Flux gate generator article

Abstract:
A flux gate generator was constructed and tested out. The results of the measurements are reported. They do not falsify the usual belief in energy conservation. The results are compared with the literature and possible improvements of the generator are discussed.

 1) Introduction
In order to test out the correctness of previous predictions regarding the behaviour of a Brown-Ecklin generator we undertook the task to build up a modified flux generator and to test it out.

 2) Setup
Acc. to a previous proposal [1] we built up a two circuit Brown-Ecklin generator. The revolver drum-like rotor housing of the fluxgate cores was made of plastic and contained four laminated iron cores. One stator side contained 4 holes for 2cm diameter and 2cm length cylindical magnets. They could be filled with stronger neodymium magnets of 1 cm length of 3 weaker neodymium magnets of 0.7cm length. The magnetic circuits were closed at the backside of the magnets by a piece of undefined steal from the stuff. Later we used two pieces 2cm x 2cm x 6cm PERENORM. The other side of the stator contained 2 U shaped laminated transformer cores each with two coils wound on the horseshoe legs. Each coil had about 500 turns (8 layers of wires, 2.2cm coil length each at the first and lowest layer, 0.8mm diameter wire thickness, inner resistance of each coil 7-8 Ohm), i.e. per U core 1000 and in sum (if switched in serie) 2000 turns. The cross section of the horseshoe was 2 x 2 cm2 .The distance between each U legs was 2cm as well. The material ( firm could not be identified) had a saturation of probably about 1,5 T which is typical for such parts used as transformer cores. The driving motor was a LEAR JET capstan motor. Its efficiency was not specified. However, it is known that such motors have typical efficiencies of 30%(long life version) to 50%(high efficiency versions).
As current source of the motor we used a very stable standard electronic power supply in the voltage regulation mode. As load of the generator we used usual electronic resistors switched in serie on a board.
For measurement of input motor voltage and current we used two standard 4.5 digit multimeter. The output voltage and frequency of the generator was measured by HAMEG scopes (20Mhz bandwidth and higher). The frequency was controled additionally by a multimeter containing a frequency counter. Some pictures of the generator are shown in fig.2a - ???. The complete interconnections of all instruments are shown in fig.3.

 Tab.1: list of used materials

iron cores:
material:VACOMAX, firm:VACUUMSCHMELZE, ~200000, saturation 0.7 T, each core 36 lamellas of 0.5mm thickness isolated by grafitti paint from a spray dose, length 5cm of lammellas, diameter of the core~2cm;the iron was baked out 5h under H2 at 1100 degree Celsius for 5 hours

 magnets:
a) strong neodymiums: NE 201 from IBS[2]; energy product E x H =250 kJ/m, remanence Br=1.2T, diameter, 2cm length 1cm
b) weaker neodymiums: NA 002 from IBS[2]; energy product E x H =80 kJ/m, remanence Br=680mT, diameter 2cm, length 0.77cm

 irons to close the flux:
material: PERENORM firm:VACUUMSCHMELZE, saturation 1.5T

 motor:
LEAR JET capstan motor part nr. 357-9102-001 type CDM ID 131039
typical data: 24V DC; without load:108mA, 4640 rpm; with load: 450mA, 3400 rpm

 3) Experiment and results
The experiment was designed in the following way: we used the generator as a tool to measure the efficiency of the capstan motor. We measured the efficiency of the motor acc. to the definition

where netto power input = input with load - input without load. Input loads (DC current!) were calculated acc. to P=U.I where P= power, U=voltage and I=current. In order to include the inner resistance of the coil the output power of the generator was approximated acc. to

If a diode was in the circuit the circuit we used the formula
 

If we get efficiencies significantly higher than expected for these motors then energy conservation in the usual sense would be falsified and overunity of the generator is probable.
After the first playing around we saw that maximum output was achieved if the airgap between the U shaped coil cores and the flux cores was bigger (i.e. 5.6mm)and tight (i.e. 0.5mm) between the magnets and the rotor. All coils were switched in serie. Furthermore, it was important that the left magnets of the circuit had the same polarity and the right magnets the opposite (contrary to fig. 1). We used strong and weaker neodymium magnets, but the measurements were done with the weaker ones because this reduces mechanical losses by vibrations. All measurements were done at constant 1875 rpm (i.e. 31.25Hz) which is equivalent to a AC current of 125Hz from the generator seen at the scope screen shown in fig. 4
The complete copy of the protocol in chronologial order of the relevant measurements can be found in the appendix. Fig.5a-c shows efficiency vs. resistance calculated from our first measurements from 14.6.97 measured with or without a diode in the circuit in each direction, comp.appendix. After this measurements we saw that the value of the input power without load shifted with time due to a slight wandering of the lammellas in the core and other imperfections of construction. Therefore, we calculated worst and best case efficiencies taking the better (higher) value and the more worse (lower) value of input power without load- measured before or after the measurements under load. The deviations of these values were the most biggest factor responsible for the error bars of measurement, therefore we neglected the other factors.
We saw that the efficiency was slightly (significant?) enhanced with diodes but the direction of the diode polarisation seemed to be insignificant. Some efficiency values were higher than 50%.
Because we got best values of higher than usual capstan motor efficiencies we had a closer look to this values and measured alternatively under load and without load at the most promising resistance values, comp. data in appendix from 6.7.97. Now, the efficiencies were about 40% .
Another interesting feature could be seen at decreasing low load resistances, comp. data in appendix from 12.7.97: Although the load decreases and the current rises the power to drive the motor decreases indicating lower back torque of the stator coils at higher current. However, it is clear from these measurements that the power delivered from the coils under this conditions decreases as well.

4) Discussion
Acc. to our generator measurements we found no region of significant overunity efficiency which would falsify the usual belief in energy conservation. However, the AC wave form shows an assymmetry which has been calculated qualitatively previously. Surely the wave form could be made more similar to the form calculated if the distance between the rotor cores (or the diameter of rotation) would be bigger during one revolution. Therefore, we believe that our model ansatz is correct principally although it needs modification to be numerically correct.
However, the output energy values calculated by the theory deviate a order of magnitude from the reality measured here. It is clear that the model presented has the weak point that the model network of magnetic resistances has been assumed with only less backing by a three dimensional field calculation. Furthermore, the non-linear behaviour of the cores is neglegted. Therefore, we see the following possibilities to increase the efficiency of the generator:
1) Possibly the back torque of the horsehoe coils is too high because the saturation of the cores is too high compared with the flux which can go through the rotor cores. If we can calculate typical currents of 40mA and higher in the coils the H-field of the coils is ~10A/cm which means that the iron (typ. saturation values 1A/cm) of the U cores would be in saturation at 1.5T where the coils are.
2) By reducing the length of the rotating iron cores the magnetic dipol moment of the core can be reduced as well. As a consequence the torque exerted on the flux gate cores should decrease as well.
If we compare our results with the known facts and summarize than we have to say that our measurements do not falsify the conservative belief in energy conservation which is contrary to other observations which claim to have measured "negative" incremental efficiencies which is quite contrary to usual energy conservation because these generators accelerate if power is drawn from it. (However, until now, no generator is known which have absolute efficiency = output/input grater than 1.) Some of such observations were made by Marinov [3](recently deceased). Acc.to his considerations and observations it is important to have big coils (i.e. big inductivities in the circuit) to shift the phase of the current in the coil that Lenz's law inverses in effect and the generator becomes self-accelerating. Futhermore, the effect exists only at higher rpm. Some of Marinov constructions can be found in fig.6a-???. Greg Watson [4] means that it is important to have a magnetic design which makes sure that the flux gates are attracted by the coil under current during the closing phase of the magnetic cycle.
Similar observations of negative incremental efficiency has been made by Pete J. Aldo [5] who used another so called SAG-flux gate design proposed by Brown [6].
Therefore, we believe that the problem of the energy balance in electromechanic enngieering is still open for further research.

Acknowledgement: We thank Mr. Thiede for doing the biggest part of the mechanic work.

 Bibliography:
[1] W.D. Bauer The Brown-Ecklin Overunity Generator - A Theoretical Analysis

 [2]Magnetismus - Dauermagnete Werkstoffe und System
Catalog by IBS Magnet Ing.K.H.Schroeter Kurfuerstenstr.92 D-12105 Berlin

 [3] S. Marinov
all references below are self publications of S. Marinov at "East-West Publishers" Graz Austria

 1)The thorny way of truth IV 1991 p.8
The perpetuum mobile "Il nicolino di Veneto" VENETIN COLIU

 2)Deutsche Physik Vol.1 No.1 1992 p.40
The self accelerating generator VENETIN COLIU

 3)Deutsche Physik Vol.2 No.5 1993 p.5
When will the self accelerating generator VENETIN COLIU become a perpetuum mobile ?
4) Deutsche Physik Vol.2 No.7 1993 p.15
The self accelerating generator VENETIN COLIU VI

 5) Deutsche Physik Vol.3 No.10 1994 p.8
The self accelerating generator VENETIN COLIU VII

 6) Deutsche Physik Vol.3 No.10 1994 p.37
The generator VENETIN COLIU VI coupled with a Robert Adams motor

 7) Deutsche Physik Vol.3 No.11 p.35
Discovery of an important additional cause for the anti-Lenz effect in the generator VENETIN COLIU.

 [4] Greg Watson's homepage
description of the DNMEC generator

 [5] Pete J. Aldo, pers. communication

 [6]Brown, Paul The magnetic distributor generator 1982
copy of a report, 1982
entry: Dr. Nieper Gravity Folder
was available from "list of shielding theory of gravity papers" at
Admiral Ruge Archives of biophysics and future science
Keith Brewer Library, Richland Center, Wisc. 53581 USA
 
 

Appendix: Copy of the protocol of our measurements

data from 14.6.97 below: 125 Hz AC, electric circuit acc. to fig.3 .
1.run: without diode

R /Ohm	Umot / V	Imot / mA	Upeak / V
INF	14.557	281	16
396	15	317	14.5
374	15	317	14.5
352	15.001	319	14
330	15.001	316	13.7
308	14.9388	316	13.7
286	14.9388	318	13.5
264	15.0363	323	13
242	15.0728	325	13
220	15.0731	324	12.7
198	15.1044	328	12.3
176	15.1457	335	12
154	15.146	336	11.3
132	15.279	338	10.5
110	15.328	344	9.7
88	15.382	344	8.7
66	15.341	347	7.5
44	15.305	346	5.5

2.run below : with diode in circuit (direction not identified), other things dito

R /Ohm	Umot / V	Imot / mA	Upeak / V
INF	14.34	270	15.3
396	14.55	287	13.3
374	14.55	287	13.1
352	14.513	287	13
330	14.513	288	12.9
308	14.524	291	12.8
286	14.524	290	12.6
264	14.521	292	12.5
242	14.521	292	12.3
220	14.569	294	12
198	14.621	297	11.8
176	14.622	302	11.3
154	14.703	301	10.8
132	14.704	301	10
110	14.745	307	9.5
88	14.746	305	8.8
66	14.85	311	7.3
44	14.9	320	5.7

3. run below: with diode in circuit (direction opposed to last run), other things dito

R /Ohm	Umot / V	Imot / mA	Upeak / V
INF	14.331	276	13.5
396	14.54	290	13.5
374	14.54	291	13.5
352	14.54	288	13.2
330	14.54	290	13
308	14.57	289	13
286	14.62	293	12.6
264	14.62	296	12.4
242	14.62	293	12.3
220	14.62	297	12.1
198	14.71	302	11.8
176	14.71	303	11.4
154	14.71	302	11
132	14.71	307	10.2
110	14.82	311	9.5
88	14.82	311	8.6
66	14.93	319	7.4
44	15.01	325	5.8
INF	14.15	272	15.2

Run from 6.7.97 below: now with PERENORM at the back side of the magnets, other things dito
notations: 0 =no diode, + =diode one direction, - =diode opposite direction;

R /Ohm	Umot / V	Imot / mA	Upeak / V	efficiency/%	diode
INF	14.755	289	17.5	--	0
352	15.041	312	14.5	39-30	+
INF	14.633	287	17.5	--	0
INF	14.66	282	17.5	--	0
352	15.008	310	14.5	33-40	-
INF	14.642	288	17.5	--	0
352	14.953	310	14.5	41-40	+
INF	14.609	288	17.5	--	0
352	14.904	310	14.5	41-40	-
INF	14.629	287	17.5	--	0
INF	14.5	285	17.5	--	0
330	14.87	310	14.5	38-34	+
INF	14.5	284	17.5	--	0
330	14.835	309	14.5	36-35	-
INF	14.445	284	17.5	--	0
330	14.832	309	14.5	35-34	+
INF	14.423	283	17.5	--	0
330	14.817	287	14.5	35-39	-
INF	14.469	287	17.5	--	0
396	14.788	308	15	42	+
INF	14.459	287	17.5	--	0
396	14.781	307	15	44-39	-
INF	14.43	285	17.5	--	0
396	14.76	307	15	40	+
INF	14.42	285	17.5	--	0
396	14.765	307	15	40	-

data from 12.7.97 (no diode) measurement at low resistances

R /Ohm	Umot / V	Imot / mA	Upeak / V	AC/ Hz
INF	14.77	347	16.5	125
22	15.86	423	3.3	125
INF	14.68	337	16.5	125
11	15.71	420	1.75	125
INF	14.77	329	16.5	125
INF	14.73	318	16.5	125
44	15.916	403	5.8	125
22	15.889	408	3	125
11	15.889	400	1.6	126
22	15.889	401	3	125
11	15.889	398	1.6	126

---