Experiment with two lamps and a pulse generator
A very interesting question arose as a result of the experiments with two identical bulbs, the first of which (EL1) is included in the circuit of the power pulse generator GG, and the second (EL2) — connected to its output. The first lamp was included in two ways: directly in the circuit of the power supply (Fig. a), and via a diode bridge rectifier with a capacitor (Fig. b). In the first case, the results of the experiment agreed with the theory on marginal values and the second — the theory has failed and explanation from the author yet.
When you turn on the scheme (Fig. a) it is possible to achieve a ratio of the capacity of dissipation of the second lamp to the first — 9..9,5. EL2 lamp is lit on the third power, and EL1 — 3..4% (its spiral slightly reddish). 9 ratio of 9.5..when one third of the power dissipation, taking into account the nonlinearity of the resistance of the filament, gives the theoretical basis.
When you turn on the scheme (Fig. b) this ratio increases to 16..20; EL2 lamp glows a little brighter than in the first case, and the filament spiral lamp EL1 never seen. Here, the theory is powerless.
As the generator GG used classic poluostrova diagram (see figure below). The chip is generating at its terminals 5 and 7 of rectangular pulses, which are amplified by the output stage MOSFET two-Oh. The scheme is quite reliable and stable in performance, it can be recommended for use in other devices.
Capacitors C1, C5 electroless, which should be designed for a voltage of 400V; C2, C3 is electrolytic, the voltage of 25V. Special attention should be paid to the switching transistors VT1 and VT2 to achieve maximum efficiency, only use a IRF840. For lamps up to 100W can be put on radiators.
Setting the schema is simple. Trimming resistors R3 and R4 need to maximize the ratio of capacity when the lamp EL2 will glow as vividly as possible, and EL1 barely smolder. When you turn on (Fig. b) she did not blush.