GB2092743A - Apparatus for Measuring Attenuation in an Optical Fibre - Google Patents

Abstract

The apparatus comprises a section 12 which generates and transmits optical pulses to the fibre 11, a beam splitter 14 and photodetection, amplification and signal processing devices 15. The detector 15 contains an integrator circuit 40a, connected downstream from a photodetector 21a, and which supplies an output signal which is the time-integral of one electrical signal, from the photodetector 21a. The pass-band of the integrator circuit 40a may be changed by altering its reactive component values, with consequent optimisation of the opposing factors of maximum inspectable fibre length and spatial resolution. The use of two similar circuits 16a, 16b, with one photodetector 21a irradiated by backscattered light from the fibre 11 and the other 21b shielded from light, enables the subtraction of spurious signals by a differential amplifier 20. A logarithmic amplifier 34 enables a linear output to be generated at terminal 37. <IMAGE>

Description

SPECIFICATION Apparatus for Measuring Backscatter Resulting from Attenuation in an Optical Fiber The present invention relates to an apparatus for measuring the attenuation to which an optical signal propagating along an optical fiber is subject, which measuring is carried out according to a measuring technique which is known as the "backscattering method" or "optical reflectometry in the domain of the time (OTDR)".

As is known, the said method consists substantially in throwing a light pulse into the fiber and analyzing the optical power diffused backwards (backscattered), which originates an optical return signal having a theoretical decreasing exponential behaviour. By carefully considering the thus obtained signal it is possible to evaluate a plurality of characteristics of the said fiber, such as for example: the value of the attenuation constant, eventual variations of the value of the attenuation constant as a function of the longitudinal coordinate in case a connection is obtained by cascade-connecting more fibers to each other, failures which have taken place during and after the installation, attenuations concentrated in well localized positions such as for example the junctions, etc.

The presently available apparatuses based on the backscattering method substantially comprise: a section of generation and transmission of the impulsive optical signal; an optical power splitter which allows simultaneously to transmit the impulsive optical signal generated by the optical fiber and to receive the backscattered optical signal and a detection section generally comprising a photodetector (exposed to the backscattered optical signal transmitted to the photodetector itself by the power splitter), an amplifier and, conveniently, a data procurement system.

The optical signal generation and transmission section is formed generally by an electronic circuit which pilots in impulsed condition a photoemitting element (for example a semiconductor type LASER) with predetermined period of repetition and pulse duration; generally, this is not connected with any relevant disadvantages.

The optical power splitter may be made from any optical directional coupler which, in this specific case, is traversed in two directions by the optical signal, and namely, the pulse generated by the transmission section passes from the optical source to the fiber, and the backscattered optical signal passes from the fiber to the photodetector.

Actually, to the backscattered optical signal may be added a spurious optical signal formed by light reflected from the inlet surface of the optical fiber during the short time of transmission of the optical pulse. Whilst the backscattered signal has an average power which is by 50 dB lower than the peak power of the signal introduced into the fiber, the spurious signal, deriving directly from a reflected signal, can have a power up to a level which is by 35 . 40 dB higher than that of the corresponding backscattered signal.Such phaenomenon produces an immediate saturation of the receiving section and renders it impossible to have trustworthy data at least as long as the photodetector and the amplifier do not come out from the zone of saturation; therefore, an initial fiber section remains unscanned, whose length, of at least some ten meters, is, anyway, not evaluable with precision. At present time, this disadvantage is avoided by disconnecting in a synchronous manner the electric supply to the input of the receiving section (which is not always easy to be done when the voltage levels are high), or by trying to obtain the best possible optical adaptation between the power splitter and the optical fiber, or by suitably taking advantage of the polarization of the incident radiation and the backscattered radiation.Presently, neither the technical solutions described hereinabove, nor their combinations are completely exempt from disadvantages.

The photodetector used is generally selected among a photomultiplier tube and an avalanche photodiode (APD). The photomultiplier may be used also without any amplifier situated downstream, since it supplies by itself a signal directly detectable by the data procurement system and having characteristics of high sensitivity with wide pass-hand. However, the photomultiplier has the disadvantages consisting in that it requires high focusing voltages for its operation and that it is somewhat cumbersome and fragile. These disadvantages are particularly felt inasmuch as the apparatus may in many cases be used for tests on the field, where the transportability (weight, overall dimensiones and sturdiness) is an essential requirement.The avalanche photodiode must always be followed by the amplifier and has the disadvantage that it always has to be thermally stabilized owing to its high sensitivity to ambient temperature variations; on the other hand, the use of a p-i-n diode photodetector is not to be proposed because the signal-noise ratio which would be obtained would be unacceptable, since the high spatial resolution required by the measuring involves very high values of the pass-hand width of the receiver (100 MHz). The term "spatial resolution" is to be intended as the possibility to distinguish from one another two signals coming from adjacent fiber sections.

The amplifier must correspond to the burdensome requirement of having very high dynamics; this requirement results in being even more burdensome when we consider that to analyze an optical connection section with an attenuation in optical dB having a value A, the amplifier must have a dynamics in electric dB higher than 4 A. The dynamics of the detection section is limited in the upper part by the linearity of the amplifier; this limitation is rendered more burdensome by the fact that the backscattered signal has very wide band and therefore the amplifier cannot be of the "tuned load" type. In the lower part, the limitation of the dynamics is determined by the noise.In other words, when one evaluates the behaviour of the backscattered signal coming from the most distant region of the section, the result of the measuring suffers from an uncertainty connected with the ratio between the useful signal and the level of the noise generated in the detection section by the photodetector, by the amplifier and by their polarization; unfortunately, owing to the high pass-band value of the amplifier, the noise is high.

Because of the limits mentioned hereinabove, the apparatuse of the type described herein which are available at present time are not capable of measuring section attenuations considerably higher than 20 dB, whilst it would instead be highly convenient to have the possibility to measure the behaviour of the attenuation in sections having total attenuations of 40 dB or more.

The object of the present invention is to provide an apparatus for measuring the attenuation of an optical signal which propagates along an optical fiber, which apparatus, yet carrying out the measuring in accordance with the teachings of the backscattering method, will avoid the abovementioned disadvantages of the known apparatuses.

This object is attained by the present invention which provides an apparatus for measuring the attenuation of an optical signal propagating along an optical fiber, of the type comprising: a section of generation and transmission of an impulsive optical signal apt to be injected into the said fiber; a.Section for the detection of the optical signal coming from said fiber and for the conversion of said optical signal into an electric signal by means of a first photodetector; and an optical power splitter arranged to directionally connect said generation and transmission section to the said fiber and the said fiber to the said detection section; characterized in that said detection section comprises a first integrator circuit connected downstream of said first photodetector and which generates an output signal which is the integral in time of said electric signal obtained by the conversion of said optical signal coming from said fiber.

For a better understanding of the present invention a preferred embodiment thereof will now be described, by way of non limiting example, with reference to the accompanying drawings, in which: Figure 1 is a functional diagram, partially a block diagram, of an apparatus according to the theachings of the present invention; and Figures 2a, 2b and 2c are diagrams, on each of which there is diagrammatically shown the behaviour in time of electric signals taken in predetermined points of the diagram of Fig. 1.

Referring now in particular to Fig. 1, reference numeral 10 indicates generally an apparatus according to the teachings of the present invention, which apparatus is capable of measuring the attenuation to which an optical signal propagating along an optical fiber 11 is subjected. The apparatus 10 substantially utilizes, as measuring technique, the method known as "backscattering method" or OTDR, and in particular comprises: a section 12 of generation and transmission of optical signals of impulsive characterwith a given period of repetition; an optical signal focusing system 13; an optical power splitter 14 of the known type commonly used for carrying out measurings according to the method specified hereinabove; and a detection section 1 5 for detecting the optical signal backscattered by the optical fiber 11.

More in detail, section 12 is conveniently formed by an optical source 17, such as for example a semiconductor type laser piloted by the output of a suitable circuit. The emission by the source 17 of the optical signal towards the focusing system 1 3 is controlled by a periodical electric signal of impulsive type, supplied by a pulse generator, through a block 19 which imposes onto the impulsive signal present at the output of the generator 18 a time delay of predetermined duration (t). Conveniently, block 19 may be made with a line delay, or by means of one or more monostable trigger circuits. The behaviour of the delayed impulsive signal present at the output of the block 1 9 determines the behaviour of the corresponding optical signal emitted by the source 17 and is diagrammatically shown in Fig. 2a.In this Figure, in fact, there is shown, as a function of'the time "t", the behaviour of the current "A" of the output signal of the block 19, and "T" indicates the period of repetition of the pulses forming the signal itself.

Detection section 1 5 has two branches 1 6a, 1 6b, substantially equal to one another, whose outputs are connected to the inverting input (-) and the non inverting input (+), respectively, of a differential amplifier 20. Branch 1 6a comprises a photodetecting diode 21 a, for example of the p-i-n type, inversely polarized through a resistor 22a connected to a terminal 23a; this latter is, in its turn, connected, in a not shown manner, to a positive pole of a direct type electric voltage source.

On the diode 21 a there is focused, by means of a focusing system not shown, an optical beam 24 coming from the optical power splitter 14. Beam 24 is substantially formed by the sum of the optical signal backscattered by the fiber 11 and partially transmitted to the diode 21 a through the power splitter, and the spurious optical signal due to an undesired partial reflection of the impulsive signal transmitted by the source 17 onto the input face of the fiber 11 connected to the splitter 14 itself. Said optical signal determines within the diode 21 a the passage of a current D, whose behaviour in the course of the time is shown in Fig. 2b. In particular, 1max indicates the maximum value assumed by the current ZD and due exclusively to the contribution of said spurious signal. 1o, instead, indicates the initial value from which begins the decreasing exponential behaviour determined by the optical signal backscattered by the fiber 11. As can be clearly seen in Fig. 2b, the portion of the signal i, having an exponential behaviour has raggings due to lack of omogeneity and/or imperfections present in the fiber 11, or to slight unfitnesses between adjacent sections of the fiber 11. K1 indicates the initial time constant of the said signal close to a decreasing exponential function.

The junction point between the diode 21 and the resistor 22a is connected through a wide band and low noise amplifier 25a to the junction point between a first terminal of a condenser 26a and a resistor 27a, to the inverting input of an amplifier 28a having a high input impedance (conveniently with a J-FET input), and to to a first terminal of a resistor 29a. The second terminal of the condenser 26a is directly connected to earth, whilst the second terminal of the resistor 29a is connected to earth through a switch 30a whose actuation is controlled by the signal present at the output of the generator 18.In particular, switch 30a is is maintained open for the whole duration of the period of detection of the backscattered optical signal and may conveniently be formed by a suitable solid state electronic circuit, or by an electromechanical component or by any other technique.

The second terminal of the resistor 27a is connected to the output of the amplifier 28a and to the input of a suitable insulation stage 31 a.

Insulation stage 31a may be formed in anyway provided it has an output impedance notably constant at varying equivalent generator input impedance. By way of example, stage 31 a could be of the type with following, by conveniently utilizing a field effect transistor (FET) in the configuration with follower. Fig. 2c shows the behaviour in the course of time of a voltage signal (v,) taken at the output of the insulation stage 31a.

The output of the insulation stage 31 a is connected, as already said before, to the inverting input (-) of the differential amplifier 20 whose output may be connected, by means of a signal amplifier 32 having the function of an active band-limiting filter, to the input 33, conveniently the unverting input (-), of a suitablelogarithmic amplifier 34, of which an input 35, conveniently the non inverting input (+), is connected to a terminal 36; to this latter there is supplied, in a way not shown, a suitable continuous signal. The logarithmic amplifier 34 has an output connected to a terminal 37 which may be connected to a suitable data procurement system or a visualization system, not shown.

The condenser 26a, the resistor 27a and the amplifier 28a are dimensioned in such a manner as to define on the whole an integrator circuit 40a for the electronic signal i,.

Branch 1 6b of the detection section 1 5 has substantially the same structure as the branch 16a; therefore, the components and the circuits of the branch 1 6b are indicated by the same reference numerals, but accompanied by the suffix b instead of the suffix a. In particular, in the example shown in Fig. 1 the branch 1 6b differs from the branch 1 6a only as regards the input and output zones. In the input zone of the branch 1 6b it is noticed that the diode 21 b is accomodated within a container 41 which prevents any optical radiation reaching the photosensitive surface of the diode 21 b.In the output zone of the branch 1 6b it is noticed that the output of the insulation stage 31 b is connected to a first terminal of a variable resistor 42 which has a second terminal connected to earth and a cursor which is connected to the non-inverting input (+) of the differential amplifier 20.

The operation of the apparatus 10 will now be described by considering first, for the sake of simplicity, only the behaviour of the branch 1 6a of the detector section 1 5 and then the modifications of the behaviour which are due to the presence of the branch 16b, taking into account that the differential amplifier 20 amplifies the difference between the signals which are present at the outputs of the branches 1 6a and 16b. The description of the theoretical part concerning the backscattering, in the fiber 11, of the impulsive optical signal injected into this latter is omitted, since this theoretical part may be considered completely known.We only desire to point out that, as already said before, the optical signal backscattered and supplied to the diode 21 a has a theoretical exponential decreasing behaviour and exhibits in practice slight deviations from the exponential curve owing to the lack of homogeneity, imperfections, more or less marked unfitnesses in the section formed by the fiber 11. Should the section be formed by the cascade junction of a plurality of small differences between the respective attenuation constants, the backscattered signal would in this case be composed of more successive sections with decreasing exponential behaviours and having each an own value of the decay constant.

Assuming, for the sake of simplicity, to have a fiber with an attenuation per length unit of approximately constant value, the signal jD generated by the diode 21 a (branch 16a) has the behaviour shown in Fig. 2b. The integration in the course of time of the signal D by the integrator circuit 40a allows obtaining the signal v8 shown in Fig. 2c, which signal has a first almost vertical section, of amplitude v80, due to the integration of the reflected pulse (1rnax)' and a second section with increasing exponential behaviour. In particular, since the integral of an exponential function is still an exponential fuction with the same time constant (K1 in the specific case), the second section maintains, in the form of a time constant K1, the information pertinent to the attenuation constant of the fiber 11. By suitably dimensioning the value of the capacity of the condenser 26a it is possible to obtain through the integration of the signal a beneficial compression of the dynamics. The limit to said compression is imposed by the fact the value of the condenser 26a affects also the pass-band of the integrator circuit 40a, and this band has to be dimensioned as a function of the desired spatial resolution.

The output signal of the pulse generator 1 8 gives rise to the rapid discharge of the condenser 26a with a time advance having the value t1 (see Fig. 2c) relative to the transmitted optical pulse, closing the switch 30a and thus putting the resistor 29a (of conveniently low value) in parallel connection with the condenser 26a itself. As has been better explained in the foregoing, switch 30a remains open during the whole period of detection of the backscattered optical signal.

Since branch 16b has substantially the same structure as the branch 16a, it is deduced that the treatment of the signal beginning from the diode 21 b up to the output of the insulation stage 31 b takes place in a manner analogous to that described with reference to the branch 16a. In branch 1 6b the diode 21 is maintained in darkness, so that in this branch, only the darkness current of the diode 21 b itself is amplified.By conveniently adjusting the position of the cursor of the variable resistor 42 there is then possible to supply to the non inverting input of the differential amplifier 20 an electric signal which compensates, from the point of view of the average temporal value calculated on a period having a duration equal to the time constant RC of the integrator circuit 40, the corresponding spurious electric signal determined by the darkness current of the diode 21 a and supplied to the inverting input of said differential amplifier 20 through the branch 16a.

The signal which its present at the output of the differential amplifier 20 is supplied, through the amplifier 32, to the inverting input (-) of the logarithmic amplifier 34, to the non inverting input (+) of which there arrives the signal which is present at the terminal 36. The logarithmic amplifier 34 converts the difference between the input signals into a corresponding signal which, owing to the exponential behaviour of the voltage supplied to its inverting input, is linear with slope and proportional to the value of the attenuation constant of the fiber 11. To this end, the value of the reference voltage which is present at the terminal 35 has to be suitably adjusted.

From the analysis of the characteristics of the apparatus 10 formed in accordance with the teachings of the present invention it can be observed that all the disadvantages specified hereinabove are eliminated.

The most critical part of the measuring, just as with the known apparatuses, is the part which corresponds to the most remote section of the fiber 11, inasmuch as the backscattered signal coming therefrom is weaker and therefore less protected against the noise. However, the influence of said noise is very small relative to the known apparatuses, inasmuch as the integrator circuit behaves as a low-pass filter and therefore it eliminates the contributions to the noise outside its own pass-band. Accordingly, by restricting said pass-band one obtains an increase of the maximum inspectable length of the fiber, the precision remaining equal; as already said before, such restriction of the band involves a worsening of the spatial resolution.In certain cases it may be necessary to restrict the band even beyond the limit which would be imposed by the desired resolution, in order to obtain an information, even indirectly on rather long distances, about the attenuation of the end portion of a very long section. To this end, the integrator circuit 40 may be made in such a way as to render it possible to vary the pass-band; it will then be possible to utilize said integrator circuit with a more ample pass-band in order to inspect in detail short section, with a more restricted pass-band for the purpose mentioned hereinabove. Indicatively these variations mat be achieved by using different values of the resistor 27a or the condenser 26a. The pass-band of the integrator circuit 40b will conveniently be varied in the same manner.

The main disadvantage of the known apparatuses consisting in that the total measurable attenuation is limited by the dynamics of the receiving amplifier, results in being eliminated, inasmuch as in the apparatus 10 the range of the useful signal v, is no more determined univocally by that of the instantaneous backscattered optical signal. In particular, the range of v, may be notably compressed, even if the integrator circuit 40 has a gain higher than one; this is obtained by acting onto the values of the components of the circuit itself. Such possibility involves, the dynamics of the amplifier remaining equal, a considerable increase of the length of the inspectable optical fiber section.On the other hand, the apparatus 10 may conveniently be used for checking by a single measurement the total behaviour of the attenuation on an entire connection section formed by more fiber pieces jointed to each other.

The detailed check of the single junctions could, instead, be carried out during the exposure, piece by piece, by using any of the apparatuses available on the market.

The disadvantage due to the reflected pulse of the input face of the fiber 11 is practically eliminated, inasmuch as the useful signal v8 is no more affected by the instantaneous peak power of said reflected signal, but by its total energy which is small because of the brief duration of the pulse itself. On the contrary, in some cases the fact that the signal v8 which has to be processed is quickly increased from the value zero to a value VBO (as shown in Fig. 2) may be beneficial. It is clear, anyway, that in case of the contribution of VBO being, in percent, eccessive relative to the dynamics of the amplifier 32, it will always be possible to maintain such contribution within the limits of the desired value by simply adjusting the value of a suitable reference voltage to be substracted from the voltage VB.

It may be said that thanks to the restriction of the pass-band and the consequent improvement of the reaction signal/noise, it is no more absolutely necessary to utilize a photomultiplier or an avalanche diode (APD), but a p-i-n- diode may be used as a photodetector, with all the advantages deriving therefrom.

The wide band and low noise type pre- amplifiers 25a, 25b allow to conveniently amplify the electric signal to the respective inputs, without notably affecting the dynamics, since said electric signal is still of low level. Conveniently, amplifiers 25a, 25b could be made by using field effect transistors (FET), in fact, it is known that amplifiers made in this manner lend themselves very well for directly following the photodiodes 21 a, 21 b, if these latter, in their turn, are of the p-i-n type.

The provision of the insulation stages 31 a, 31 b immediately downstreal of the respective integrator circuits 40a, 40b prevents the initiation of parametric type oscillations due to the periodic opening and closure of the switches 30a, 39b.

Finally, it is clear that modifications and variations may be made to the apparatus 10 described hereinabove, without departing from the scope of the present invention.

For example, although being advisable at present time to use as photodetector a p-i-n diode, any other photodetector may be used, for example a photomultipliern an avalanche diode (APD), or also other devices. The photodetector 21 b may be also of a type different from that of the photodetector 21 a, provided it has such characteristics that, in absence of the optical beam 24, the signal at the output terminal 37 of the detection section 1 5 may be substantially of zero value. In particular, the diode 21 b may be of the p-i-n type, even if the diode 21 a is of the avalanche type.

The condenser 26, which in the present description has been considered as being a discrete component, could, under given working conditions, be physically formed by the sole stray capacity at the input of the amplifier 28.

The adjustment of the level of the signal generated by the branch 16b may be carried out by supplying to the photodetector 21 b a suitable optical radiation, constant in the course of the time, of conveniently adjustable level.

Ampiifiers 25a, 25b may be omitted; in this case, the photodiodes 21 a, 21 b are directly connected to the inputs of the respective integrator circuits 40a, 40b.

One or both circuits formed by the photodetectors 21 a, 21 b and by the wide band amplifiers 25a, 25b, may be substituted by phototransistors, inclusive the field effect transistors.

The control signal of the switches 30a, 30b could be substituted by an optical signal having a slight advance relative di the impulsive signal injected into the optical fiber 11. Since Fig. 2 clearly shows that during the operation of the apparatus a periodic condition is established, it is obvious the instead of a delay of the pilot signal of the optical source 1 7 relative to the control signal of the switches 30a, 30b, it is possible to have a delay of this latter relative to the former. It is also clear that such a relative delay could be obtained by acting onto the optical part, for example onto the focusing system 13, rather than onto electric part of the apparatus, without departing from the scope of the present invention.The use of the resistors 29a, 29b and of the switches 30a, 30b for giving rise to rapid discharge of the condensers 26a, 26b is not strictly necessary for the operation of the apparatus 10 and may be substituted by other suitable circuits or devices.

Conveniently, each resistor 27a, 27b could be substituted by a plurality of resistors having a progressively decreasing resistive value and apt to be selected by means of a suitable commutator.

The insertion of decreasing resistive values would lead to a corresponding increase of the pass-band of the integrator circuits 40a, 40b, so that, on conveniently short sections, the apparatus 10 would allow to identify even small localized defects. Exactly the same may said as regards the condensers 26a, 26b.

The apparatus 10 could be provided with a further commutator, by means of which it should be possible the eliminate the integrators 40a, 40b. In this case, the apparatus would behave substantially as any of the known apparatuses available on the market at present time. This may be conceived as a particular case of what has been said in the preceding paragraph.

The diagram shown in Fig. 1 could conveniently be modified by inserting a drift circuit downstream of each integrator circuit 40a, 40b or downstream of the differential amplifier 20. In such configuration, the first amplification of the signal ZDW at low level, would be carried out so as to minimize the introduced noise. The signal obtained at the output of the drift circuit would, therefore, have the same behaviour as the signal jD' butasignal-noise ratio better than that which would be-obtained with a wide band amplifier.

The diagram of Fig. 1 could be further modified by inserting therein commutation means arranged to insert, disinsert at will the said drift circuit.

Finally, the behaviour in the course of the time of the- envelope of the transmission pulse could conveniently be modified to meat any requirement whatever.

Claims (28)

Claims

1. An apparatus suitable for measuring the attenuation of optical signals propagating along an optical fiber, and which includes: a device for the generation and transmission of optical pulses; a device which serves to transmit the pulses from the generation and transmission device to the fiber end and to direct backscattered light from the fiber to a detection system comprising a photodetector and an integrator circuit, which is connected to the photodetector, and which integrates with respect to time each electricai pulse generated by the photodetector.

2. An apparatus as claimed in claim 1, characterized in that said detection section (15) comprises a second photodetector (21 b) not exposed to any optical radiation and a differential stage (20), both of them being connected in such a manner that the darkness current of said second photodetector is processed as is processed said electric signal generated by said first photodetector (21 a), and said differential stage (20) carries out the difference between said two electric signals resulting from said processings.

3. An apparatus as claimed in Claim 2, -characterized in that said differential amplifier (20) has inputs (+, -) connected, respectively, to outputs of a first and a second branch (1 6a, 1 Sb), which also are part of said detection section (15), each of which begins, respectively, with the said first and second photodetector (21 a, 21 b).

4. An apparatus as claimed in Claim 3, characterized in that into said first branch (16a) there is inserted said first integrator circuit (40a) and that into said secondbranch (16b) there is inserted said a second integrator circuit (40b).

5. An apparatus as claimed in Claim 4, characterized in that said first and/or second integrator circuit (40a, 40b) comprise an amplifier (28a, 28b) and use, as integrating capacity, the capacity of at least a condenser (26a, 26b) connected to the input of said amplifier (28a, 28b)-.

6. An apparatus as claimed in Claim 4, characterized in that said first and/or second integrator circuits (40a, 40b) comprise an amplifier (28a, 28b) and use as integrating capacity the parasitic input capacity.

7. An apparatus as claimed in Claims 5 or 6, characterized in that said amplifier (21 a, 21 b) is of high input impedance.

.

8. An apparatus as claimed in any of the Claims from 5 to 7, characterized in that inserte in parallel to each said untegrating capacity is a branch comprising a switch (30a, 30b) whose actuation takes place with the same period of repetition as said impulsive optical signal.

9. An apparatus as claimed in Claim 8, characterized in that the actuation of said switch (30a, 30b) is electrically controlled by means of an impulsive electric signal supplied by a pulse generator (18) pertaining to said section (12) for the generation and transmission of said impulsive -optical signal.

10. An apparatus as claimed in Claim 9, characterized in that said pulse generator (1 8) is connected, through a delay circuit (19), to an optical source (17), and that said control signal of said switch (30a, 30b) is taken upstream of said delay circuit (19).

11. An apparatus as claimed in Claim 10, characterized in that said delay circuit (19) is formed by a-delay line or by one or more monostable trigger circuits.

12. An apparatus as claimed in Claim 8, characterized in that the actuation of said switch (30a, 30b) is controlled by means of an electric signal obtained by converting an optical signal coming from an optical source (17) pertaining to said section (12) of generation and transmission of said impulsive optical signal, or from said optical power splitter (14).

13. An apparatus as claimed in any of the Claims from 5 to 12, characterized in that said first and/or second amplifier (28a, 28b) comprise a feedback resistor which may be selected from a plurality of resistors through first commutation means.

1 4. An apparatus as claimed in any of the Claims from 4 to 13, characterized in that connected downstream of each said first and second integrator circuit (40a, 40b) is a respective insulation stage (31 a, 31 b) in which the value of the output impedance is substantially insensible for the variations of the circuit upstream, which correspond to the opening and closure operations of said switch (30a, 30b).

1 5. An apparatus as claimed in Claim 14, characterized in that each said insulation stage (31 a, 31 b) comprises a field effect transistor (FET) connected into the follower type configuration,

1 6. An apparatus as claimed in any of the Claims from 3 to 15, characterized in that a wide band amplifier (25a, 25b) is connected immediately downstream of each said photodetector (21 a, 21 b).

1 7. An apparatus as claimed in Claim 16, characterized in that each said photodetector (21 a, 21 b) is formed by a p-i-n type photodiode and that each said wide band amplifier (25a, 25b) comprises a field effect transistor (FET).

18. An apparatus as claimed in Claim 16, characterized in that at least one of said photodetectors (21 a, 21 b) is formed by a phototransistor.

19. An apparatus as claimed in any of the Claims from 3 to 18, characterized in that on at least one of said branches (16a, 16b) there provided gain adjusting means (42).

20. An apparatus as claimed in any of the Claims from 2 to 9, characterized in that inciding on said second photodetector (

21 b) is an optical radiation, constant in time and of adjustable intensity.

-21. An apparatus as claimed in any of the preceding Claims, characterized in that said detection section (15) comprises, as output amplifier, a logarithmic amplifier (34) connected downstream of said differential stage (20).

22. An apparatus as claimed in Claim 21 n characterized in that said logarithmic amplifier (34) has an input (35) on which is applicable a voltage signal apt to be substracted from the signal which said logarithmic amplifier receives from said differential stage (20).

23. An apparatus as claimed in Claim 21 or 22, characterized in that between said differential stage (20) and said logarithmic amplifier (34) there is interposed a linear amplifier (32).

24. An apparatus as claimed in Claim 23n characterized in that said amplifier (32) is a differential amplifier having an input on which a voltage signal of said differential amplifier (20) is applied.

25. An apparatus as claimed in any of the preceding Claims, characterized in that it comprises a drift circuit in any section downstream of said integrator circuit (40a).

26. An apparatus as claimed in Claim 25, as dependent on any of the Claims from 22 to 24, characterized in that said drift circuit is interposed between said differential stage (20) and said logarithmic amplifier.

27. An apparatus as claimed in Claim 26, characterized in that it comprises second commutation means arranged to disinsert said drift circuit.

28. An apparatus for measuring the attenuation of an optical signal propagating along an optical fibern substantially as described hereinabove with reference to the annexed drawing.