DE3305482A1 - AMPLIFIER COMPENSATED BY PRE-MAGNETIZING CURRENT - Google Patents

Description

The invention relates to a non-turn-off emitter follower single-ended push-pull circuit (SEPP circuit) of class B.

An emitter follower SEPP circuit is generally operated in class B from the viewpoint of efficiency. In such a circuit, it is essential to pass reactive current through the output transistors in order to smoothly connect the upper and lower transfer parameters (positive and negative half-cycles), that is, to reduce crossover distortion. In a conventional circuit of this type, crossover distortion can occur because one of the transistors is turned on when the other of them is turned off. In order to overcome this problem, a class B non-disconnection circuit has recently been used, in which such disconnection is due to the presence of a constant reactive current, the st ^; indiq in the otherwise non-

'ii Vfc »

-2-- G

conductive transistor flows, is not possible. A servo circuit is used to generate this reactive current. With this arrangement, crossover distortion is actually reduced. However, nothing is draw on a current distortion arising from the non-linearity of the current transfer characteristic, and in relation to on a voltage distortion, which is based on the exponential transfer characteristic of the transistor, done.

There is also the disadvantage that thermal runaway or overturning can occur if there is no complete compensation for the reactive current is created. In the arrangement according to the prior art, the reactive current changes with changes There is little in the signal levels and the ambient temperature or move short periods of time.

Furthermore, it is extremely difficult to obtain such a Circuit to be dimensioned and designed, as the above-mentioned thermal compensation is very sensitive. The interpretation is particularly in the case of a conventional non-shutdown Class B circuit that does not use positive feedback difficult. In any case, it can be in either way designed circuit of the temperature compensation may only be imperfect.

To explain the prior art on which the invention is based, reference is made to FIGS. 1-3. In these show:
30th

1 shows the diagram of an exemplary conventional

Class> B SEPP circuit;
FIG. 2 is a graph showing a characteristic of the circuit of FIG. 1; FIG.

3 is a graph showing a characteristic curve of a other usual circuit.

In Fig. 1 the basic structure of a conventional nichtabschaltbaren SEPP Schaitung Class B is shown where may be adder in error amplifier A ₁ and A ₂ having a gain factor less than one, voltage generation circuits B ₁ and B _2, the voltage ·, an input signal source C and a bias voltage source V _ß for bipolar transistors Q ₁ and Q ₂ are provided.

In Fig. 1, a reactive current I ^ is composed of two currents I ^ and I _E2 , which are present when there is no signal at the input terminal IN. Currents Ι _β1 , I _ß2 are supplied from an energy source V _ß. If the base-emitter voltage of each transistor is V _ßE and the emitter _{resistance R E} , the following equation results:

V-V

I _d , \ ^ E _{1)

_{1)

^R E

When an input signal current i _i flows, the current i _{£ 1 is} increased. Assuming a current amplification factor h ~ .. of the transistor Q ₁ , the following equation is obtained:

¹ EI ^{= h} fe1 ' ¹ I

This current i _{£ 1} generates a voltage parallel to the resistor R _£. This then turns the input voltage V ^ ₁ to amplifier A ₁ :

^V i1 ■ < ^V BE - V * ¹ El ■ ^R E

= < ^V BE - V ^{+ h} fe1 - ¹ I - ^R E 30

The presence of this voltage causes the base of the

Transistor Q _{2 is} reverse biased and would turn off the transistor if the amplifier A _{1 were} not provided . If the gain of the amplifier A. is set to one, then the voltage V,. positive feedback to the base of transistor Q ₁ under these circumstances, whereby the input voltage to transistor Q _{1 is} increased. In this way it becomes constant constantly

Reactive current I without reverse biasing transistor Qp, delivered. A very similar process occurs when the input current i. being reversed to in the direction to flow in which the transistor Qp is turned on.

Fig. 2 shows a current transfer characteristic with respect to the input signal current i. in the circuit indicated in FIG. In general, when the emitter current of a transistor increases to high current levels, the current amplification factor h _fe decreases and consequently the resulting characteristic is highly non-linear, resulting in a high level of current distortion. Further, if the gains of the amplifiers A. and B _{1 are set} to one as described above, the positive feedback ratio becomes 100% and the stabilizing effect of the resistor Rr is completely lost (R ^ becomes zero in equation (1)) . Therefore, the reactive current ij becomes unstable and oscillations can occur. In practice, the gain factors are set to values below one, but the circuit is still quite unstable with regard to temperature.

If a non-deactivatable circuit arrangement of the class B is not applied and instead by omitting the amplifier A ,, Ap in Fig. 1 in an effort to rely on the To reduce distortion based on non-linear current transfer characteristics, a constant voltage control is effected, a transmission characteristic as shown in FIG. 3 is established. However, remains in this case too, a distortion based on the exponential transfer characteristics of the transistors is obtained.

In a conventional SEPP circuit of class B ^{using a constant voltage control,} 1.) vO an overlap distortion based on the exponential transfer characteristic and an on / off operation t'l'-r output tr. ·., Generally Loren br · dormant distortion. Self

if a non-shutdown, a constant current control class B SEPP circuit is used, 2) a distortion based on a non-linear transfer characteristic occurs. Furthermore, 3.) is without consideration Depending on the type of control, a temperature compensation for the reactive current is necessary, a complete compensation however, it is not possible. 4.) Periods longer than several tenths of a minute are also required until the Reactive current becomes constant after the application of energy. Furthermore, 5.) the reactive current fluctuates with the presence or absence of the input signal, and the magnitude of the reactive current deviates from a set value to a large extent off when there is a large signal "Finally, 6.) the operating point is unstable and it changes with the ambient temperature as well as the presence or absence of the input signal due to the disadvantages specified under 3.) to 5.) above.

To the disadvantages and shortcomings outlined To avoid it, it is an object of the invention to provide a no-turn-off emitter follower SEPP circuit of the B, which has an extremely low distortion and which completely compensates for the temperature of the reactive current makes unnecessary.

Furthermore, the regulation of the reactive current setting of an output amplifying element in one Amplifiers of the SEPP type (single-ended push-pull type) mainly only made by hand stranding a rheostat. Because the reactive current setting circuit a If a thermal compensation element, such as a varistor or thermistor, is included, the reactive current will increase over time and change temperature. Typically, several tenths of a minute are necessary to maintain a constant reactive current state after energy has been applied. Another disadvantage is that there are fluctuations in the work tpunkts, di ·? nuf Pe <je lande run η ρ η ο inei> Fituianyssignals

based, a so-called thermal distortion (thermal cross-modulation distortion) can result. Even if a push-pull arrangement in which the transistors are class A are operated, is used, odd harmonics of the input signal are amplified '^ although even harmonics can be suppressed to a certain extent).

In view of the above, it is another object of the invention to provide a bias control circuit for an amplifier capable of controlling a transistor amplifier circuit by maintaining the DC reactive current of an amplifying element at a substantially constant value regardless of to stabilize the temperature, and thus suitable _s is a cross-over distortion, switching distortion, etc. to reduce..

To achieve these and other objects of the invention, an amplifier circuit having a non-shutdown function is provided Single-follower single-ended push-pull training (SEPP training) of class B created the extreme has a low distortion factor and a temperature compensation for the reactive current is unnecessary. In detail, according to the invention, a first and a second error amplifier, one for the positive and one for the negative half-period, each of which is provided has three input terminals, two being non-inverting terminals and one being an inverting terminal.

Connection is. Each one inverting connection

is connected to the output electrode of the corresponding reinforcing element. The output electrodes of the reinforcing elements are via loads, resistive loads, or current mirror circuit loads

connected to an output terminal. Thus becomes 35

to the non-inverting input terminals of the error amplifiers a tension is placed which corresponds to the size of the in the respectiveij '-' ii ViU ^ .uürkereltMiiiHit f 1 i c ^ endiMi represents reactive current.

ΑΛ.

ι The first, non-inverting input terminals each Error amplifiers receive the input signal level shifted by a predetermined value, while the second, inverting input terminals renders the output signal, which is also shifted by a predetermined amount, received. The outputs of the error amplifiers are connected to the Input signal for application to the input electrodes of the amplifying elements is summed.

Furthermore, according to the invention, an amplifier circuit is created, in which there is a bridge circuit between the output electrodes of the amplifying elements and an output terminal inserted is'c. An error amplifier · generates an error output signal that is the difference as one to the input signal level and represents a voltage proportional to an output voltage detected by the bridge circuit. A variable bias circuit controls the DC bias of the amplifying element in accordance with the error signal generated thereby. With this arrangement it is not necessary to provide any temperature compensation element such as a varistor or thermistor. Furthermore, no reactive current regulation is required because the direct bias current of the amplifying element is kept constant at all times.

The subject matter of the invention is explained with reference to the drawings. Show it:

4 shows the diagram of one designed according to the invention Circuit;

5 is a graph showing a characteristic curve

the circuit of Fig. 4;
6 shows a diagram for a special embodiment of the circuit of FIG. 4;

7 and 8 are diagrams for further special configurations the circuit of Fig. 4;

• te

9 shows a circuit diagram for a further preferred embodiment

approximate form according to the invention; FIG. 10 shows a diagram for a special embodiment of the circuit of FIG. 9;
11 shows a graph of a transmission characteristic of the

Circuit of Fig. 9;
12 to 14 circuit diagrams for further embodiments

according to the invention;
FIG. 15 shows a modification of the bridge circuit from FIG. 12.

4 shows the basic structure of a SEPP circuit according to the invention with two error amplifiers A ₁ ¹ and Ap ¹ , each of which has three input connections a, b, c and a ', b ¹ , c'. The emitters of the transistors Q ₁ , Q ₉ are connected to one another via resistors R _P , '^ - t.

the connection point of the two resistors is connected to the output terminal AUS. The emitters of the transistors Q ₁ , Q2 are also connected to the input terminals a and a ^{1 of} the error ^{amplifiers A. 1} and A ₂ ¹ . The input terminals c and c ¹ are connected to the input cycle IN _{via bias voltage sources V β.} The connection point of the two resistors Rr is connected to the remaining input connections b, b ^{1 of} the amplifiers A ₁ ¹ , A ₂ ¹ via bias voltage sources Vu. The outputs of amplifiers A ₁ ¹ and Ap ^{1 are connected} to voltage generator circuits B and B ₀ , which consist of voltage adders, in which the outputs of amplifiers A ₁ ¹ , A ₂ ¹ are summed with the input signal.

If a negative signal with respect to the input connections a - c of the amplifier A ₁ ^{1 is applied} to the connection b, the amplifier A ₁ ¹ opens the input connection b, ie it makes the impedance imposed on the connection by the amplifier essentially infinite, so that the signal at connection b has no influence on the amplifier output.

Furthermore, the amplifier A ₁ ¹ brings the input connection c to the 35th

Open when a negative signal with respect to input terminals b and a is applied. In a similar way 3 opens ύ '-. R amplifier A _? 'the input terminal b ¹ if

ι a positive input signal with respect to the input connections a 'and c ^{1 is applied} to the connection b ¹ while connecting the connection c ^; opens when a positive input signal with respect to the input connections a ¹ and b ^{1 is applied.}

With the above arrangement, feedback loops which have the terminals a, a ¹ and c, c ^{1 of} the amplifiers A ^ ¹ and A ₂ ¹ as input terminals are formed. This means that the reactive current 1 ^, if there is no input signal, can be represented as:

V _B

^R E

When a positive signal is entered and a to i _$ . R ^ the same negative signal is inevitably applied between the terminals a and c, then the input terminal b is opened. Assuming a gain factor o <for the amplifier A ₁ ¹ , an input voltage V. with reference to the signal component becomes:

Previous year V ₀ ♦ AV _{BE +} i _s R _E - ~ V _g (3)

where is:

ο an output voltage,
AVgr a change from Vnr,
1 the signal current and

V is an input voltage between terminals a and c,

which can be expressed as
30th

^V g = ^V i - < ^V o ⁺ W

Equation (4) is substituted into equation (3) to give the following equation:

^V i = ^V o ^{+ AV} BE ^{+ 1} S ¹¹ E ""["ι - ^(V 0 ^{+ 1} S ^

Ordering this equation, V is obtained as:

AV _RF
V - V - i R - (5)

_ If oc in equation (5) becomes large, the effect of ο

Vnr, the one on the exponential transfer characteristic distortion based member is canceled. Therefore, in the case of a linear load i will not be distorted, and consequently V and V analog, that is, they follow one another, which largely eliminates distortion.

If, on the other hand, one considers the operation of the amplifier Α « ¹ , the input connection c ^{1 is} opened because a signal which is approximately positive with (i.Rr + V) is inevitably applied with reference to the connections b ¹ and a ¹ . Therefore the amplifier Ap ¹ acts as an error ^{amplifier having the terminals b 1} and a ¹ as its inputs. This becomes the reactive current

I. ₌ L_

^R E

So the result is equal to equation (2) above.

If a negative signal is input at the input terminal IN, a completely similar process takes place, only changing the sign of the various signals.

The circuit of FIG. 4 has the transfer characteristic shown in FIG. 5, which is that of one without disconnection working class B SEPP circuit in which the resulting characteristic curve linear i-st.

FIG. 6 shows a special embodiment of the circuit of FIG. 4, the error ^{amplifier A. 1 being formed} by two transistors Q ₃ and Q ₅ , the respective bases of which .'ji-Mi F. ingrt.'Kjsansch I i! " .r, <wi α mid ·: corresponded. The error

ι amplifier A ₂ ¹ is formed by the transistors Q ₆ and Q ₈ , the bases of which ^{correspond to the input terminals a 1} and c ^1, respectively. The output currents of the constant current sources I ₁ and I ₂ have the relationship I ₁ > I ₂ -

In the particular embodiment of FIG. 7, four input terminal type error amplifiers are used, an amplifier of ideal performance is formed by adjusting the error amplifiers.

The circuit shown in FIG. 8 is obtained in that the ohmic load on the error amplifier of FIG Current mirror circuit loads is replaced. This circuit has a high gain factor and is suitable for manufacture as an integrated circuit.

In any case mentioned above, ^{bias voltages applied between the terminals c or c 1} and b or b ¹ or d or d ^{1 are} set to a value that provides a desired reactive current magnitude, they are not necessarily limited to a value of V _D as he is is used in the circuit of FIG.

Further, even if the constant voltage control point (the input terminal in Figs. 6, 7 and 8) is opened and I ^ is converted into an input signal source, i.e., a constant current control circuit is formed, a stabilized one SEPP circuit, which has the features of reactive power servo control and class B operation without shutdown, can be obtained.

Furthermore, the first and second reinforcing members can have a Darlington construction.

As has been explained, according to the invention the reactive current and distortion components of the SEPP circuit of class B are detected simultaneously (in real time) and amplified as error signals and fed back.

ν ν y ν "

4b-

As a result, the distortion at the SEPP output terminal greatly reduced, the need for further temperature compensation of the reactive current is eliminated, and the reactive current is kept stable at a constant value immediately after the energy is applied. The reactive current also remains stable at a predetermined value when a large signal is introduced. Furthermore, the output impedance the circuit is very significantly reduced, so that the load on the previous stage is very much reduced.

In addition, by using a no-shutdown SEPP circuit arrangement, Class B without switching distortion, a high performance SEPP circuit having a low distortion factor and a has much higher stability than known circuits.

9 shows a circuit diagram for a further embodiment according to the invention. Here, an input signal Vj is not only used as a base input to an output amplification _{transistor Q 1a} via a variable bias circuit 1a and for controlling the DC bias of transistor Q ₁₃ , but also as a base input to a

ι a

second output amplification transistor Q ,. placed, in the latter case via a constant voltage source

25 "2a, a resistor R _ß , a constant voltage source 2b and a variable bias circuit 1b. The transistors Q, | _a and Q _(1t) are connected in a class B emitter follower SEPP arrangement, with the emitters of the two transistors connected to one another via resistors R. and R _{1h are} connected A signal ν 'generated at the connection point between the resistors R ₁ and R ₁ is transmitted via a ι a \ υ ο

Resistance Rp to the circuit output ν, which can be used to operate a load, for example a loudspeaker (not shown). The resistors R _{1 a} , R ₀ and "ι a c

^R 1b ' ^R 2 ^s * ^nc ^ in each case between the emitters of the assigned transistors Q ₁ ^, Q _1ti and the circuit output AUS in series cj-i v'Miil tor .. Wi'! Er ', Läridi »K ₃ ^ and W . as well as R- ,, and R .. lie

parallel to the respective series-connected circuits, the resistors R ₃ , and R ₄ . in series with the emitters of the corresponding transistors Q., and Q ₁ ^. Thus, two bridge circuits are formed, one of which has the resistors R _1a> R ₂ , R ₃₉ and R ₄₃ and the other the resistors R ^, R ₂ . i ^ and R ₄ ^ includes.

One input to an error amplifier 3a is connected to a connection point of the resistors R ₃ and R-, the one input of an error amplifier 3b is connected to a connection point of the resistors R _3b and R _4b . As the other inputs to the error amplifiers 3a, 3b come in level through a constant _{voltage V Q} (supplied from the constant voltage sources Za and 2b),.

or voltages shifted below the input signal v. The output signals from the error amplifiers 3a, 3b are each applied as control signals via circuits 4a and 4b to one of the variable bias circuits 1a and 1b. The circuits 4a, 4b lead the error outputs of the error amplifiers 3a, 3b to the variable biasing circuits 1a, 1b in a selected manner in response to outputs from zero passage comparators 5a and 5b, the voltage ν 'from point a and a voltage v _Q from point d of the bridge circuits received as entrances, too. The circuit also includes constant current sources 6a and 6b.

Let us first consider the case of the positive half-cycles of the input signal Vj. Assuming that I _{d is} the reactive current of the transistor CL and that its output current is I, the following results for the voltage developed between points c and d:

^V cd = ^R i _a n _d + i _a > + R ₂ i _a (6)

Furthermore, the developed between points b and a Stress can be expressed as:

VWV ^u if V * -

W _ 4a wp

^V ba -K ^ ⁷ TTa * ^{cd 2} -

^R 1a * ^R 4a _T. ^R 1a ^R 4a " ^R 2 ^R 3a . _M ^{= R} 3a ^{+ R} 4a ^{d R} 3a ^{+ R} 4a ^{a {?)}

Assuming that the condition R ^ _a : R ₂ = Ro _a : R ^ _{3 is} fulfilled, that is, that a balanced state is maintained, then equation (7) can be rewritten as: R _1q R _Aa Ro R · ,,

ν - _ ^{1a 4a} τ - J- ^3a _ τ (g)

It follows that it is possible _{to detect the reactive current I ^ flowing through the transistor Q 1} by detecting the voltage between points b and a. The same is of course true for the case of the negative half-periods.

If the voltage V _fl at point a and the voltage V ^ at point b meet the condition V = V ^, d-.h-. for positive half-cycles or a zero-valued input signal, the zero crossing comparator 5a switches the circuit 4a on (conductive state), while for V, <Vj, ie for

α Q

negative half cycles, the comparator 5a switches off the circuit 4a (non-conductive state). On the other hand, the zero crossing comparator 5b operates in a complementary manner, so that it switches the circuit 4b on when Vj = V and switches it off _{when V ^ <V &.} Thus, positive and negative feedback loops operate simultaneously because, when no signal is applied, the condition V _a = V _{d is} met.

Assuming that the error signal inputs to the error amplifiers 3a, 3b each v. and V ₁ , so is the

la ID

parallel - / around resistor R _B generated voltage InRg-35

^! d

When the gain factors of the variable bias circuits 1a, 1b and the error amplifier 3a, 3b are sufficiently large, the error signal inputs become v ^ _a
and V ₁ J ₅ small enough to be neglected. In
In this case, the following equation (10) can be obtained from equation (9):

I - ^R 3a + ^R 4a _{R. Mon)}

^d " ² i ^ft ia ^ß 4a ^{) BB} **""■

Since the current I ^ is constant, the reactive current I ^ will always be be constant regardless of the presence or absence of an input signal.

Let us now examine the relationship between the output signal voltage v _Q and the input signal voltage v. be considered for positive half-periods. _{Assuming that the voltage v f} is present between the input and output terminals of the variable bias circuit 1a, that the transistor
Q _{1a has} the _{base output voltage v ß} and that the voltage Vnr is present between the base and emitter of the transistor Q., the following equation is obtained:

^v i ^{+ v} f = ^V B

= ^V BE ^{+ R} ia ^ a ^{+ 1} ^ ^{+ v} o '

= ^V BE ^{+ R} 1a ^ a ^{+ I} d ^{) + v} o

From equation (11), the following equation (12) can be obtained:

^v o ^{= v} i ^{+ v} f "[Ve ^{+ R} 1a ⁽ⁱ a ' ⁺ * d> ^{+ V} B

^ ^v i ^{+ v} f - [ ^(R 1a ⁺ R ₂ ^ a ^{+ R} U ^l d ^{+ V} Be]

-V6-

Furthermore, the error signal output v> from the error amplifier

i a

ker 3a can be expressed as:

i - v -

^V 4a

(13)

When the gain of the variable bias circuit 1a and the error amplifier 3a with

o <(= ν * / v,), then the following τ ι a

Equation can be obtained:

Substituting equation (14) into equation (12) one obtains:

^v o ■ ^v i

^R 1a ^R 4a

^V BeJ

Hence (equation 15):

(1 ₊ *) v ₀ . (1 ^) V ₁ - (V _{0 +} R ₂ I,

From equation (15), the following equation (16) is obtained:

^R 1a ^R 4a ^R 3a ^{+ R} 4a

- + * ^V BE

^V BE j

(16)

If ot is sufficiently large (ex ^ \ or ex - »- co), then equation (16) can be rewritten as:

^V d '

^{Since v} «= ν« ¹ - Ro ^ o »the following equation can be used

OO Co.

will be obtained:
V = ^v o ⁺ Va

* V ₁ - (I _d + V _D ) (18)

^{1 R} 3a ^{+ K} 4a ^{d ü}

As stated above, I ^ and V _{Q are} constant so that the amount of distortion in ν 'approaches zero. Since R _{2 is} a fixed value resistor, the distortion of the circuit output also approaches zero. For negative half-cycles, the input signal voltage v .. and the output signal voltage v _Q likewise have an analogous relationship.

A special embodiment of the circuit of FIG. 9 is shown in the circuit diagram of FIG. 10, the same reference symbols denoting equivalent components. The transistors Q _2a , Q ₃₃ and Q _2b> Q _3b each form an error amplifier 3a or

3b. The resistors R, R. each form the changing

Sa 5 0

Chen bias circuits 1a and 1b, respectively, while transistors Q _4a , Q _4b represent the zero cross comparators 5a and 5b, respectively.

During the positive half cycles, transistor Q is ₄₉ . switched off (V ₃ = V _b > V _d ), the transistors Q _2a and Q _2b work in the usual way. When no input signal is applied, transistors Q _2a > Q _2b operate normally, where

= I.

_3a = I ₄₃ . For negative half periods, V _b

which is why the transistor Q. is completely switched on and the transistor Q _{2a is} switched off, so that no voltage is generated in parallel with the resistor R. But because the

-aa

Transistors Qgu, and Q _{3t) on} the negative side operate normally, a current flows from them into the resistor Ru, so that a voltage is generated across it, which causes a feedback process.

Neither the operation of the previous stage nor that of the SEPP circuit itself is adversely affected by the presence or absence of the input signal, since h ^=! 2a ^{+ X} 3a ^+! 4a ^{and l} b ⁼ * 2b ^+! 3b ⁺ Ub are ^constant . Although the transistors Q ₁ and Qu are inverted

Have Darlington connection, they can also be connected in a conventional Darlington connection, and field effect transistors can be used as active elements of the previous stage.
15th

11 shows the transfer characteristic of a circuit according to the invention, a linear relationship being obtained by the positive-side transfer characteristic 41 and the negative-side transfer characteristic 42. It can be clearly seen from the graph that the invention has created a class B SEPP circuit which operates without disconnection (a voltage ν in FIG. 11 is v _g = v. - V ₀ ¹ ).

FIGS. 12 to 14 show circuit diagrams for further embodiments according to the invention, with the same reference numerals as in FIG. 10 denoting components that are similar to FIG. The circuit of Fig. 12 lacks zero cross comparators 5a and 5b, while the positive and negative variable bias circuits are formed by a transistor Qg ₃ and resistors Rg _g and Rg and a transistor Qcu and resistors R ₅ ^ and Rg. Furthermore, a positive and a negative error amplifier consist of transistors Q ₆ and Q ₇ or Qg, and Q _7b . The operating points of transistors Q ₆₃ and Q _6b are essentially set to shutdown.

-vö-

In the circuit of FIG. 13, the zero crossing comparators 5a and 5b of FIG. 9 have also been omitted; a positive and a negative variable bias circuit can be of transistors Qc-, Q _9a and resistors R _5a, R _6a and _5b of transistors _Q> Q _gb and resistors ^R 5b ^'R 9b 9 ^ebi arcs. ^E * ^ne error signal amplification for positive and negative reactive components is brought about by transistors Q _7a and Q _8a or Q _7b and Qg _b , while an error signal amplification for positive and negative signal components is brought about _{by transistors Q 5a} and Q _Qa or, Q _5b and Qg _b will.

In the circuit of Fig. 14, positive and negative variable bias circuits are provided by transistor Qg and resistors R ₅ and R ₆ and transistor Q _5, respectively. and resistors R ₅ . as well as R ^. educated. Transistors Qg and Q ₇ and Q _6b and Q _7b form positive and negative error signal amplifiers, respectively. The transistors Q ₁₀ and Q 10b operate in the same way as the zero cross _{comparators of the circuit shown in FIG.} The operating points of these transistors are essentially set to shutdown. If the control point is set to the emitter side of transistors Q _5a and Qg _b in the circuits of Figures 12-14, an output signal without distortion will be provided.

The bridge circuit shown in FIG. 15, which is modified compared to the bridge circuit of FIG. 12, consists of the resistors R _1a> R _1b as well as R ₂ and is the star-delta conversion of the circuit of FIG. 12.

As explained in detail above, according to the invention it is not necessary to adjust any temperature insert compensating elements, such as a varistor or thermistor, and no reactive current regulation is required, since the pre-magnetizing direct current of the amplifying

the elements making up the circuit is kept constant. In addition, a constant direct current is provided voltage value is supplied immediately after the power is switched on. Thermal distortion is also eliminated. Furthermore, since the transfer characteristic of the circuit is linear, any crossover distortion is reduced. Furthermore, no switching distortion is caused, since an amplifier arrangement without Class B disconnection is used.

Claims (9)

Amplifier compensated by bias current Claims Single-ended push-pull (SEPP) circuit of the emitter filter type, characterized by first and second amplifying elements which are emitter follower type and connected and biased in a class B single-ended push-pull arrangement by first and second Loads interposed between output electrodes of the first and second amplifying elements and an output terminal, respectively, through first and second ones Signal summer with first, connected to an input terminal inputs and each to the Input electrodes of the first and second amplifying element connected outputs, through a first and second error signal amplifier means, each for the first and second amplifying element are provided and each of which has a first and a second non-inverting input terminal as well has an inverting input terminal, of which the inverting input terminals are each connected to the output electrodes of the first and second reinforcing elements vvvV * »' are connected, the first and second error signal amplifier means each connected to the second inputs of the first and second signal summer Has outputs, through first and second means, the level of one present at the input terminal Shift signal and between the input terminal and the first non-inverting input terminal the first and second error signal amplifier means are inserted, and by first and second Means that shift the level of a signal present at the output terminal and between the second non-inverting input terminals of the first and second error signal amplifier devices are inserted, wherein when amplifying a -between the inverting input terminal and the first non-inverting Input terminal of the first error signal amplifier device generated error signal by the first error signal amplifier device the second error signal amplifier means is connected between its inverting input terminal and its second non-inverting input terminal is amplified and at amplifying one between the inverting input terminal and the first non-inverting terminal of the second error signal amplifier device generated error signal by the second error signal amplifier device the first error signal amplifier means has an inverting input terminal between it as well amplifies the error signal generated at its second non-inverting input terminal. 2. Circuit according to claim 1, characterized marked that the tensions of each a signal level shifting device are the same. 3. Circuit according to claim 1, characterized characterized in that the first and second Load each include an ohmic load. 4. A circuit according to claim 1, characterized characterized in that the first and second Load each include a current mirror circuit. 5. A circuit according to claim 1, characterized characterized in that the first and second amplifying element each comprises a bipolar transistor. 6. Circuit according to claim 1, characterized characterized in that both the first and the second amplifying element comprise Darlington pair bipolar transistors. 7. Amplifier circuit, marked by an amplifying element, by a resistance bridge circuit which is inserted between an output electrode of the amplifying element and an output terminal and comprises a first, second, third and fourth resistor, the first and second of which are connected in series between the output terminal and the output electrode of the amplifying Elements are connected and the third and fourth resistors are connected in series with one another and in parallel with the first and second resistors, by an error signal amplifier device which supplies an error signal which is a difference between the levels of an input signal to the amplifying element and one of the
Bridge circuit represents detected signal, and by a bias changing means which controls a DC bias of the amplifying element in accordance with a Rollerdusgangssujna I generated by the error signal amplifying means. 8. A circuit according to claim 7, characterized in that the bias changing device a comparator with a first and second input connection, of which the one with a connection point between the first and second resistor, the other with the output terminal is connected, and a switching circuit which selectively connects the output of the error signal amplifier means connects an input electrode of the amplifying element in response to an output of the comparator, includes. 9. A circuit according to claim 7, g e k e η η ζ e i c. π η e t by one of the level of the applied between the input terminal and the error signal amplifier means Input signal shifting device.