Some of the results of this experiment can hardly be explained by modern physics, because classical ideas on powered through the diode, the circuit can not be higher voltage than the supply voltage. The experiment is a continuation of a series of experiments on nuclear magnetic resonance (NMR) in the inductor and can be repeated in any home laboratory. Its scheme is shown in figure (1) and includes: generator of short pulses GG1, smoothing capacitors C1 and C2, an isolation diode VD1, two voltmeter DC voltage V1 and V1, the power source U1, and, depending on the variant, THV1 pulse transformer or resonant circuit L1C3.

The operation of the circuit is to periodically connect the load to the inductor (transformer THV1 or the coil L1) key from GG1 to the supply voltage U1. In this case we measure two voltages: the first voltage, which is measured by voltmeter V1, and strore — voltage after the decoupling diode VD1, which is measured by the voltmeter V2. It should be noted that V2 is shunted by two condensers C1 and C2, which are completely smooth all the possible emissions from the operation of the key and its load. At first glance there is nothing unusual: the voltage V2 to just below V1, exactly as much as the passport is the voltage drop across the diode VD1 (depending on the diode it may be 0.4-0.8 In). But if you follow some conditions, the voltage V2 is more than V1, and this excess may reach a value of 1.6 times!

The main condition that the voltage V2 is more than V1, is sufficiently short pulse in the load, with a rather steep front, and recession. It can provide this generator (GG1 scheme), and the optimal location of the contacts SA1 and SA2 in it, found by the author is "1111" and "0001" respectively. The optimal clock frequency may be in the range of 7-10 kHz. Thus, the potentiometer R1 should be withdrawn to its outermost position, providing the minimum resistance, and the switch SW1 is open. All this allows to obtain the minimum pulse width and the optimal ratio between their packs.

The author observed that the pulses longer than 100 NS in this experiment is a very important element GG1 is the output key. From the values of its output capacity, front rise and fall will depend on the parameters of the excess. The best solution from a known author, is the transistor TF27S60 (AOTF27S60); with it, the excess voltage can reach values of 60%. Works worse IRF3205 in which this parameter was 20%. Even weaker results were in IRF840 — 10% or less. The resulting effect is so sensitive to output capacitance key generator GG1 that the accession of a standard oscilloscope probe to its drain can be reduced by 20% or more.

Some additional conditions will be described below.

The experimental setup with the load shown in figure (1a). If the switch contact S1 is open, the circuit operates in the classical manner and no voltage V2 is not observed. To have an effect it is necessary to close the contact, then the voltage V2 will gradually increase. The author, at a voltage of 20 V and a current consumption of 2 mA, the value on the voltmeter V2 reached 32 V (with transistor TF27S60). We used a pulse transformer THV1, which is described in detail here.

You need to add that and other similar transformers on ferrite cores also worked well. It is important to have two windings and the primary must have more turns of wire, and the secondary — several times more. It is also assumed that the wire windings — copper.

The waveform at the drain key generator GG1, as shown in figure (2). There are well visible emissions NMR with a duration of about 50 NS, resembling a self-induced EMF. In this case, the frequency is not so important, because it will depend on the output capacitance of the transistor. In this case it was a little more than 5 MHz. It is known that for the ferrite core with such a permeability maximum operating frequency may not exceed 50-100 kHz. These emissions and determine the effect of voltage, when V2 becomes greater than V1.

The experimental setup with the load in the resonant circuit L1C3 is shown in figure (1b). A flat coilused in this experiment has an inductance of about 25 µh and is made of entirely of copper wire. Despite the fact that the resonant frequency of the circuit about 170 kHz on the waveform we can observe the emission of NMR in the order of 50 NS. Moreover, if the generator GG1 to configure the resonant frequency of this circuit (or its harmonics), then the whole effect disappears. Need some detuning, which is adjusted for the individual circuit.

Waveforms at the drain key generator GG1 shown in figures (3-4). The clock frequency selected there, about 20 kHz for better visibility of the pulses, the more optimal is still of the order of 10 kHz. On the waveform also clearly visible unipolar emission of NMR with a duration of about 50 NS, which determines the effect.

Fig.2. The oscillogram of the NMR emissions across the primary winding of the pulse transformer. Probe 1:10

Fig.3. The oscillogram of the emission of NMR resonant circuit. Scan 500 NS. Probe 1:10

Fig.4. The oscillogram of the emission of NMR resonant circuit. Scan 5 ISS. Probe 1:10

Waveforms obtained at low supply voltage of about 12-15 V. At a higher voltage to them is not possible, because the oscilloscope was switched off. In this case, the current consumption from the power source U1 does not exceed 3-4 mA. Here we must note that at large duty cycle, and in this experiment it is 1000 or more, the pulse current can have values of 3-5 A.

Through numerous experiments established that the effect of exceeding the voltage V2 occurs only if the inductive load is present, the NMR emissions, and stimulating her pulses less than 100 NS in duration. In the case of a classical emission of EMF of self-induction, or of the work resonant circuit, the effect is not observed.

It is also observed that the effect a bit better if decoupling diode VD1 is not installed on the positive supply (as in diagram), and for minus. To explain this result, it is not possible.

In this experiment, may have been able to find one of the ways of removal of the additional energy generated by the coil material in MRI. This can be done, for example, with a threshold element, which will "drain" the excess voltage to the load or samosejka scheme, or, in the case of a transformer, use it strochnoy the winding, periodically loading it on a diode bridge with capacity, and then using the received charge.