Driving signals

I want to describe this part of the generator because I consider that this is the part which is more difficult to build.  Therefore here I want to make a summary of the different possibilities that I think that it may work in order to excite the electromagnets.

We have to differentiate between the 1902 patent,  no. 30378, and the 1908 patent no. 44267:

  • In the 1902 patent Figuera stated that it just required one single signal “intermittent or alternating current” to excite the machine.
  • In the 1908 patent Figuera described the use of two opposite signals (unphased 180º) to excite the electromagnets.

This is a huge difference and I still do not know if the second patent is just an upgrade or optimization of the initial design or they are just two different machines. Or that in 1902 Figuera did not disclosed all the details, but he did it in 1908, few days before dying, or just that the 1908 is an optimized design fitted to get an AC output while the 1902 design was just getting pulsed electricity to get into a battery. Who knows…

1902 Patent (no. 30378)

In this patent Figuera stated that it was needed an intermitent or alternating current to excite all the electromagnets. In order to create an “intermittent or alternating current” [It is curious to note that Figuera used the word “alternating current”, not the word “alternate current”, something weird… ] there are many options:

  • A rotational mechanical device to pulse the current
  • A relay to pulse the current
  • A VFD (Variable Frequency Drive) to pulse the current
  • Any other electronic circuit design as a PWM (Pulse Width Modulation)
  • For the case of alternating current maybe the AC from the mains or a rectified AC current with a diode bridge
  • Etc…

Figuera offered very few details in his 1902 patents. He did not defined the type of intermittent current or whether it should get to zero voltage or not. That´s why I prefer the 1908 patent because there are more details and supposedly it is an optimization of the original design. Maybe an optimization or maybe just an adjustment to create AC in the output that was getting world spread in those days instead of a pulsed output from the 1902 patent which needed to be accumulated in a battery before its use. Who knows…

It is my belief that Figuera in the 1902 patent may have used some kind of coil interconnections to get a different filter in each coil taking profit of the internal filtering of the inductance and resistance of each coil to unphase his one input signal and create, perhaps, two partially unphased signals in each set of confronted electromagnets. The more coils that are transversed the greater the filter : filter constant = (L1+L2+…+Ln)/(R1+R2+….Rn) . You may see what I mean reading the section C in the chapter below.

1908 Patent (no. 44367)

In this case Figuera defined the use of two opposite signals: when one was at maximum the other was at minimum, when one was increasing the other was decreasing.

Figuera solved this situation with the use of a mechanical commutator based on many intermediate contacts varying the resistance to each set of coils in order to split the current to one or the other set. The problem with the mechanical commutator is that it is very difficult to build it, especially for amateur people. There are some other options to get those two opposite signals:

A) Develop a mechanical commutator as that described in the patent

Some possible designs for the commutator could include the use of a commutator extracted for an old motor, some brushes and slip rings


Easiest Commutator

Other design to assure a robust contact while spinning will require to have all the brushes static:

Or other possible design:

colector y anillo Figuera

Take into account that Figuera represented a simplified commutator design to “make easy the understanding” as he literally wrote. I guess he could have used one commutator and two resistors arrays in parallel to get symmetrical signal to each set of coils.

Signals with Figuera 1908 commutator with 2 resistors

Note that the governing equation for the current in this configuration is the Ohm´s Law applied to each side:

Minimun current: path with highest impedance; current going through regulator and electromagnet.

Maximun current: path with lowest impedance; current circulating just along electromagnet.

B) Using a stepper motor driver to get the two driving signals

Stepper motor drivers, with microstepping,  create two or more 90º unphased signals which can be used to drive the electromagnets. If the output is AC-type maybe a full rectification bridge may be needed to convert it into an always positive modulated DC-type signal. Those drivers are conrolled by an external pulse generator which manage the driver timing and its frequency.

C) Two unphased square signals feeding the coils

If you feed two square signals (in opposition) created with two coupled transistors (in opposition: when one is ON the other is OFF and the contrary) those signals once that reach each set of coils suffer a filter ( filter with time constant = L/R , as consequence of the inductance and resistance of the coils) which convert then into two opposite sawtooth signals, as needed in the patent.

The electronic driving circuit for those signals could be quite simple: a 555 chip (or other) to create a low power pulsed signal (frequency regulator). This signal could go directly to a power transistor to drive one set of coils. This same pulsed signal from the 555 chip also should go to a logical “NOT” gate in order to invert it. This inverted signal should go to a second transistor to create the second square signal in opposition to the first one. See a schematic in this link.

Opposed squared signals

The previous schematic has a null base current (reaching zero voltage during the OFF pulses in the squares waves). The next schematic is an improvement which allows to modulate the base current to different values above zero in order to tune the signals with different frequencies and different maximum and minimum values:



Also, applying this same principle to two half AC waves (obtained from one AC signal with four diodes (see this schematic), one in each set of electromagnets oriented to take half wave in each set of electromagnets) the result is analogous, but here we are forced to use the frequency of the AC network (50 Hz or 60 Hz), while in the previous case with the pulsed signals we could adjust the frequency to get the proper response for our set of value for inductance (L) and resistance (R) and the filter that they create:

Two_half_AC_waves Two_half_AC_waves_filtered_RL

If you want to test this configuration adding a DC bias to move the signal away from zero level, you may use the schematic in this link, or, also this one with a resistor to adjust the DC bias.


Other posible design with DC current plus a bias (offset above zero): Setting one or two batteries in series synchronized with a commutator or an electronic switching element the final voltage can be adjusted to achieve two opposed squares waves well above zero level and without heat looses. See this schematic with three batteries and one commutator, or this other schematic with two batteries and two synchronized commutators.


D) Using the rotor of a motor as commutator

Taking an universal motor and just using the rotor adding two static brushes at opposite sides, each of them connected to each set of electromagnets. The field coils have to be disconnected and the rotor moved with other external motor coupled to its shaft. The rotor coils will act as sequential resistance/inductance sets which will create the two 180º unphased signals will rotating.


E) Using a center tapper transformer and a DC offset to get the two opposite signals

For example some examples are those creating a DC bias (diode bridge + capacitor) and then using an intermediate tap transformer. This circuit just creates two opposite signals with an intermediate tap transformer, and, then adds a DC offset created with a diode bridge and a capacitor. The intermediate tap transformer allows to take out two opposite signals, one in each of its sides, while feeding a DC offset through the intermediate tap. A simple sketch is here:

Offset AC


or even simpler, I have found that the rheostat is not need so you can reduce the heat losses of the original circuit. That rheostat was just creating a voltage drop that you can also manage with a different transformer in the DC offset. You just need to provide a transformer for the DC offset signal with a higher output voltage than the one with the intermediate tap, just taking into consideration that the DC offset voltage must be a little greater than that of the intermediate transformer. For example with 220V input a intermediate transformer (T1) with turn ratio 20:1 and a transformer for the DC bias (T2) with ratio 33:1 is apparently fine in simulation (supposing that in set of coils has 2.3 ohms and 23 mH, two sets used) :


even changing the intermediate tap transformer by two common transformers interconnected:



F) Implement an electronic circuit with transistors to power sequentially each of the resistor described in the patent.

This subject is described in the Patrick Kelly Ebook – Free Energy Devices and Info (version December-2015). The aim is to crate a sequential firing method with transistor to introduce a power signal to the required resistor in each moment. For example the next image is a possible implementation for the driving circuit.




G) Simple circuit to create a DC offset to add to a AC signal

If we could create a DC offset , or bias , and then add a AC we could build the two required signals being always above zero and working in opposition. The DC offset is just to create a base magnetic field.  In one electromagnet the DC offset will add up to the AC signal magnetic field, and in the other they will subtract. When the AC signal reverses its polarity the situation will be with the first subtracting and the second adding. This is:  DC + AC in one , and DC – AC in other.

Figuera_DC_plus_AC_and_DC_minus_AC_two_signals (1)

Even, I think that the DC offset may be substituted by a two permanent magnets. The magnet will create the same magnetic field that the DC signal was creating.

Figuera_Magnet_plus_AC_and_Magnet_minus_AC_two_signals (1)

H) Implement a magnetic amplifier

In order to modulate the two power signals to the electromagnets with a small control signal driven by electronic circuit. Magnetic Amplifiers avoid using resistors and therefore you may skip wasting energy in those elements.

Idea: use a rectified AC signal as input to a magnetic amplifier in order to regulate the output signal with negative feedback, (negative gain) so that an increase in the input will make a decrease in the output (see below)

Some important ideas about mag amps:

“The magnetic amplifier, like the vacuum tube and the transistor, is an electrical control valve where a smaller current controls another circuit´s larger current”

“With a magnetic amplifier you can control AC load current only. For DC applications it is possible to control an AC current and rectify the output”

“Magnetic amplifier control circuits should accept AC input signals as well as DC input signals. The DC input signal is called “bias”. The most effective way to apply bias to a saturable core and also allow AC input signals to control the magnetic amplifier is to use a bias winding”

A few links to magnetic amplifiers theory are: link1 , link2, link3, link4, link5, link6 

A kind of elementary design could be (note: you will need a higher frequency AC signal source because you will try to modulate it with a 50-60 Hz AC signal, therefore you will need a higher frequency in the source signal, maybe 400 Hz or higher, I do not know for sure):

Mag amp for two unphased signals

The schematic is just to show the main idea. It is not a working design because I am not an expert (maybe someone more skillful into mag amps may design a working device…)

I) Building a variable relunctance toroid to split the current (like a variac but for DC signals)

Some users are postulating that a good way to avoid the power losses in the resistors a good option is to build a DC rheostat to alter the current into both sets of electromagnets by varying the reluctance in a toroidal coil.

Basically it is a big toroidal core with one winding with two extremes and, like in a variac, some intermediate taps which are fed back and forth in the way described in the patent with the commutator (or by an electronic switching circuit) , and, that toroidal coil is connected to each set of electromagnets by its extreme taps. I just include here this design in case of being useful because the aim of this webpage is to collect all possible methods to excite the system. But I am not in position to assure if this or other methods are right or wrong.


J) Using a push-pull amplifier

Some users of the forum suggests to user a push-pull amplifier for this task. I just post here this option but I have no idea of those devices, some kind of an audio amplifier or so.

The idea is create two low power opposite signals with a signal generator (there are some for PCs) and then take those signals and amplify them with the amplifier.


K) Forming each electromagnet with a group of coils connected in serie and powering them sequentially

Some users suggest that maybe Figuera was trying to transmist with his 7 sets of electromagnets that he could have used a single electromagnet with 7 coils connected in series with intermediate tap points. Then he could have been applying power to these coils sequentially in each electromagnet: in one electromagnet going from 7,6,5…1 coil powered (decreasing magnetic strength) while at the same time in the other electromagnet increasing magnetic strength by powering 1,2,3…7 coils which modify the number of turns (N) and also the current (i) which transverse the coils. With this implementation you can get rid of the heat losses created by the resistors. A formula for the force of a solenoid/electromagnet is: