FR2468254A1 - METHOD FOR MODIFYING THE GAIN OF AN AMPLIFIER ACCORDING TO THE VOLTAGE OF A SIGNAL AND CIRCUITS FOR CARRYING OUT THIS METHOD - Google Patents

Abstract

A method of changing the gain of an amplifier as a function of the voltage of a signal, amplifier which has a linear characteristic or a non-linear characteristic to cover a larger dynamic range, to compress or expand display or measurement values in a predetermined display or dynamic range; the voltage amplitudes u2 coming from an amplifier 8 of known embodiment are applied to an amplification circuit 6 having at least two different characteristics 1, 2 which are connected so as to present, for at least one predetermined transition voltage 4 , a continuous transition but with an inflection point, and to form, for display or measurement, voltage amplitudes u3 which have a different display or dynamic appearance in partial ranges of the applied voltage amplitudes u2.

Description

1.

  HOW TO CHANGE THE GAIN OF AN AMPLIFIER IN DIFFERENCE

SIGNAL VOLTAGE AND CIRCUITS FOR THE IMPLEMENTATION

IN THE WORK OF CFE PROCEDE.

The invention relates to a method for modifying the gain of an amplifier as a function of the voltage of a signal, an amplifier which has a linear characteristic or a non-linear characteristic to cover a greater dynamic range, to compress or dilate d values. or display

  measurement in a display range or dynamic predeter-

undermined. The invention also relates to circuits for

implementation of this Process.

In the field of electronics, we must amplify

  proud of the temporary appearing electrical voltages-

  display their peak value (amplitude), the dynamic range to be treated is often clearly

greater than 40 dB.

So, for example, in testing materials at

using ultrasonic pulses, the ultrasonic pulse

  re received is converted, using an electro-acoustic transducer, into a voltage pulse; this voltage pulse (signal voltage) is amplified and its amplitude is displayed and evaluated. In many cases, the display is made on a cathode oscilloscope, so that one has the duration and the amplitude of the pulse for the evaluation and the

representation.

  It is known to amplify the voltage pulses (signal voltage) in a linear or non-linear fashion, in particular logarithmically (J and H. Krautkrâmer,

WerkstoffprUfung mit Ultraschall, 3rd Edition, Edi-

  tions Springer 1975, p. 204 and following).

When amplifying with a linear characteristic, a signal voltage proportional to the sound pressure is amplified so that between the amplified voltage u2 and the voltage to be amplified ul, there is always a constant relationship called gain uv = - V - 2

which represents the slope of the linear characteristic.

The drawback of this type of amplifier is the low usable dynamic range corresponding to the ratio of the maximum and minimum amplitudes to be displayed of the signal voltage. During amplification with a logarithmic characteristic, the amplified voltage u2 is amplified

proportionally less when the trend

  sion to amplify u1 believes. We generally have U2 = C1 lgu1 + c2 Such a logarithmic amplifier can be

  designed so that its output voltage u2 corresponds to

  responds to the measurement in dB, we then have

u2 = 20 logu1 + c.

  The advantage of the logarithmic amplifier is that the usable dynamic range is large and that the display of amplitudes

voltage is in dB.

For many fields of application, such

  that the material test using ultra-pulses

the logarithmic amplifier presents the

  agrees that low signal voltages, which originate from disturbing influences or small reflectors which are not wanted to be taken into account, are displayed covering the higher signal voltages which need to be evaluated. If, for example, an electro-acoustic transducer delivers a voltage proportional to the sound pressure between 0.05 V and 5 V, corresponding to a dynamic range of 40 dB, when the ordinate axis of the cathode oscilloscope is subdivided in five

  parts, -the value of 5 V will correspond to the cin-

fourth graduation, and the value of 0.05 V will be displayed

chée at O dB corresponding to the 0 graduation. If, for the evaluation, the range from -20 dB to 40 dB, which corresponds to 0.5 V - 5 V, is particularly large, we only have 50% of the amplitude of the ordinates for this range, and the smaller impulses going from 0 to 20 dB, corresponding to

  0.05 V - 0.5 V, occupy the other 50%.

To suppress weak signal voltages, it is also known to design the amplifier so that it amplifies only input voltages which exceed a threshold voltage (J. and H. Krautkrâmer, Werkstoffprufung mit Ultraschall, 3rd Edition, Editions Springer 1975, p. 209). However, this method has the significant disadvantage that all the information contained in the signal voltage which does not exceed the

  threshold values are completely lost.

The object of the invention is a method which compresses

the dynamic range for low voltage pulses

  which expands it for large voltage pulses, without monotonic transition, so that the advantages of non-linear amplification, in particular

logarithmic, be kept.

  This problem is solved according to the invention thanks to the fact that the voltage amplitudes coming from an amplifier of known embodiment are applied to an amplification circuit having at least two different characteristics which are connected so as to present, for at least one predetermined transition voltage, a continuous transition but with an inflection point. and to form, for display or measurement. voltage amplitudes which have a different display or dynamic appearance in different

Dartiels ranges of the applied amplitudes of tension.

The invention will be better understood on reading

of the following description of two embodiments

  Referenced but not limiting and in which reference is made to the appended drawings in which: FIGS. the to represent features to be obtained by the method of the invention; fig. 2 is a diagram of an exemplary embodiment of a test device and a circuit for implementing the method of the invention; fig. 3 is a diagram illustrating the offset of the operating point of the circuit for implementing the method of the invention, and FIG. 4 is a diagram of a second mode of

realization of a circuit for the implementation of the pro-

assigned of the invention. The ultrasonic signal produced in a room W

is received, as shown in fig. 2, by a trans-

electroacoustic conductor 9, i.e. the sound pressure signal is converted into a signal

  of electrical voltage u1, possibly by the in-

through an adjustment element 10 graduated in dB.

  The signal voltage u1 is applied to the amplifier.

tor 8 having a logarithmic characteristic; if, as for example, in the prior art, the dynamic range is equal to 40 dB, when displaying

  the signal voltage using a cathoscope

dique 11, a range of 20 dB occupies 50% of the amplitude of the ordinates. If now, according to the invention, an amplifier 6 having two characteristics is placed between the amplifier 8 and the cathode oscilloscope II

  which connect at a point and which pre-

feel a weak amplification V1 = tgal = 0.4 up to a transition voltage 4 (fig. la) for characteristic 1, and a greater amplification V2 = tga 2 = 1.6 from the transition voltage

  for characteristic 2, and if the transition voltage

tion is at 50% of the dynamic range for example.

  5 V, all signals going from 0 to 20. dB, so

up to 2.5 V, at the output of the amplifier loga-

  rithmic are displayed on the graduations 0-1 of the ordinate of the cathode oscilloscope, and the signals going from 20 dB to 40 dB are displayed on the graduations 2 to 5. For signal voltages going from 0 to 20 dB, therefore only 20% of the scale is needed, while 80% of the scale is available

  for signal voltages from 20 to 40 dB.

  The choice of the value of the transition voltage

  is only a matter of adapting to basic needs

  ticks. If, as shown in fig. lb. range

  display ranges from 0 to 50 dB, and you want to

  take the signals from 0 to 30 dB on a gradation on the ordinate and display the signals from 30 to 50 dB on

  the graduations 1 to 5 of the scale, the cutting voltage

sition is chosen equal to 3V and the characteristics of the amplifier, which are not necessarily linear, are chosen correspondingly. We will therefore choose here a first characteristic with the gain Vla = tg a 1 = 1/3, and a second characteristic

with the gain V2a = tga 2a 2.

  The two characteristics are then designated

by la and 2a.

On the other hand, the method can be extended so as to provide another transition with another characteristic passing through a second transition voltage, as shown in FIG. the (here

characteristic 1b is an unrelated characteristic

  neary). Alternatively, the method can be implemented so that the transition voltages are predetermined so that it can be determined

  also evaluation criteria; for example,

  that at a first transition voltage 4 (fig. le), that is to say the characteristic lb highly compressed, are displayed the signals coming from the structure of the material (granulation noise), while between this transition voltage and a second transition voltage 5 (characteristic 2b) appear the signal amplitudes for an importance class

  secondary, and that the signal amplitudes which correspond

  lay, for example, faults and whose dynamic range requires a large part of the scale on the cathode oscilloscope are displayed with a third characteristic 3b. Fig. illustrates it a

such implementation. In this example, for the first

first range corresponding to the granulation noise, there is the first quarter of the first graduation (0 - 0.25) corresponding to 0 - 10 dB; for displays of secondary importance, there are graduations ranging from 0.25 to 2 corresponding to 10 dB - 30 dB, and, for important displays ranging from 30 dB to 50.dB, there are graduations 2 - 5. A suitable circuit for implementing the method is shown in FIG. 2 in the form

of a block 6.

  On input 13 of this circuit is applied the

signal voltage u2 delivered by the logarith amplifier-

mic 8. The operational amplifier 7 provides a

basic planning which is given by the elements of

  feedback R1, R2, R3 and R4, R3 and R4 being in pa-

rallies on R2, and the overall amplification of the ampli-

  operational ficitor is reduced to gain V1. The poten-

  tiometer R6 in block 15 makes it possible to adjust a voltage threshold 4 (fig. la - le) on the base of transistor 17 so that the collector-emitter circuit is

blocked. The output voltage u3 of the operating amplifier--

rational is also applied, via resistor R5 and diode D, based on the

  transistor. When the output voltage u2 of the amplifier

operational feedback factor exceeds the voltage threshold on the basis of the switching transistor, the collector-emitter circuit becomes conductive, as a result of which the feedback branch formed

by R3 and R4 becomes ineffective and the feedback-

  is no longer determined only by resistors R2 and R1,

  and the gain increases from the value V1 to a value V2.

  Thanks to another switching transistor 18, the poten-

  point 12 of operational amplifier 7 is so-

  offset simultaneously so that the two characteristics

ticks intersect at a desired point (fig. 3). If the potential of point 12 was not shifted in a predetermined manner, during the transition of transistor 17 to the conducting state, one would obtain a steeper characteristic which would cut the abscissa axis for any voltage value (for example OV ) (characteristic 2c of fig. 3). However, it is necessary that

the two characteristics intersect in point 4.

For this reason, when entering the conduc-

  of transistor 17, there is simultaneously a predetermined potential offset Au at point 12, whereby

  characteristic 2c is offset parallel to it

  even so as to cut characteristic 1 in point

4 to give characteristic 2.

  The device can be extended by means of several circuits, corresponding to block 15, connected in parallel. for which other voltage thresholds 5 are respectively set, and which increase by

  steps the overall amplification of the amplifier

  rational feedback. Such another connected circuit

in parallel is indicated schematically by block 16.

  In another arrangement shown in FIG. 4, at least two amplifiers 18a and 18b are connected

  in parallel, the amplifier 18a having the characteristic

  teristics 1 (fig. la) and, from a transition voltage 4, which, for example, can be determined by an arrangement of diodes 19, we pass to a second amplifier 18b having a higher gain, and the

output voltages of the two amplifiers are addi-

operated, for example in a network 20, so as to obtain a steeper characteristic from the transition voltage 4. The necessary transition voltage can be produced for example by means of

a Zener diode or a series of diodes 19 (event-

adjustable).

In the examples described, the amplifier 8 is

  an amplifier having a logarith-

  mique, as is currently the case for

non-destructive testing of materials using

ultrasonic impulses. However, this process can

also be used with a linear amplifier 8.

If, for example, you have to display or represent linear voltage amplitudes ranging from 0 to 5 V, the voltage range going from 0 to 3 V being compressed on a graduation and the voltage range going from 3 to V being expanded on graduations 1 to 5, for a dynamic range of 5 V the transition voltage must be chosen equal to 3 V. Characteristic 1 must

  then have a slope of tg, = 1/3 and the character-

téristique 2 a slope of tga 2 2.

The process according to the invention and the circuits

  for its implementation allow to compress

small amplitudes' of voltage or low respectively

voltage pulses, which appear domi-

especially during logarithmic amplification,

in a proportion which corresponds to the pra-

ticks and decompress or expand larger voltage amplitudes over a scale length

more important.

As it goes without saying, and as it results from elsewhere

  already from the above, the invention is not limited to

  ment to those of its modes of application and embodiments which have been more particularly envisaged; she in

embraces, on the contrary, all variants.

Claims (7)

1.- Process to modify the gain of an amplifier as a function of the voltage of a signal, amplification   tor which has a linear characteristic or a nonlinear characteristic for covering a greater dynamic range, for compressing or expanding display or measurement values within a predetermined display or dynamic range, characterized in that the voltage amplitudes (u2), coming from an amplifier (8), of known embodiment   are applied to an amplification circuit (6),   having at least two different characteristics (1,2) which are connected so as to present, for at least a predetermined transition voltage (4), a transi- continuous but with an inflection point, and to form, for display or measurement, voltage amplitudes (u3) which have a different display or dynamic appearance in partial ranges   applied voltage amplitudes (u2). 2.- Method according to claim 1, characterized in that the characteristics (1,2) are linear characteristics.   3.- Method according to claim 1, characterized in that the first characteristic (1) is a nonlinear characteristic, in particular a character- logarithmic teristics, and, by the fact that the second characteristic (2) is a linear characteristic. 4.- Circuit for implementing the following process any one of claims 1 to 3, characterized by the fact that a first channel (R2) and at least a second (RY, R4) feedback channels are connected in parallel to an operational amplifier (7) and determine a first amplification characteristic (1), and by the fact, that beyond a voltage threshold (4)   which can be pre-selected, the second counter channel reaction (R3, R4) is made ineffective and a second amplification characteristic (2) is determined by the fact that switching off the   second way of feedback is through diary of an electronic switch (17), and by the fact that at the same time as switching off the second feedback channel, an electronic controller (18) applies to the input (12) of the am -   operational.plifier (7) a bias voltage   value such that the second amplification characteristic (2) passes through the corresponding point at the transition voltage (4).   5.- Circuit according to claim 4, characterized in that one or more feedback channels (16) are connected in parallel to the existing channel (15), these channels becoming effective beyond transition voltages ( 5) pre-selectable but respectively more important and determining other   amplification characteristics (3b).   6.- Circuit according to claim 4 for the implementation of the method according to claim 2, characterized in that the feedback channels (R1, R2, R3, R4) are constituted by resistors ohmic and provide linear characteristics.   7.- Circuit for implementing the method according to claim 1, characterized in that at least two amplifiers (18a, 18b) having different characteristics are connected in parallel, that the response thresholds of the amplifiers correspond to different voltage values,   and by the fact that the output voltages of the ampli   suckers are added in a network (20) or a similar item.