How to make a powerful high voltage pulse generator. High-voltage generators with inductive energy storage. Advantages and disadvantages

Rectangular pulse generators are widely used in radio engineering, television, automatic control systems and computer technology.

To obtain rectangular pulses with steep fronts, devices are widely used, the principle of operation of which is based on the use of electronic amplifiers with positive feedback. These devices include the so-called relaxation generators - multivibrators, blocking generators. These generators can operate in one of the following modes: standby, self-oscillating, synchronization and frequency division.

In standby mode, the generator has one stable equilibrium state. An external trigger pulse causes the waiting generator to jump to a new state that is not stable. In this state, called quasi-equilibrium, or temporarily stable, relatively slow processes occur in the oscillator circuit, which eventually lead to a reverse jump, after which a stable initial state is established. The duration of the quasi-equilibrium state, which determines the duration of the generated rectangular pulse, depends on the parameters of the generator circuit. The main requirements for waiting generators are the stability of the duration of the generated pulse and the stability of its initial state. Waiting generators are used primarily to obtain a certain time interval, the beginning and end of which are fixed, respectively, by the front and fall of the generated rectangular pulse, as well as for expanding pulses, for dividing the pulse repetition rate, and for other purposes.

In the self-oscillatory mode, the generator has two states of quasi-equilibrium and does not have a single stable state. In this mode, without any external influence, the generator sequentially jumps from one state of quasi-equilibrium to another. In this case, pulses are generated, the amplitude, duration and repetition frequency of which are determined mainly only by the parameters of the generator. The main requirement for such generators is high frequency stability of self-oscillations. Meanwhile, as a result of changes in supply voltages, replacement and aging of elements, the influence of other factors (temperature, humidity, interference, etc.), the stability of the frequency of self-oscillations of the generator is usually low.

In synchronization or frequency division mode, the frequency of repetition of the generated pulses is determined by the frequency of the external clock voltage (sinusoidal or pulsed) supplied to the generator circuit. The pulse repetition rate is equal to or a multiple of the clock voltage frequency.

A generator of periodically repeating rectangular pulses of a relaxation type is called a multivibrator.

The multivibrator circuit can be implemented both on discrete elements and in an integrated design.

Multivibrator on discrete elements. In such a multivibrator, two amplifying stages covered by feedback are used. One feedback branch is formed by a capacitor and a resistor , and the other and (Fig. 6.16).

states and ensures the generation of periodically repeating pulses, the shape of which is close to rectangular.

In a multivibrator, both transistors can be in active mode for a very short time, since as a result of positive feedback, the circuit jumps into a state where one transistor is open and the other is closed.

Let us assume for definiteness that at the moment of time transistor VT1 open and saturated, and the transistor VT2 closed (Fig. 6.17). Capacitor due to the current flowing in the circuit at previous moments of time, it is charged to a certain voltage. The polarity of this voltage is such that to the base of the transistor VT2 a negative voltage is applied relative to the emitter and VT2 closed. Since one transistor is closed, and the other is open and saturated, the self-excitation condition is not satisfied in the circuit, since the gains of the cascades
.

In this state, two processes take place in the circuit. One process is associated with the flow of the capacitor recharge current from the power supply through the resistor circuit - open transistor VT1 .The second process is due to the charge of the capacitor through a resistor
and the base circuit of the transistor VT1 , as a result, the voltage at the collector of the transistor VT2 increases (Fig. 6.17). Since the resistor included in the base circuit of the transistor has a greater resistance than the collector resistor (
), capacitor charging time less capacitor recharge time .

open, because its base is connected to the positive pole of the power source through a resistor .

Basic
and collector
transistor voltage VT1 while not changing. This state of the circuit is called quasi-stable.

At the point in time as the capacitor recharges, the voltage at the base of the transistor VT2 reaches the opening voltage and the transistor VT2 switches to the active mode of operation, for which
. When opening VT2 collector current increases and decreases accordingly.
. Decrease
causes a decrease in the base current of the transistor VT1 , which in turn leads to a decrease in the collector current . Current reduction accompanied by an increase in the base current of the transistor VT2 because the current flowing through the resistor
, branches off to the base of the transistor VT2 and
.

After the transistor VT1 leaves the saturation mode, the self-excitation condition is fulfilled in the circuit:
. In this case, the process of switching the circuit proceeds like an avalanche and ends when the transistor VT2 goes into saturation mode, and the transistor VT1 - in cut-off mode.

In the future, a practically discharged capacitor (
) is charged from a power source through a resistor circuit
- the base circuit of an open transistor VT2 exponentially with time constant
. As a result, over time
there is an increase in the voltage across the capacitor before
and the front of the collector voltage is formed
transistor VT1 .

Closed state of the transistor VT1 ensured by the fact that initially charged to voltage capacitor through an open transistor VT2 connected to the base-emitter gap of the transistor VT1 , which maintains a negative voltage at its base. Over time, the blocking voltage at the base changes as the capacitor recharged through the resistor circuit - open transistor VT2 . At the point in time transistor base voltage VT1 reaches the value
and it opens.

In the circuit, the self-excitation condition is again satisfied and a regenerative process develops, as a result of which the transistor VT1 goes into saturation mode VT2 closes. Capacitor is charged to a voltage
, and the capacitor almost empty (
). This corresponds to the time , from which the consideration of processes in the scheme began. On this, the full cycle of the multivibrator operation ends, since in the future the processes in the circuit are repeated.

As follows from the timing diagram (Fig. 6.17), in a multivibrator, periodically repeating rectangular pulses can be removed from the collectors of both transistors. In the case when the load is connected to the collector of the transistor VT2 , pulse duration determined by the process of recharging the capacitor , and the duration of the pause - the process of recharging the capacitor .

Capacitor recharge circuit contains one reactive element, so , where
;
;.

Thus, .

Recharge process ends at time , when
. Therefore, the duration of the positive pulse of the collector voltage of the transistor VT2 is determined by the formula:

.

In the case when the multivibrator is made on germanium transistors, the formula is simplified, since
.

Capacitor recharge process , which determines the length of the pause between transistor collector voltage pulses VT2 , proceeds in the same equivalent circuit and under the same conditions as the process of recharging the capacitor , only with a different time constant:
. Therefore, the formula for calculating similar to the formula for calculating :

.

Usually, in a multivibrator, the pulse duration and pause duration are adjusted by changing the resistance of the resistors and .

The duration of the fronts depends on the opening time of the transistors and is determined by the charge time of the capacitor through the collector resistor of the same shoulder
. When calculating a multivibrator, it is necessary to fulfill the condition of saturation of an open transistor
. For transistor VT2 without current
capacitor recharge current
. Therefore, for a transistor VT1 saturation condition
, and for the transistor VT2 -
.

Frequency of generated pulses
. The main obstacle to increasing the pulse generation frequency is the long duration of the pulse front. Reducing the duration of the front of the pulse by reducing the resistance of the collector resistors can lead to non-fulfillment of the saturation condition.

With a high degree of saturation in the considered multivibrator circuit, there may be cases when, after switching on, both transistors are saturated and there are no oscillations. This corresponds to the hard regime of self-excitation. To prevent this, you should choose an open transistor operating mode near the saturation limit in order to maintain sufficient gain in the feedback circuit, and also use special multivibrator circuits.

If the pulse duration equal to the duration , which is usually achieved at , then such a multivibrator is called symmetrical.

The duration of the front of the pulses generated by the multivibrator can be significantly reduced if diodes are additionally introduced into the circuit (Fig. 6.18).

When, for example, the transistor closes VT2 and the collector voltage begins to increase, then to the diode VD2 a reverse voltage is applied, it closes and thereby disconnects the charging capacitor from the collector of the transistor VT2 . As a result, the capacitor charge current no longer flows through the resistor , and through a resistor . Therefore, the duration of the front of the collector voltage pulse
now determined only by the process of closing the transistor VT2 . Diode works the same way. VD1 when the capacitor is charged .

Although in such a circuit the front duration is significantly reduced, the charge time of the capacitors, which limits the duty cycle of the pulses, remains practically unchanged. Time constants
and
cannot be reduced by lowering . Resistor in the open state of the transistor through an open diode is connected in parallel with the resistor .As a result, when
the power consumed by the circuit increases.

Multivibrator on integrated circuits(Fig. 6.19). The simplest circuit contains two inverting logic elements LE1 and LE2, two timing chains
and
and diodes VD1 , VD2 .

At the point in time
and at the exit LE2
. As a result, the input LE1 through a capacitor , which is charged to a voltage
, voltage is applied and LE1 goes to zero
. Since the output voltage LE1 decreased, then the capacitor starts to disintegrate. As a result, the resistor a negative polarity voltage will appear, the diode will open VD2 and capacitor quickly discharge to voltage
. After the end of this process, the input voltage LE2
.

At the same time, the process of charging the capacitor takes place in the circuit and over time, the input voltage LE1 decreases. When at a moment in time voltage
,
,
. Processes begin to repeat themselves. The capacitor is charging again. , and the capacitor discharged through an open diode VD1 . Since the resistance of an open diode is much less than the resistance of resistors , and , capacitor discharge and going faster than their charge.

Input voltage LE1 in the time interval
determined by the process of charging the capacitor :, where
;
is the output resistance of the logic element in the unity state;
;
, where
. When
, the formation of a pulse at the output of the element ends LE2, hence the pulse duration

.

The duration of the pause between pulses (time interval from before ) is determined by the process of charging the capacitor , That's why

.

The duration of the front of the generated pulses is determined by the switching time of the logic elements.

On the timing diagram (Fig. 6.20), the amplitude of the output pulses does not change:
, since the output impedance of the logic element was not taken into account in its construction. Given the finiteness of this output resistance, the amplitude of the pulses will change.

The disadvantage of the considered simplest multivibrator circuit based on logic elements is the hard mode of self-excitation and the possible absence of an oscillatory mode of operation associated with this. This disadvantage of the circuit can be eliminated if an additional logic element AND is introduced (Fig. 6.21).

When the multivibrator generates pulses, then the output LE3
, insofar as
. However, due to the hard mode of self-excitation, such a case is possible when, when the power supply voltage is turned on, due to the low rate of voltage rise, the capacitor charge current and turns out to be small. In this case, the voltage drop across the resistors and may be less than the threshold
and both elements LE1 and LE2) will be in a state where the voltages at their outputs
. With this combination of input signals at the output of the element LE3 there will be tension
, which through a resistor applied to the input of the element LE2. As
, then LE2 is transferred to the zero state and the circuit begins to generate pulses.

To build rectangular pulse generators, along with discrete elements and integrated circuits, operational amplifiers are used.

so the voltage at the inverting input
depends not only on the voltage at the output of the amplifier, but is also a function of time, since
.

A positive voltage is applied to the non-inverting input
. Voltage
remains constant, and the voltage at the inverting input
increases over time, tending to the level
, since the process of recharging the capacitor takes place in the circuit .

However, for now
, the state of the amplifier determines the voltage at the non-inverting input and the output remains at the level
.

At the point in time the voltages at the inputs of the operational amplifier become equal:
. Further slight increase
leads to the fact that the differential (difference) voltage at the inverting input of the amplifier
turns out to be positive, so the output voltage decreases sharply and becomes negative
. Since the voltage at the output of the operational amplifier has changed polarity, the capacitor subsequently recharges and the voltage on it, as well as the voltage at the inverting input, tend to
.

At the point in time again
and then the differential (difference) voltage at the input of the amplifier
becomes negative. Since it acts on the inverting input, the voltage at the output of the amplifier abruptly again takes on the value
. The voltage at the non-inverting input also jumps
. Capacitor , which by the time charged to a negative voltage, recharges again and the voltage at the inverting input increases, tending to
. Since at the same time
, then the voltage at the output of the amplifier remains constant. As follows from the timing diagram (Fig. 6.23), at the time the full cycle of the circuit operation ends and in the future the processes in it are repeated. Thus, at the output of the circuit, periodically repeating rectangular pulses are generated, the amplitude of which at
is equal to
. Pulse duration (time interval
) is determined by the capacitor recharge time according to the exponential law from
before
with time constant
, where
is the output impedance of the operational amplifier. Because during the pause (interval
) the capacitor is recharged under exactly the same conditions as during the formation of pulses, then
. Therefore, the circuit operates as a symmetrical multivibrator.

happens with time constant
. With a negative output voltage (
) open diode VD2 and the capacitor recharge time constant , which determines the duration of the pause,
.

The standby multivibrator or single vibrator has one stable state and provides the generation of rectangular pulses when short trigger pulses are applied to the input of the circuit.

Single vibrator on discrete elements consists of two amplifying stages covered by positive feedback (Fig. 6.25).

transistor VT1 depends on the collector current of the transistor VT2 . Such a circuit is called an emitter-coupled single vibrator. The circuit parameters are calculated in such a way that in the initial state, in the absence of input pulses, the transistor VT2 was open and saturated, and VT1 was in cutoff mode. Such a state of the circuit, which is stable, is ensured when the following conditions are met:
.

Let us assume that the one-shot is in a stable state. Then the currents and voltages in the circuit will be constant. transistor base VT2 through a resistor connected to the positive pole of the power supply, which in principle ensures the open state of the transistor. To calculate the collector
and basic currents, we have a system of equations

.

Determining from here the currents
and , we write the saturation condition in the form:

.

Considering that
and
, then the resulting expression is significantly simplified:
.

On a resistor due to the flow of currents ,
voltage drop is generated
. As a result, the potential difference between the base and emitter of the transistor VT1 is defined by the expression:

If the scheme satisfies the condition
, then the transistor VT1 closed. Capacitor while charged to voltage. The polarity of the voltage across the capacitor is shown in fig. 6.25.

When the transistor VT1 open, capacitor is connected to the base-emitter region of the transistor VT2 so that the base potential becomes negative and the transistor VT2 goes into cutoff mode. The switching process of the circuit is of an avalanche-like nature, since at this time the condition of self-excitation is fulfilled in the circuit. The switching time of the circuit is determined by the duration of the processes of switching on the transistor VT1 and turn off the transistor VT2 and is fractions of a microsecond.

When the transistor closes VT2 through a resistor collector and base currents stop flowing VT2 . As a result, the transistor VT1 remains open even after the end of the input pulse. At this time, the resistor voltage drops
.

The state of the circuit when the transistor VT1 open and VT2 closed, is quasi-stable. Capacitor through a resistor , open transistor VT1 and resistor is connected to the power source in such a way that the voltage on it has the opposite polarity. The capacitor recharge current flows in the circuit , and the voltage on it, and therefore on the base of the transistor VT2 tends to a positive level.

Voltage change
is exponential: where
. Initial voltage at the base of the transistor VT2 determined by the voltage to which the capacitor is initially charged and residual voltage on the open transistor:

The voltage limit to which the voltage at the base of the transistor tends VT2 , .

Here it is taken into account that through the resistor not only the current of recharging the capacitor flows , but also the current open transistor VT1 . Hence, .

At the point in time voltage
reaches trigger voltage
and transistor VT2 opens. Appeared collector current creates an additional voltage drop across the resistor , which leads to a decrease in voltage
. This causes a decrease in the base and collector currents and a corresponding increase in voltage
. Positive increment of transistor collector voltage VT1 through a capacitor transferred to the base circuit of the transistor VT2 and contributes to an even greater increase in its collector current . The circuit again develops a regenerative process, ending with the fact that the transistor VT1 closes, and the transistor VT2 goes into saturation mode. This completes the impulse generation process. The pulse duration is determined by putting
: .

After the end of the pulse, the process of charging the capacitor proceeds in the circuit through a circuit of resistors
,and emitter circuit of an open transistor VT2 . At the initial moment, the base current transistor VT2 equal to the sum of the capacitor charge currents : current , limited by the resistance of the resistor
, and the current flowing through the resistor . As the capacitor charges current decreases and, accordingly, the base current of the transistor decreases VT2 tending to a stationary value determined by the resistor . As a result, at the moment of opening the transistor VT2 voltage drop across the resistor turns out to be greater than the stationary value, which leads to an increase in the negative voltage at the base of the transistor VT1 . When the voltage across the capacitor reaches
the circuit returns to its original state. The duration of the process of recharging the capacitor , which is called the recovery stage, is determined by the relation .

Minimum repetition period of single vibrator pulses
, and the maximum frequency
. If the interval between input pulses is less than , then the capacitor will not have time to recharge and this will lead to a change in the duration of the generated pulses.

The amplitude of the generated pulses is determined by the voltage difference across the collector of the transistor VT2 in closed and open states.

A single vibrator can be implemented on the basis of a multivibrator if one feedback branch is made not capacitive, but resistor and a voltage source is introduced
(Fig. 6.27). Such a circuit is called a single vibrator with collector-base connections.

To the base of the transistor VT2 a negative voltage is applied and it is closed. Capacitor charged to voltage
. In the case of germanium transistors
.

Capacitor , acting as a boost capacitor, charged to a voltage
. This state of the circuit is stable.

When applied to the base of the transistor VT2 unlocking pulse (Fig. 6.28) in the circuit, the processes of opening the transistor begin to proceed VT2 and closing the transistor VT1 .

When switching the circuit, the front of the output pulse is formed, which is usually removed from the collector of the transistor VT1 . In the future, the process of recharging the capacitor takes place in the circuit .Voltage on it
, and hence the voltage at the base transistor VT1 changes exponentially
,where
.

When at a moment in time the base voltage reaches the value
, transistor VT1 opens, the voltage on its collector
decreases and closes the transistor VT2 . In this case, a cutoff of the output pulse is formed. The pulse duration is obtained by putting
:

.

As
, then . Cut duration
.

Subsequently, the capacitor charge current flows in the circuit through a resistor
and the base circuit of an open transistor VT1 . The duration of this process, which determines the recovery time of the circuit,
.

The amplitude of the output pulses in such a one-shot circuit is almost equal to the voltage of the power source.

Single vibrator on logical elements. To implement a one-shot on logical elements, NAND elements are usually used. The block diagram of such a single vibrator includes two elements ( LE1 and LE2) and a timing chain
(Fig. 6.29). Inputs LE2 combined and it works like an inverter. Output LE2 connected to one of the inputs LE1, and a control signal is applied to its other input.

its input circuit. The circuit generates a square wave at a short-term decrease (time ) input voltage
. After a time interval equal to
(not shown in Figure 6.29), at the output LE1 the voltage will increase. This voltage jump through the capacitor passed to the input LE2. Element LE2 switches to state "0". Thus, at the input 1 LE1 after a time interval
tension starts
and this element will remain in the state of one, even if after the expiration of time
voltage
will again become equal to the logical "1". For normal operation of the circuit, it is necessary that the duration of the input pulse
.

As the capacitor charges output current LE1 decreases. Accordingly, the voltage drop across :
. At the same time, the voltage increases
aiming for tension
, which when switching LE1 to state "1" was less
due to the voltage drop across the output resistance LE1. This circuit state is temporarily stable.

At the point in time voltage
reaches the threshold
and element LE2 switches to state "1". To input 1 LE1 a signal is given
and it switches to the log state. "0". At the same time, the capacitor , which is in the time interval from before charged, begins to discharge through the output resistance LE1 and diode VD1 . After the time has passed , determined by the process of discharging the capacitor , the circuit returns to its original state.

Thus, at the output LE2 a rectangular pulse is generated. Its duration, depending on the time of decrease
before
, is determined by the relation
, where
- output impedance LE1 in state "1". Circuit recovery time , where
- output impedance LE1 in state "0"; - internal resistance of the diode in the open state.

and the voltage at the inverting input is small:
, where
voltage drop across the diode in the open state. At the non-inverting input, the voltage is also constant:
, and since
, then the output voltage is maintained constant
.

When applied at the time input pulse of positive polarity with amplitude
the voltage at the non-inverting input becomes greater than the voltage at the inverting input and the output voltage jumps to
. In this case, the voltage at the non-inverting input also increases abruptly to
. Simultaneously diode VD closed, condenser begins to charge and a positive voltage rises at the inverting input (Fig. 6.32). Till
voltage is maintained at the output
. At the point in time at
there is a change in the polarity of the output voltage and the voltage at the non-inverting input takes its original value, and the voltage begins to decrease as the capacitor discharges .

As
, then
.

The recovery time of the circuit is determined by the duration of the capacitor discharge process from
before
and taking into account the accepted assumptions
.

Oscillators on operational amplifiers provide the formation of pulses with an amplitude of up to tens of volts; the duration of the fronts depends on the bandwidth of the operational amplifier and can be fractions of a microsecond.

A blocking oscillator is a relaxation-type pulse generator in the form of a single-stage amplifier with positive feedback created using a transformer. The blocking generator can operate in standby and self-oscillating modes.

Standby operation blocking-generator. When operating in standby mode, the circuit has a single steady state and generates square wave pulses when a trigger pulse is applied to the input. The steady state of the blocking generator on a germanium transistor is carried out by including a bias source in the base circuit. When using a silicon transistor, a bias source is not required, since the transistor is closed at zero voltage at the base (Fig. 6.33).

base voltage decreases. Such a connection is realized by appropriately connecting the beginning of the transformer windings (in Fig. 6.33, shown by dots).

In most cases, the transformer has a third (load) winding to which the load is connected. .

The voltages on the transformer windings and the currents flowing in them are interconnected as follows:
,
,
,
where
,
– transformation coefficients;
- the number of turns of the primary, secondary and load windings, respectively.

The duration of the process of turning on the transistor is so short that during this time the magnetization current practically does not increase (
). Therefore, the equation of currents in the analysis of the transition process of turning on the transistor is simplified:
.

base current
and the actual current flowing in the base circuit of the transistor,
.

Thus, the initial change in the base current
as a result of the processes occurring in the circuit, leads to a further change in this current
, and if
, then the process of changing currents and voltages is avalanche-like. Therefore, the condition for self-excitation of the blocking generator:
.

In the absence of load (
) this condition is simplified:
. As
, then the self-excitation condition in the blocking generator is satisfied quite easily.

The process of opening the transistor, accompanied by the formation of the front of the pulse, ends when it goes into saturation mode. In this case, the condition of self-excitation ceases to be satisfied and, subsequently, the top of the pulse is formed. Since the transistor is saturated:
, then a voltage is applied to the primary winding of the transformer
and reduced base current
, as well as the load current
, turn out to be constant. The magnetization current during the formation of the pulse top can be determined from the equation
, whence, under zero initial conditions, we obtain
.

Thus, the magnetization current in the blocking generator, when the transistor is saturated, increases in time according to a linear law. In accordance with the current equation, the collector current of the transistor also increases linearly
.

As time passes, the degree of saturation of the transistor decreases as the base current remains constant.
, and the collector current increases. At some point in time, the collector current increases so much that the transistor switches from saturation to active mode and the condition for self-excitation of the blocking generator begins to be satisfied again. Obviously, the duration of the pulse top is determined by the time during which the transistor is in saturation mode. The saturation mode boundary corresponds to the condition
. Hence,
.

From here we get the formula for calculating the duration of the top of the pulse:

.

Magnetizing current
during the formation of the top of the pulse increases and at the end of this process, i.e. at
, reaches the value
.

Since the voltage of the power source is applied to the primary winding of the pulse transformer during the formation of the peak of the pulse , then the amplitude of the pulse on the load
.

When the transistor switches to active mode, the collector current decreases
. A voltage is induced in the secondary winding, resulting in a decrease in the base voltage and current, which in turn causes a further decrease in the collector current. A regenerative process develops in the circuit, as a result of which the transistor switches to cutoff mode and a pulse cutoff is formed.

The avalanche-like process of closing the transistor has such a short duration that the magnetization current during this time practically does not change and remains equal
. Therefore, by the time the transistor closes in the inductance stored energy
. This energy is only dissipated in the load , since the collector and base circuits of a closed transistor are open. In this case, the magnetizing current decreases exponentially:
, where
is the time constant. flowing through resistor current creates a reverse voltage surge on it, the amplitude of which
, which is also accompanied by a voltage surge at the base and collector of a closed transistor
. Using the previously found relation for
, we get:

,

.

The process of dissipation of energy stored in a pulse transformer, which determines the recovery time of the circuit , ends after a time interval
, after which the circuit returns to the initial state. Additional surge of collector voltage
may be significant. Therefore, in the blocking generator circuit, measures are taken to reduce the value
, for which a damping circuit consisting of a diode is included in parallel with the load or in the primary winding VD1 and resistor , whose resistance
(Fig. 6.33). When the pulse is formed, the diode is closed, since a voltage of reverse polarity is applied to it, and the damping circuit does not affect the processes in the circuit. When a voltage surge occurs in the primary winding when the transistor closes, a forward voltage is applied to the diode, it opens and current flows through the resistor . As
, then the surge of the collector voltage
and reverse voltage surge on are significantly reduced. However, this increases the recovery time:
.

Not always a resistor is connected in series with the diode , and then the burst amplitude is minimal, but its duration increases.

impulses. We will consider the processes occurring in the scheme, starting from the moment of time when the voltage across the capacitor reaches the value
and the transistor will open (Fig. 6.36).

Since the voltage on the secondary (base) winding remains constant during the formation of the top of the pulse
, then as the capacitor charges, the base current decreases exponentially
, where
is the resistance of the base-emitter region of a saturated transistor;
is the time constant.

In accordance with the current equation, the collector current of the transistor is determined by the expression
.

It follows from the above relations that in the self-oscillatory blocking oscillator, during the formation of the pulse top, both the base and collector currents change. As you can see, the base current decreases over time. The collector current, in principle, can both increase and decrease. It all depends on the relationship between the first two terms of the last expression. But even if the collector current decreases, it is slower than the base current. Therefore, when the base current of the transistor decreases, the time comes , when the transistor leaves the saturation mode and the process of forming the top of the pulse ends. Thus, the duration of the pulse top is determined by the relation
. Then we can write the equation of currents for the moment when the formation of the pulse top is completed:

.

After some transformations we have
. The resulting transcendental equation can be simplified under the condition
. Using the series expansion of the exponential and restricting ourselves to the first two terms
, we obtain a formula for calculating the duration of the top of the pulse
, where
.

During the formation of the top of the pulse due to the flow of the base current of the transistor, the voltage across the capacitor changes and by the time the transistor closes, it becomes equal to
. Substituting into this expression the value
and integrating, we get:

.

When the transistor switches to the active mode of operation, the self-excitation condition begins to be satisfied again and an avalanche-like process of its closing takes place in the circuit. As in the waiting blocking generator, after closing the transistor, the process of dissipation of the energy stored in the transformer takes place, accompanied by the appearance of surges in the collector and base voltages. After the end of this process, the transistor continues to be in the closed state due to the fact that a negative voltage of the charged capacitor is applied to the base . This voltage does not remain constant, because in the closed state of the transistor through the capacitor and resistor recharging current flows from the power supply . Therefore, as the capacitor recharges the voltage at the base of the transistor increases exponentially
, where
.

When the base voltage reaches
, the transistor opens and the process of pulse formation begins again. Thus, the duration of the pause , determined by the time the transistor is in the off state, can be calculated if we put
. Then we get
.For a blocking oscillator based on a germanium transistor, the resulting formula is simplified, since
.

Blocking generators have a high efficiency, since practically no current is consumed from the power source in the pause between pulses. Compared to multivibrators and single vibrators, they allow you to get a larger duty cycle and a shorter pulse duration. An important advantage of blocking generators is the possibility of obtaining pulses whose amplitude is greater than the power supply voltage. To do this, it is enough that the transformation ratio of the third (load) winding
. In a blocking generator, in the presence of several load windings, it is possible to carry out galvanic isolation between the loads and receive pulses of different polarity.

The blocking generator circuit is not implemented in an integrated design due to the presence of a pulse transformer.

Scheme and theories of action

As shown in fig. 3.2, a current limiting transformer T1 is connected to a bridge rectifier D1-D4 and charges an external storage device - capacitor C through an overvoltage protection resistor R18. An external storage capacitor is connected between the discharge ground and the spark gap electrode G1. The load in this project is not included as standard, but between the discharge ground and the spark gap electrode G2. Note that the load is complex, usually highly inductive (not in all cases) with little resistance from the Load inductor wire. The electrodes of the spark gap G1 and G2 are located at a distance greater than 1.2-1.5 times the breakdown distance at a given voltage.

The third trigger electrode TE1 is discharged with a short, low energy, high voltage pulse at G2, creating a voltage peak that ionizes

Rice. 3.2. Schematic diagram of a pulse generator

Note:

Special note regarding diodes D14, D15. The polarity can be reversed to produce a larger trigger effect with a low impedance load, as is the case with a can warp device, wire blaster, plasma weapon, etc.

Attention! If the load impedance is too high, energy can be sent back through the diodes and transformer T2 and cause the failure of these components.

Note that the circuit ground and common wire are isolated from each other.

The discharge ground is connected to the chassis and ground through the green wire of the power cord.

To ensure greater security, it is recommended to use push-buttons as switch S3, which is switched on only when pressed.

If the device is located in a place where unauthorized personnel have access, it is recommended to use a key switch as S4.

a gap between G1 and G2, which leads to the discharge of the energy stored in the external capacitive storage into the load with complex resistance.

The charge voltage of the external capacitive storage is set by the resistor divider circuit R17, which also outputs a signal for the voltmeter Ml. The charge voltage is set in series with R17 control variable resistance R8. This control signal sets the off level of comparator II, which sets the DC bias of transistor Q1. In turn, Q1 controls the relay, which turns the relay off. The contacts of the de-energized relay RE1 remove the energy supply to the primary winding T1. When R8 is set to a predetermined value, it automatically maintains a certain voltage level in external capacitive storage. The safety button S3 provides the possibility to delay the charging of the external capacitor manually.

The red LED LA1 lights up when the power is turned on. The yellow LED LA2 lights up when the charge reaches the set value.

The triggering electrode circuit is a special capacitive discharge (CD) system, where the energy of the capacitor C6 is directed to the primary winding of the pulse transformer T2. On the secondary winding T2, a sequence of positive high voltage pulses is generated, which is fed to capacitors C8 and C9 through decoupling diodes D14 and D15. These high voltage direct current pulses induce ionization in the gaps by discharging through the trigger electrode TE1. At the input of this circuit is a voltage doubler, consisting of capacitors C4, C5 and diodes D8 and D9. Start switch S1 energizes the circuit, causing the spark gap to fire immediately. The silicon triode thyristor SCR removes the charge from C6, the unlocking current to the SCR supplies the dinistor DIAC, the bias to which is set by the variable resistance R14 and the capacitor C7.

A step-down voltage transformer 12 V TZ feeds the control circuit, which also includes relay RE1. If the system does not have 12V, it can only be started by activating RE1 manually. The rectifier on diodes D10-D13 rectifies an alternating voltage of 12 V, which is then filtered by a capacitive filter C1. Resistor R5 decouples power for control through the zener diode Z3, Z4, which is necessary for the stable operation of the comparator circuit. The energy storage is powered by 115 V AC, with fuse F1 activated, and the 115 V AC power is switched on by switch S4.

Comment

In our lab at Information Unlimited, the energy storage equipment includes 10 racks of oil capacitors. Each rack holds 50 32µF 4500V capacitors connected in parallel to achieve a total capacitance of 1600µF or about 13000J at 4000V per rack. All 10 racks connected in parallel give 130,000 joules. It is very important at these energy levels to properly connect and assemble the system with the necessary arrangement and thickness of wires to obtain pulses of hundreds of megawatts. To protect personnel from hazardous voltage, anti-explosion shields are installed around the storage racks.

The charge time for one rack is about 10 minutes. With such a charge, it would be impractical to use 10 racks, since it would take almost 2 hours to charge them. We use a 10,000 V, 1 A current charging system that allows all 10 racks of oil capacitors to be charged to store an energy of 130,000 J for 1 minute . Such a high-voltage charger can be purchased by special order.

Device pre-assembly procedure

This section assumes that you are familiar with the basic tools and have sufficient assembly experience. The pulse generator is assembled on a metal chassis 25.4 × 43.2 × 3.8 cm, made of galvanized iron 1.54 mm thick (22 gauge). It uses a 6500V, 20mA current limited RMS transformer. It is necessary to follow the given drawing as closely as possible. You can use a more powerful transformer, then you will have to change the size of the device. We propose to connect in parallel up to 4 previously used transformers; to get a charging current of 80mA. A voltmeter and controls are mounted on the front panel. It is recommended to replace S4 with a key switch if the unit is located in a location where unauthorized personnel have access.

When assembling the device, observe the following sequence of actions:

1. If you purchased a kit, lay out and identify all components and structural parts.

2. Cut out a board with a 0.25 cm grid perforation and dimensions of 15.9×10.8 cm (6.25×4.25 inches) from the blank.

Rice. 3.3. Pulse Generator Circuit Board

Note:

The dotted line shows the connections on the back of the board. Large black dots show holes in the board that are used to install components and connections between them.

3. Insert elements as shown in fig. 3.3, and solder them to the leads of the elements, to those contact pads, where necessary, as you move from the lower left edge to the right. The dotted line shows the wire connections on the back of the board as shown in the circuit diagram. Avoid wire bridges, potential shorts, and cold soldering as these will inevitably cause problems. Solder joints should be shiny and smooth, but not ball-shaped.

4. Connect the circuit board with wires to the following points (see figure 3.3):

– to chassis ground with #18 vinyl insulated wire 20 cm long;

– with TE1 high voltage wire 20 kV, 10 cm long;

- with R18 resistor, #18 vinyl insulated wire, 20 cm long;

– with anodes D3 and D4 with 30 cm #18 vinyl-insulated wire (circuit ground);

- with TK (2) 12 V DC wire in vinyl insulation #22 20 cm long;

– with M1 voltmeter (2) 20 cm long #22 vinyl insulated wire. Check all connections, components, location of all diodes, semiconductors, electrolytic capacitors CI, C2, C4, C5, C7. Check solder quality, potential short circuits, cold solder spots. Solder joints should be smooth and shiny, but not ball-shaped. Check this carefully before turning on the device.

5. The assembly of the spark gap is carried out as follows (Fig. 3.4):

– fabricate the BASE1 base from 1.4 mm thick (20 gauge) galvanized iron sheet with dimensions of 11.4 x 5 cm (4.75 x 2 inches);

– Fabricate two brackets BRKT1 from 1.4 mm thick galvanized iron sheet (20 gauge) measuring 6.4 x 3.2 cm (2.5 x 1.25 inches) each. Bend the edge in the form of a visor measuring 1.9 cm;

– Make two BLK1 blocks of polyvinyl chloride (PVC) or similar material 1.9 cm thick and 2.5×3.2 cm (1×1.25 inches). They must have good insulating properties;

- Make a block BLK2 from Teflon. It must withstand a high voltage trigger pulse;

– Carefully solder the COL1 flanges to the BRK1 brackets. Adjust the armature to ensure accurate alignment of the tungsten electrodes after assembly of the device. At this point, you will need to use a propane gas blowtorch, etc.;

- grind off the sharp ends from the eight screws. This is necessary to prevent breakage of the PVC material due to corona discharge formed at sharp ends at high voltage;

- pre-assemble the parts, drill the necessary holes in them for assembly. For correct placement, follow the drawing;

Rice. 3.4. Spark gap and ignition device

Note:

The spark gap is the heart of the system, and it is there that the energy stored by the capacitors over the entire charge period a is quickly released into the load in the form of a high power pulse. It is very important that all connections are capable of withstanding high currents and high discharge voltages.

The instrument shown here is designed for the HEP90 and is capable of switching at up to 3000 joules (with a properly adjusted pulse), which is usually sufficient for effective experimentation with mass transfer devices, can bending, wire blasting, magnetism, and other similar projects.

By special order, a high energy switch capable of operating with 20000J of energy can be supplied. Both switches use a high voltage trigger pulse that depends on the high line load impedance. This is not usually a problem for moderately inductive loads, but can be a problem with low inductance loads. This problem can be solved by placing several ferrite or ring cores in these lines. The cores react very strongly to the triggering pulse, but they reach saturation during the main discharge.

The design of the spark gap must take into account the mechanical forces that result from strong magnetic fields. This is very important when working with fjj energy and will require additional funds to reduce the inductance and resistance.

Attention! When conducting experiments, a screen should be installed around the device to protect the operator from possible fragments if the device breaks down.

For reliable starting, the starting gap must be set depending on the charge voltage a. The gap must be located at least 0.6 cm from the bracket. If the inclusion is unstable, you need to experiment with this value.

– Attach the large block lugs LUG1 to each side of the BRKT1 brackets. The connection must be made carefully, since the impulse current reaches kiloamperes;

– temporarily set the main gap to 0.16 cm and the trigger gap to 0.32 cm.

The order of the final assembly of the device

Below are the final assembly steps:

1. Fabricate the chassis and panel as shown in fig. 3.5. It would be wise to cut a square hole in the panel to accommodate the voltmeter before fabricating the panel. The voltmeter that is used requires a 10 cm square hole. Other, smaller holes can be identified from the drawing and drilled after joining the chassis and panel.

Note:

Fabricate the front panel from 1.54 cm (22 gauge) galvanized iron sheet measuring 53.34 cm x 21.59 cm (21 x 8.5 in.). Bend 5 cm on each side to connect to the chassis as shown. Make a hole for the voltmeter.

Fabricate a 1.54 cm thick (22 gauge) galvanized iron chassis measuring 55.88 x 27.9 cm (22 x 15 inches). Fold 5 cm on each side and make a 1.25 cm peak. The total size will be (25x43x5cm) with a 1.25 cm peak on the bottom of the chassis.

Make smaller holes and holes for connections in the course of further work.

The visor running from the given part of the chassis is not shown in the figure.

Rice. 3.5. Drawing for the manufacture of the chassis

2. Try on the control panel and drill the necessary holes for controls, indicators, etc. Pay attention to the insulating material between the chassis and parts of the device, see fig. 3.6 part of PLATE1. This can be achieved with a small amount of RTV Silicone Sealant at room temperature curing. Drill appropriate holes as you work, checking for correct location and dimensions.

Rice. 3.6. General view of the assembled device

Note:

The wires are shown slightly elongated to ensure clarity of images and connections.

The dotted lines show the elements and connections located under the chassis.

3. Try on the remaining parts (see fig. 3.6) and drill all necessary holes for mounting and placement. Pay attention to the fuse holders FH1 /FS1 and the insulation of the input power cord BU2. They are located on the underside of the chassis and are shown in dotted lines.

4. Provide sufficient space for high voltage components: transformer output pins, high voltage diodes and resistor R18. Please note that the high voltage diodes are mounted on the plastic board with RTV double sided adhesive tape.

5. Replace the control panel. Secure the circuit board with a few pieces of tape coated with RTV sealant when you are sure everything is OK.

6. Make all connections. Pay attention to the use of wire nuts when connecting terminals T1 and T2.

Preliminary Electrical Tests

To perform preliminary electrical tests, follow these steps:

1. Short-circuit the output terminals of the transformer with a high voltage clamp wire.

2. Remove the fuse and install a 60 W barreter (electrovacuum current stabilizer) in its holder as a ballast resistor for the test period.

3. Set the switch S4 (see Fig. 3.7) to the off state, move the axis of the switch combined with the variable resistance R8 / S2 to the “off” position, set the variable resistances R14 and R19 to the middle position and turn on the device in a 115 V AC network by plugging the COl power cord into a power outlet.

4. Turn the axis of the combined variable resistance switch R8 until it turns on and watch the lamps LA1 and LA2 light up.

5. Press the charge button S3 and make sure that the relay RE1 turns on (clicking sound is heard) and the lamp LA2 is off for the time that the button S3 is pressed.

6. Turn on S4 and press S3, note that the barreter, turned on according to step 2, is lit at full heat.

7. Press the start button S1 and observe a flash between the trigger electrode TE1 and the main discharge gap between G1 and G2. Pay

Rice. 3.7. Front panel and controls

note that the axis of variable resistance is set to the average value, but by turning the axis clockwise, you can increase the discharge.

Main tests

To test, follow these steps:

1. Unplug the power cord and turn off S2 and S4.

2. Connect a 30µF, 4kV capacitor and a 5kΩ, 50W resistor as C and R as shown in Figure 2. 3.6.

3. Remove the ballast lamp and insert a 2A fuse.

4. Set the trigger gap to 0.32 cm and the main gap to 0.16 cm.

5. Connect a high accuracy voltmeter through an external capacitor.

6. Turn on the device and turn on S2 and S4. Press button S3 and check that the external capacitor is charged to 1 kV before RE1 is switched off. Note that in the normal state LA2 is on and off only for the duration of the charge cycle. When the set charge is reached, the LA2 LED turns on again, indicating that the system is ready.

7. Rotate R8/S2 30° clockwise and note that the voltage rises to a higher value before charging stops.

8. Press button S1 and observe an instantaneous strong arc in the main gap that occurs when energy is directed to the external load.

9. Charge the device to 2500V by measuring the voltage with an external voltmeter connected across the capacitor. Adjust R19 so that the front panel voltmeter reads 2.5 at full scale 5. Make a note on the front panel so you know where the voltage is 2500V. voltmeter. Repeat step 8, observing an intense arc when discharging. Repeat the charge and discharge cycles at different voltages to familiarize yourself with the control of the device.

This completes the verification and calibration of the device. Further operations will require additional equipment, depending on the project in which you are experimenting.

Useful bottom equipment mathematical relationships

System storage energy:

The ideal current boost is achieved in LC systems. Use a factor of 0.75 when using oil capacitors and lower values ​​for photo and electrolytic capacitors. Peak current time to 1 A cycle:

magnetic flux

A \u003d area of ​​\u200b\u200bthe edge of the coil in m 2; Le = distance between poles in m; M = mass in kg. Force:

Acceleration: Speed:

where t is the time to reach the peak current.

The current pulse generator (PCG) is designed to generate multiple repetitive current pulses that reproduce the electro-hydraulic effect. The basic schemes of the GIT were proposed back in the 1950s and have not undergone significant changes over the past years, however, their component equipment and the level of automation have been significantly improved. Modern GIT are designed to operate in a wide range of voltage (5-100 kV), capacitor capacitance (0.1-10000 μF), stored storage energy (10-106 J), pulse repetition rate (0.1-100 Hz).

The above parameters cover most of the modes in which electro-hydraulic installations for various purposes operate.

The choice of the GIT scheme is determined in accordance with the purpose of specific electro-hydraulic devices. Each generator circuit includes the following main blocks: power supply - transformer with a rectifier; energy storage - capacitor; switching device - forming (air) gap; load - working spark gap. In addition, the PCG circuits include a current-limiting element (this may be resistance, capacitance, inductance, or their combined combinations). In PCG circuits, there can be several forming and working spark gaps and energy storage devices. The GIT is powered, as a rule, from an alternating current network of industrial frequency and voltage.

GIT works as follows. Electric energy through the current-limiting element and the power supply enters the energy storage - capacitor. The energy stored in the capacitor with the help of a switching device - an air forming gap - is pulsed to the working gap in a liquid (or other medium), on which the electric energy of the storage device is released, resulting in an electrohydraulic shock. In this case, the shape and duration of the current pulse passing through the discharge circuit of the PCG depend both on the parameters of the charging circuit and on the parameters of the discharge circuit, including the working spark gap. If for single pulses of special PCGs, the parameters of the charging circuit circuit (power supply) do not significantly affect the overall energy performance of electrohydraulic installations for various purposes, then in industrial PCGs, the efficiency of the charging circuit significantly affects the efficiency of the electrohydraulic installation.

The use of reactive current-limiting elements in the PCG circuits is due to their ability to accumulate and then release energy into the electrical circuit, which ultimately increases the efficiency.

The electrical efficiency of the charging circuit of a simple and reliable circuit (PCG with a limiting active charging resistance (Fig. 3.1, a) is very low (30-35%), since the capacitors are charged in it by pulsating voltage and current. Introduction to the circuit of special voltage regulators (magnetic amplifier, saturation inductor), it is possible to achieve a linear change in the current-voltage characteristic of the capacitive storage charge and thereby create conditions under which energy losses in the charging circuit will be minimal, and the overall efficiency of the PCG can be increased to 90%.

To increase the total power when using the simplest PCG circuit, in addition to the possible use of a more powerful transformer, it is sometimes advisable to use a PCG that has three single-phase transformers, the primary circuits of which are connected by a "star" or "triangle" and are powered by a three-phase network. The voltage from their secondary windings is supplied to separate capacitors that operate through a rotating forming gap for one common working spark gap in the liquid (Fig. 3.1, b) [-|]. .4

When designing and developing PCG of electrohydraulic installations, the use of the resonant mode of charging a capacitive storage from an alternating current source without a rectifier is of considerable interest. The overall electrical efficiency of resonant circuits is very high (up to 95%), and when they are used, an automatic significant increase in operating voltage occurs. It is advisable to use resonant circuits when operating at high frequencies (up to 100 Hz), but this requires special capacitors designed to operate on alternating current. When using these schemes, it is necessary to observe the well-known resonance condition

W \u003d 1 / l [GS,

Where is the co-frequency of the driving EMF; L is the inductance of the circuit; C is the capacitance of the circuit.

A single-phase resonant PCG (Fig. 3.1, c) can have an overall electrical efficiency exceeding 90%. GIT allows you to get a stable frequency of alternating discharges, optimally equal to either single or double the frequency of the supply current (i.e., 50 and 100 Hz, respectively) when powered by industrial frequency current. The application of the circuit is most rational (. with a power of the supply transformer of 15-30 kW. A synchronizer is introduced into the discharge circuit of the circuit - an air forming gap, between the balls of which there is a

A rotating disk with a contact that triggers the forming gap when the contact passes between the balls. In this case, the rotation of the disk is synchronized with the moments of voltage peaks.

The circuit of a three-phase resonant PCG (Fig. 3.1, d) includes a "three-phase step-up transformer, each winding on the high side of which operates as a single-phase resonant circuit n ^ one common for all or for three independent working spark gaps with a common synchronizer for three forming gaps This circuit allows you to get a discharge alternation frequency equal to three or six times the frequency of the supply current (i.e. 150 or 300 Hz, respectively) when operating at industrial frequency. The circuit is recommended for operation at PCG powers of 50 kW or more. since the charging time of a capacitive storage (of the same power) is less than when using a single-phase PCG circuit. However, a further increase in the power of the rectifier will be advisable "only up to a certain limit.

It is possible to increase the efficiency of the process of charging the capacitive storage of the PCG by using various schemes with a filter capacitance. The PCG circuit with a filter capacitance and an inductive charging circuit of the working capacitance (Fig. 3.1, (3)) makes it possible to obtain almost any pulse alternation frequency when operating on small (up to 0.1 μF) capacities and has an overall electrical efficiency of about 85%. This is achieved by the fact that the filter capacitance operates in the mode of incomplete discharge (up to 20%), and the working capacitance is charged through an inductive circuit - a choke with low active resistance - for one half-cycle in an oscillatory mode, set by the rotation of the disk on the first forming gap In this case, the filter capacity exceeds the working one by 15-20 times.

The rotating disks of the forming spark gaps sit on one shaft and therefore the frequency of alternating discharges can be varied over a very wide range, maximally limited only by the power of the supply transformer. 35-50 kV transformers can be used in this circuit as it doubles the voltage. The circuit can also be connected directly to a high-voltage network.

In the PCG circuit with a filter tank (Fig. 3.1, e), the working and filter tanks are connected in turn to the working spark gap in the liquid using one rotating spark gap - the forming gap. However, during the operation of such a PCG, the operation of a rotating spark gap begins at a lower voltage (when the balls approach) and ends at a higher voltage (when the balls move away) than specified by the minimum distance between the spark gap balls. This leads to instability of the main parameter

Discharges - voltage, and consequently, to reduce the reliability of the generator.

To improve the reliability of the PCG by ensuring the specified stability of the parameters of the discharges, a rotating switching device is included in the PCG circuit with a filter capacitance - a disk with sliding contacts for alternate preliminary currentless switching on and off of the charging and discharge circuits.

When voltage is applied to the charging circuit of the generator, the filter capacitance is initially charged. Then the circuit is closed by a rotating contact without current (and hence without sparking), a potential difference occurs on the balls of the forming spark gap, a breakdown occurs and the working capacitor is charged to the voltage of the filter capacitance. After After this, the current in the circuit disappears and the contacts open again without sparking by rotating the disk.Then, the contacts of the discharge circuit are closed by the rotating disk (also without current and sparking), and the voltage of the working capacitor is applied to the forming discharger, its breakdown occurs, as well as the breakdown of the working spark gap in the liquid. In this case, the working capacitor is discharged, the current in the discharge circuit stops and, therefore, the contacts can be opened again by rotating the disk without sparking destroying them.Then the cycle is repeated with a discharge repetition rate specified by the frequency of rotation of the disk of the switching device.

The use of PCG of this type makes it possible to obtain stable parameters of stationary spherical dischargers and to carry out the closing and opening of the circuits of the charging and discharge circuits in a currentless mode, thereby improving all the performance and reliability of the generator of the power plant.

A power supply scheme for electro-hydraulic installations was also developed, which allows the most rational use of electrical energy (with a minimum of possible losses). In the known electro-hydraulic devices, the working chamber is grounded, and therefore part of the energy after the breakdown of the working spark gap in the liquid is practically lost, dissipating on the ground. In addition, with each discharge of the working capacitor, a small (up to 10% of the initial) charge is retained on its plates.

Experience has shown that any electro-hydraulic device can effectively operate according to a scheme in which the energy stored on one capacitor C1, passing through the forming gap of the FP, enters the working spark gap of the RP, where for the most part it is spent on performing the useful work of the electro-hydraulic shock. The remaining unused energy goes to the second uncharged capacitor C2, where it is stored for later use (Fig. 3.2). After that, the energy of the recharged to the required
the value of the potential of the second capacitor C2, having passed through the forming gap FP, is discharged into the working spark gap RP and the newly unused part of it now falls on the first capacitor SU, etc.

The alternate connection of each of the capacitors either to the charging or to the discharge circuit is carried out by the /7 switch, in which the conductive plates A and B, separated by a dielectric, are connected in turn to contacts 1-4 of the charging and discharge circuits.

Mitchell Lee

LT Journal of Analog Innovation

Steep pulse sources that mimic a step function are often useful in some laboratory measurements. For example, if the slope of the fronts is on the order of 1 ... 2 ns, you can estimate the rise time of the signal in the RG-58 / U cable or any other, taking a segment only 3 ... 6 m long. The workhorse of many laboratories - the ubiquitous pulse generator HP8012B - does not reach 5 ns, which is not fast enough to solve such a problem. Meanwhile, the rise and fall times of the gate driver outputs of some switching regulator controllers can be less than 2 ns, making these devices potentially ideal pulse sources.

Figure 1 shows a simple implementation of this idea, based on the use of a flyback controller operating at a fixed switching frequency. The controller's own operating frequency is 200 kHz. Applying a portion of the output signal to the SENSE pin forces the device to operate at minimum duty cycle, producing output pulses of 300 ns. Power decoupling is important for this circuit, since the output current delivered to a 50 Ω load exceeds 180 mA. 10 µF and 200 Ω decoupling elements minimize peak distortion without sacrificing edge steepness.

The output of the circuit is connected directly to a 50 Ω termination, providing a signal swing of about 9 V across it. In the case where the quality of the pulses is of paramount importance, it is recommended to suppress the triple-pass signal by absorbing reflections from the cable and the remote load using the serial termination shown in the diagram. Serial matching, that is, matching on the transmitting side, is also useful when the circuit operates on passive filters and other attenuators designed for a certain signal source impedance. The output impedance of the LTC3803 is approximately 1.5 ohms, which should be taken into account when choosing the resistance of the series terminating resistor. Series matching works well up to impedances of at least 2 kΩ, above which it becomes difficult to provide the necessary bandwidth at the junction of the resistor and the circuit, resulting in degraded pulse quality.

In a series-matched system, the output signal has the following characteristics:

• pulse amplitude - 4.5 V;
• the rise and fall times are the same, and equal to 1.5 ns;
• pulse flat top distortion - less than 10%;
• the decline in the top of the impulse is less than 5%.

By directly connecting a 50 Ω load, the rise and fall times are not degraded. To get the best pulse shape, connect the 10uF capacitor as close as possible to the VCC and GND pins of the LTC3803, and connect the output directly to the terminating resistor using stripline technology. A characteristic impedance of approximately 50 ohms has a 2.5 mm wide printed conductor on a 1.6 mm thick double-sided printed circuit board.

Related materials

PMIC; DC/DC converter; Uin:5.7÷75V; Uout:5.7÷75V; TSOT23-6

• Linear Technology is generally a top company! It is very, very unfortunate that they were gobbled up by Analog Devices. Don't expect anything good from this. I used to meet an article by an English-speaking radio amateur. He built a generator of very short pulses with a width of a few nanoseconds and rise / fall times in picoseconds. On a very fast comparator. Too bad I didn't save the article. And now I can't find it. It was called something like "... real ultrafast comparator ...", but somehow it's not like that, it's not googled. I forgot the name of the comparator, and I don’t remember the company. Then I found a comparator on ebay, it cost about 500 rubles, in principle, it is budgetary for a really worthy device. Linear Tecnology has very interesting chips. For example LTC6957: rise/fall time 180/160 ps. Awesome! But I can hardly build a measuring device on such a mikruha.
• Isn't this the case on the LT1721? Tunable 0-10ns.

Scheme 1

The generator has been designed to use the minimum number of commonly available electronic components, with good repeatability and sufficient reliability. The oscillator option (circuit 1) is assembled on the basis of the widely used UC3525 (U1) PWM controller, which controls the Q4-Q7 field-effect transistor bridge circuit. If the lower switches of each of the half-bridges operating in antiphase are controlled directly by the outputs of the 11/14 U2 microcircuit, then bootstrap cascades on transistors Q2, Q3 are used as upper-side drivers. Such stages are widely used in most modern IC drivers and are fairly well documented in the power electronics literature. The input voltage is AC or DC (~24~220V/30-320V), supplied to the input of the diode bridge (or bypassing it in the case of a DC voltage supply), feeds the power part of the circuit. To prevent a large starting current, a thermistor Vr1 (5A / 5Ohm) is included in the power circuit break. The control part of the circuit can be powered from any source with an output voltage of +15/+25V and a current of 0.5A or more. The parametric voltage regulator on transistor Q1 can have an output voltage from +9 to +18V (depending on the type of power switches used, for example), but in some cases you can do without this stabilizer if an external power source with the necessary parameters is already stabilized. The UC3525 microcircuit was not chosen by chance - the ability to generate a pulse sequence from several tens of hertz to 500 kHz and sufficiently powerful outputs (0.5A). At least the TL494 microcircuits could not function at a frequency of less than 250 Hz in a push-pull mode (in a single-cycle mode - no problem) - the internal logic failed and the pulse sequence, as well as their duration, became chaotic.

The frequency of the pulse sequence is adjusted by a variable resistor R1, the pulse duration is adjusted using R4. The initial duration of the "dead time" is set by the resistor R3.

Scheme 2

The generator shown in diagram 2 is a complete analogue of the previous circuit and has practically no circuit differences. However, the domestic K1156EU2 chip (a complete analogue of the UC3825) used in this generator is capable of operating at higher frequencies (almost up to 1 MHz), the output stages have a greater load capacity (up to 1.5A). In addition, it has a minor difference in pinout compared to UC3525. So, the "clock" capacitor is connected to pin 6 (5 - for the 3525 chip), the timing resistor is connected to pin 5 (6 - for the 3525 chip). If pin 9 of the UC3525 is the output of the error amplifier, then in the UC3825 this pin acts as the input of the "current" limiter. However, all the details are in the datasheet for these microcircuits. It should be noted, however, that the K1156EU2 is less stable at frequencies below 200 Hz and requires more careful layout and mandatory blocking of its power supply circuits with relatively large capacitors. If these conditions are ignored, the smooth adjustment of the pulse duration near their time maximum may be disturbed. The described feature manifested itself, however, only when assembling on a breadboard. After assembling the generator on the printed circuit board, this problem did not appear.

Both circuits are easily scalable in power by using either more powerful transistors or by connecting them in parallel (for each of the switches), as well as by changing the supply voltage of the power switches. It is desirable to "plant" all power components on radiators. Up to a power of 100W, adhesive-based heatsinks were used, designed for installation on memory chips in video cards (output switches and a stabilizer transistor). Within half an hour of operation at a frequency of 10 kHz with a maximum duration of output pulses, at a supply voltage of the keys (31N20 transistors were used) + 28 V to a load of about 100 W (two 12 V / 50 W lamps connected in series), the temperature of the power switches did not exceed 35 degrees Celsius.

To build the above circuits, ready-made circuit solutions were used, which I only rechecked and supplemented during prototyping. For generator circuits, printed circuit boards were designed and manufactured. Figure 1 and Figure 2 show the boards of the first variant of the generator circuit, Figure 3, Figure 4 - images of the board for the second circuit.

Both circuits at the time of this writing were tested at frequencies from 40Hz to 200kHz with various active and inductive loads (up to 100W), at constant input supply voltages from 23 to 100V, with output transistors IRFZ46, IRF1407, IRF3710, IRF540, IRF4427, 31N20 , IRF3205. Instead of bipolar transistors Q2, Q3, it is recommended to install (especially for operation at frequencies above 1 kHz) field-effect transistors such as IRF630, IRF720 and the like with a current of 2A and an operating voltage of 350V. In this case, the value of the resistor R7 can vary from 47 ohms (over 500 Hz) to 1k.

Component ratings indicated through a slash - for frequencies above 1 kHz / for frequencies up to 1 kHz, except for resistors R10, R11, not indicated in the circuit diagram, but for which there are mounting places on the boards, jumpers can be installed instead of these resistors.

Generators do not require adjustment and, with error-free installation and serviceable components, they start working immediately after power is applied to the control circuit and output transistors. The required frequency range is determined by the capacitance of the capacitor C1. The component ratings and positions for both circuits are the same.

Figure 5 shows the assembled generator boards.

List of radio elements