The power supply of the ballast capacitor and the separation transformer gained popularity among hobbyists due to small size and the fact that they electrically not connected to the network. However, when designing such devices you must consider several factors to prevent emergency situations in the result of which may be damaged not only the power source, and load. The author of the article, summarizing the experience of creating such devices, recommends, what you should pay attention when designing and building.
In Amateur practice widely used sources with ballast the dividing capacitor and transformer [1-6]. Such a solution allows you designing compact power units. Consider some of the issues the design of such devices for example a low-power source, described in [1] (see figure).
The transformer T1 functions as a separation. He works at low the input and output voltage. Its design is very simple. The Capacitor C1 Is ballast and resistor R2 limits the current pulse on power-up. The voltage at the primary winding of the transformer to limit the Zener diodes VD1 and VD2.
In the resonant circuit consisting of capacitor C1, the primary inductance winding of the transformer L and the primary winding resistance load RH possible resonance that can lead to failure the power source.
Suppose that in the loaded source to the primary winding voltage is equal to 20 In (a typical case). This means that the primary winding the load resistance RH is about 10 times less than the capacitance |XC1| of the capacitor C1, forms a voltage divider 10:1 (approximately), ie |XC1|=10RH.
When properly designed transformer inductive reactance of the primary winding |XL| has about 10 times given to the primary winding of the load resistance RH, so the figure of merit referred to the circuit is extremely low, no resonance can not be.
A completely different situation occurs when the load is switched off (idling). If the above ratio |HS|=10RH and |XL|=10RH, |XC1|=|XL| and resonance occurs. If the input network is to apply voltage 1...2, the primary winding the unloaded transformer it is due to the resonance will increase by 10 times and more the quality of the resulting circuit is large enough, however, when applying network voltage for this rise will not. With the increase of voltage on the coil over nominal (20) the magnetic circuit of the transformer is in saturation, it the inductance is reduced and the circuit ceases to be tuned into resonance.
However, if the transformer is designed with a good margin for valid input the voltage rise can be substantial. This will cause an increase the voltage on the capacitor C1 in comparison with the work in nominal mode, and if the capacitor is selected without reserve - the breakdown can happen. There are other possible no less serious consequences.
Therefore, as for transformerless power supply with ballast capacitor, it is forbidden to operate without rated load. Conventional solution - the connection of the Zener diode to the output of a source or two counter-consistently United Zener (or symmetric) to the primary winding (see the figure).
So the problem is solved for relatively low-power power supplies. For similar powerful devices (very simple get a charger for automobile batteries [2-4]) such measures will not do. Here can be connected in parallel to the primary or secondary winding of the analog symmetric diacs [7, Fig.5,a] or provide relay protection mode idle [3].
Special attention should be paid to the choice of the ballast capacitor the rated voltage. This is the highest voltage between the plates capacitor, whereby he is capable of reliably and continuously operating. For most types of regulated nominal DC voltage. Permissible AC voltage is always less than nominal, for with the exception metalloboranes capacitors MBGC, K42-19, polypropylene C-4 and polyethylene terephthalate K73-17 rated voltage up to 250 V, inclusive, in which these parameters are equal. Therefore, when choosing the type and nominal voltage you must use the reference electric capacitors and remember that the calculation is carried out for the peak value of the AC voltage.
When I connect (or disconnect) the power supply unit to the network in the circuit occurs the transition process, which after some time is replaced the steady-state regime. Without going into the theoretical foundations of transients, note two of the act of switching:
1. The current in the inductor (the device with inductive resistance) can't change abruptly, or, otherwise, the current after switching has the same value it had at the time immediately preceding switching.
2. The voltage across the capacitor cannot change abruptly, or, otherwise, the voltage after the switching has the same meaning as that immediately before commutation.
When the power supply is connected to the network, the capacitor is not charged and the fall voltage across it is zero. The current in the inductor cannot occur instantaneously, therefore, the voltage across the resistor is zero and the supply voltage fully applied to the primary winding of the transformer, which is designed to significantly a smaller value. It occurs when you turn on high risk interturn breakdown and disappears advantage in the simplicity of a transformer winding in bulk, than he has earned wide popularity among radio Amateurs. It is especially dangerous the power supply is connected to the network, which at this point is valid amplitude or something close to this voltage.
It is pertinent to the objective of reducing the voltage across the primary the winding is connected. Current limiting resistor not save you in this in the situation.
This forces us to find a different solution to prevent the possibility interturn breakdown in the transformer and to protect the elements of the power unit from increased tenfold voltage.
The voltage limiter on the two anti-series in parallel the primary winding of the Zener diodes(see figure) allows to solve this problem. For each half-wave of the limiter works as a parametric stabilizer the voltage across the primary winding of the transformer. Ballast function is performed when this basically current limiting resistor R2. The resistor should be calculated on short-time current overload, and the Zener diodes, generally provide it.
If during normal operation the Zener open and work as stabilizers, may be a difference between the amplitudes of the rectified current pulses of positive and negative half-wave. This effect is explained by the fact that positive half stabilizes one Zener diode, and the negative of the other. You know, the voltage stabilization of two copies of the Zener even one party can vary greatly. This imposes an additional component pulsation frequency of 50 Hz, which is harder to suppress smoothing filter than 100 Hz.
To reduce additional component of the ripple arising from differences voltage stabilization, can be recommended instead of counter-consistent connection of two Zener diodes includes at least one Zener diode in the diagonal of the diode bridge parallel to the primary winding. This will keep the reliability of the unit power.
If not there are increased requirements for stability of the output voltage, we can recommend a selection of Zener diodes with a minimum voltage stabilization 1...3 is greater than the maximum amplitude of the voltage on the primary winding in the steady state. Parametric stabilizer in this the event will serve only voltage limiter at the moment of inclusion and at idle. And after the power supply output at steady state it automatically disabled, significantly increasing the efficiency of the unit.
Literature
Author: B. Sadowski, Chelyabinsk