Imagine a bizarre construction, almost theatrical in its absurdity: a thin, almost invisible thread stretches to the suspension, to which some device is attached. To it - a heavy, powerful rope, and to it - a massive load. At first glance, everything is clear: the fragile link will be the first to give in, the thread will break, unable to cope with the tension. But here is a surprise. The thin thread holds calmly, without trembling, as if it does not feel all this weight. But the thick rope, seemingly reliable and strong, cracks with a tear and eventually cannot withstand - breaks. Is it possible? The author is not responsible for the comparison with a mechanical analogue, but in radio electronics, as it turns out, such a situation is quite possible.

The experiment is based on previous research of the avalanche mode of bipolar transistors. An inductance L1 (Fig. 10a) is added to it, which increases the efficiency of the device, which is expressed in the almost complete absence of heating of the transistor VT1. This is important for comparing the power released on resistors R2 and R3 in the form of heat.

The circuit allows creating conditions under which resistor R2 (0.125 W) almost does not heat up, and the temperature on R3 (1 W) reaches 100 degrees Celsius. It turns out that the electrical power passing through resistance R2 does not correspond to the thermal power dissipated on it. Why this is very unusual, how to achieve the desired operating mode and what calculations are necessary will be shown later in the experiment.

The circuit in Figure 10a works quite simply. Capacitor C2 is charged via resistor R2 and coil L1, although the latter may, in principle, be absent. When a certain voltage is reached on C2, transistor VT1 enters avalanche mode and, opening, discharges this capacitance onto resistor R3. The inductance only slightly enhances the effect we need by changing the pulse shape, increasing the efficiency of the transistor. Also, in avalanche mode, additional charges are released.

It is important to note that for the composite transistors used here, their base must be open. This is due to the fact that all the necessary resistors are already built into them at the manufacturing stage.

The inductance L1 can be assembled as follows. On a ring of nanocrystalline, measuring 32*20*10 mm, 15 turns of wire are wound. This winding is included in the circuit. But a more effective option is when two turns are wound on top, the terminals of which are shorted together. The author's inductance L1 in this connection was 30 μH. You can also choose other options for this coil, perhaps they will be even more effective.

The elements shown in diagram 10a were selected for the experiment. TIP120 has proven itself as a good transistor VT1, the following oscillograms and power balance will be shown for it. To do this, you need to install this transistor in the circuit and gradually increase the voltage (Up) on it. You can start immediately with 280 volts, gradually increasing it until stable pulses appear on the oscilloscope. This will be the optimal operating mode of the circuit.

On the oscillograms we see the classic charge of the capacitor C2 through the resistor R2 (Fig. 13) and its subsequent dischargeand resistor R3 (Fig. 11-12), with the exception of the initial avalanche surge, which lasts about 10 ns. Inductance L1 slightly smooths out the peak of the discharge pulse.

During the experiment, the power released as heat on resistor R3 was carefully measured and recalculated. The total power consumption of the entire circuit from the power source was 1.7 W, of which 0.8 W was released on resistor R3. To confirm the latter value, an additional thermal test was carried out: a temperature sensor was connected to resistor R3, and its heating rate was compared with the heating rate when connected to a DC voltage source. Thus, the power value of 0.8 W can be considered reliable, as well as the total circuit power of 1.7 W.

In this case, resistor R2 hardly heated up, although according to calculations and logic, which will be presented below, it should dissipate 0.85 W of power and it should burn out immediately, since it was designed for only 0.125 W.

To conduct a comparative balance of the powers dissipated on resistors R2 and R3, we will present circuit 10a as its analogue, but without taking into account the avalanche mode (Fig. 10b). From here on we assume that: \[ R_2 \gg R_3 \tag{1}\] Then, with the open key SW1, the capacitor C2 is charged through the resistor R2. From classical radio engineering we know that no matter what the value of this resistor is, exactly half of the energy that the power source spends on charging the capacitor will always be dissipated on it. And since we calculate everything for one period \(T\), then all this is recalculated for power: \[ P_{R2} = {C_2 U_C^2 \over 2\, T} = {P_U \over 2} \tag{2}\] After opening the switch SW1, the accumulated energy is dissipated on the resistor R3. And we already know that this is half of the energy spent by the power source.

Thus, the power balance in the idealized circuit looks like this: of all the power consumed from the power source, half goes to heating R2, and the other half goes to heating R3. Which also looks quite logical.

In the real circuit, the transistor VT1 also heats up, which reduces the heating of the resistor R3 and reduces its overall efficiency. This is what we observe in the experiment, where slightly less than half of the total power is dissipated on the resistor R3, and about 0.05 W is spent on heating VT1. This parameter also corresponded to the observed effects - the transistor worked without a radiator and was barely noticeably warm.

The idealized model of the experimental circuit used almost completely corresponds to the real circuit, with the exception of the thermal power dissipated on the resistor R2. In an idealized circuit, such a resistor should overheat and burn out, but in a real circuit with an avalanche mode of operation of the key, we do not observe this: the resistor continues its operation and only heats up a little.

Thus, the electrical power passing through the resistance does not correspond to the thermal power dissipated on it. The author suggests that such an effect becomes possible due to the avalanche mode of the transistor, in which a very short, but still sensitive release of electrical charges occurs.