw the two MOV: 200pF, and the inductance throw the output self: 12,8 Henry, and calculated the resonance frequency, which is about 3,1KHz, and modified the driver circuit (with a parallel capacitor of 10nF on the oscillater driving capacitor) in order it gave my the possibility to range from 1,6KHz to 5,5KHz.

I tuned my circuit and observed a resonance frequence measured at 3,2KHz (which is very close to 3,1KHz, I did not expect so close because oscilloscope probe and oscilloscope itself have a parasite capacitance I estimated to 50pF or something else....). But waves are not at all sinusoidal ones: but some kind of modulated signal. I spent two days calculating the exact output power with a numerical (and manual in a big part) computing of power with integration of the voltage and intensity product. I calculated 1,84W output (a 54% COP , or 84% COP if you substract the driver power consumption). This is not sur unity at all.

I have no more time to join all my experiment characteristic now, because I ended my computations yesterday in night, and I have to go out now, but more to come about it.

Pascal


carbonprobe wrote:

Is anybody out there in MEG Land?

My latest deep thoughts:

The MEG can be considered a parallel LCR circuit. When an LCR circuit
is in resonance the voltage is maximum and the current is minimized.
So I conclude that the permeability of the core will be extremely
high at resonance. Since current is minimized - H (magnetizing force)
will be minimized, and since Voltage is maximized - B (Flux density)
will be maximized.

Thus B/H = mu = huge permeability

I couldn't understand why the MEG was using such high frequencies -
my reasoning was - if you take a look at the data sheet for any core,
the permeability drops as frequency goes up, and this would in turn
switch only a small bit of flux in he MEG.

But since the MEG is at resonance (and high permeability) then it's
switching much more flux than I had previously thought.

Anybody Agree?

Ken










Main page:
  http://groups.yahoo.com/group/MEG_builders


  To post a message to this group, send email to
  MEG_Builders@yahoogroups.com

  To contact the moderator of this group, send email to
  MEG_Builders-owner@yahoogroups.com

  To unsubscribe from this group, send email to
  MEG_Builders-unsubscribe@yahoogroups.com