JPS6353489B2 - - Google Patents

Description

[Detailed description of the invention]

The present invention relates to an apparatus for searching for a fault point or measuring transmission loss in a communication optical cable by sending out an optical pulse from one end of the cable and observing the optical pulse reflected within the optical cable. When a pulse of light is sent out from one end of an optical cable,
Inside the optical cable, two types of reflected light are reflected toward the input end side. One type of reflection is called Fresnel reflection, which is caused by the difference in refractive index that occurs at the boundary between the glass that makes up the fiber core of the optical cable and the air from the end or break point. For example, the Fresnel reflected light from a plane cut perpendicular to the axis of an optical cable is 14 dB lower than the transmitted light. If the cut plane is not vertical or mirror-like, the reflected light will be smaller than this 14 dB. Another type of reflection is reflected light called Rayleigh backscattered light. This is isotropic scattering caused by minute fluctuations existing in the fiber core of the optical cable, and a small portion of it propagates toward the input end of the optical cable and is observed as backscattered light. This backscattered light is extremely weak light, and its level is about 45 dB lower than that of the incident light. This is called Rayleigh scattering return loss. The optical loss in one relay section of an optical transmission line for communication is designed to be approximately 40 dB, so when measuring the optical loss and connection loss in the relay section using Rayleigh scattered light, the Rayleigh scattering return loss (45 dB) + relay A dynamic range of section optical loss (40 dB) x 2 + S/N margin (30 dB)/2 = 140 dB is required.
No current optical pulse testing device has this level of sensitivity. In reality, it can be said that almost all the problems in configuring an optical pulse test device are contained in the fact that this dynamic range must be realized in the time domain. The main problems are: (i) the amplifier is saturated with high-level light due to Fresnel reflection, making it impossible to detect accurate values; and (ii) the minimum detection level is reduced due to induced noise within the measuring instrument. (iii) Due to setting a high amplification factor, drift in the photodetector and first-stage amplifier deteriorates measurement accuracy. The present invention aims to improve this problem and to provide a pulse test device that can avoid saturation of the amplifier due to strong reflected light. A further object of the present invention is to provide a device that can convert weak reflected light into an electrical signal and suppress its drift to a small level even when the electrical signal is passed through an amplifier with a high amplification factor. The present invention uses a photomultiplier tube equipped with a blanking electrode as an element for converting reflected light into an electrical signal, and by applying a pulse to the blanking electrode of this electron multiplier tube, strong reflected light is generated. It is characterized in that the output of the photomultiplier tube is reduced (including stopped) when the photomultiplier tube is incident. Furthermore, the second aspect of the present invention is characterized in that, in addition to the above configuration, the converted electrical signal is digitized, the digital signal is processed to be averaged, and the zero level is determined from the result of this averaging. A detailed explanation will be given below using the drawings of the embodiment. FIG. 1 is a block diagram of an apparatus according to an embodiment of the present invention. Optical pulses emitted from a pulsed light source 1 are transmitted via a directional coupler 2 from an optical connector 3 to one end of an optical cable 4 under test. As described above, light is reflected in the optical cable under test 4, and when it returns to the optical connector 3, it is separated by the directional coupler 2 and enters the photoelectric converter 6. The photoelectric conversion element of this photoelectric converter 6 is a photomultiplier tube, and the electrode thereof is configured so that a blanking pulse is applied from a pulse generator 7. The electrical signal obtained at the output of the photoelectric converter 6 is applied to the amplifier 10 via the variable resistance attenuator 9 and amplified, and the output is converted into a digital signal by the AD converter 11. This digital signal is digitally processed by an averaging circuit 12 to produce an averaged output. Further, another output of this averaging circuit 12 is given to a zero level detection circuit 14 to set a zero level in a state where there is no optical signal, and is combined with the above averaged output and converted back to an analog signal by a DA converter 16. It will be done. This output is displayed on the display device 17. The display device 17 preferably includes a logarithmic converter and is capable of displaying the received level in decibels on the time axis. Reference numeral 18 denotes a timing circuit that provides timing signals to each circuit of this device 20. Here, the features of the present invention are that a photomultiplier tube is used as the photoelectric converter 6, that a blanking pulse is applied to the electrodes of this photomultiplier tube, and that the averaging circuit 12 is constituted by a digital circuit, and the zero level is digitally set by the output of this averaging circuit 12. FIG. 2 is a structural diagram of the electron multiplier tube used in this example. This photomultiplier tube 22 is called an image desector, and the optical signal is sent to the lens 23.
The light is focused by the photocathode 24 and converted into photoelectrons. These photoelectrons pass between the deflection electrodes 25,
It is further configured to pass through the aperture 27 and reach the first dynode 28 . These photoelectrons are amplified by a large number of dynodes following the dynode 28, and are extracted from the anode 30 as an electrical signal. Here, a blanking electrode 31 is provided around the aperture 27, and is configured to prevent photoelectrons from reaching the dynode 28 by applying a voltage to this blanking electrode. Returning to FIG. 1, the blanking pulse output from the pulse generator 7 is connected to be applied to the blanking electrode 31. Further, the generation timing and connection time of this blanking pulse can be set arbitrarily during the time when the optical pulse is received by the photomultiplier tube 22. Note that the blanking pulse is not necessarily connected to the blanking electrode 31, but may be connected to the deflection electrode 25. To explain the operation of the apparatus configured in this way, FIG. 3 is a diagram showing an example of observation results displayed on the display device 17 in response to one sending pulse. The horizontal axis is the time axis, and multiplying this by the speed of light in the optical cable gives the round trip distance. The vertical axis represents the level of reflected light received and is a linear representation in this example. In FIG. 3, F ₀ is the reflection that occurs at the input end of the optical cable 4 and is of extremely high level. F ₁ is a Fresnel reflection that occurs in a connection at the midpoint (x point) of the optical cable 4, and has a high level. F ₂ is the Fresnel reflection that occurs at the far end. curves A and B
is backscattered light due to Rayleigh scattering, and the slope corresponds to optical transmission loss. N is the noise level. In such a received waveform, F ₀ and F ₂ have high levels and saturate the amplifier 10. _F0 and _F2
Although the time when the level of amplifier 10 is high is extremely short,
Once saturated, even if the signal disappears, it takes some time for the saturated state to recover, and subsequent observation of the received waveform becomes impossible. Therefore, in _the apparatus of the present invention, the time indicated by diagonal lines in _{FIG .}
During this period, a blanking pulse is sent to reduce the output of the photomultiplier tube and control the amplifier 10 to avoid saturation. More specifically, a blanking pulse is applied to the blanking electrode 31 of the photomultiplier tube 22, and photoelectrons are prevented from reaching the dynode 28 only during this application period. The blanking pulse is sent out from the pulse generator 7, which is connected to the timing circuit 18.
The pulse light source 1 sends out pulses at the times t ₀ and t ₁ shown in FIG.
It is convenient to have a configuration in which t 2 and t ₂ can be set arbitrarily and variably by manual operation. By setting these times correctly, the portion of curve A or B can be observed without being affected by the high level portion. FIG. 4 shows the relationship between the voltage applied to the blanking electrode of the electron multiplier and the amount of attenuation of the output voltage. This was done by setting the output voltage to 0 dB when a voltage of -1500V was applied to the blanking electrode, and measuring the output voltage change by changing the voltage of the blanking electrode. From this result, with a voltage shift of about 50V,
It can be seen that there is an attenuation effect of 100 dB in voltage attenuation, that is, 50 dB in light attenuation converted to incident light. The perfect edge Fresnel reflection is approximately smaller than the backscattered light.
Since it is 30 dB higher, the attenuation effect of 50 dB is a practically sufficient value. In reality, in order to attenuate by a certain time width at any point in time, the blanking voltage is usually set to about -1530V in this example, and only when attenuation is done, a pulse of -10 to -20V is superimposed. That's good. Changing the voltage applied to the blanking electrode in a pulsed manner has an effect on the electrical signal output. The measurement results are shown in FIG. This was measured by setting the blanking pulse peak value to 10V and varying the rise time of this pulse (step pulse). For 15V and 20V, it is calculated from the measurement results at 10V. From the above measurement results, the blanking pulse has a peak value of approximately 15V and a rise time of approximately
200nS was adopted. Since the zero point of a DC amplifier fluctuates due to drift, an error occurs when logarithmic conversion is performed. As shown in FIG. 6, consider a case where the amplified output of the waveform of backscattered light has an amplifier full scale amplitude (amplitude 1), and the waveform fluctuates by δ due to drift. The error ε (dB) after logarithmic transformation at the point of loss L (dB) (round trip loss) is is expressed. The results of this calculation are shown in FIG. The horizontal axis is the round trip loss L (dB) and the optical loss 2.8dB/Km
It is also expressed as line length x (Km) when 2.8×2×x=L. Since the optical loss is determined from the loss difference between two points, the loss measurement error due to the drift of the zero point is as follows. 2% drift: Measurement error 1.0-0.1 = 0.9dB 1% drift: Measurement error 0.55-0.04 = 0.51dB 0.5% drift: Measurement error 0.28-0.02 = 0.26dB 0.2% drift: Measurement error 0.12-0.01 = 0.11dB From this, it can be seen that for a drift of 0.2%, the optical loss measurement error for 2 km is 0.1dB, and extremely strict drift stability is required. For example, 0.2% of a 5V full-scale amplifier is 10mV, and with a gain of 40dB, the input-referred drift of the amplifier is 0.1mV, so measurement errors due to drift are considered unavoidable. In response to this drift, as shown in FIG. 8, samples (indicated by marks in the figure) are also performed for the zero level of the received waveform of the backscattered light, and averaged by digital processing. In other words, take the difference in digital values from the averaging results and use the result as
Perform DA conversion and then logarithmic conversion. With this method, it is possible to correct all the drifts in the stages before the AD converter. The zero-level sample point can be both before and after the optical pulse transmission (the rising point of the received waveform of the backscattered light), but since the sample point changes depending on the length of the observation cable afterward, the previous point may be the same. I'm going. In such a system, the band of the amplifier does not need to extend to direct current, and there is no problem even if the direct current is cut off. However, in order to accurately reproduce the received waveform, it is necessary to lower the low cutoff frequency of the amplifier to some extent. When the backscattered waveform is passed through an amplifier that cuts out the DC component, the areas of the diagonal line on the lower right and the area of the diagonal line on the lower left become equal, as shown in FIG. Figure 9a shows the case where the low cut-off frequency f _L is extremely low, and Figure 9 b shows the case where the low cut-off frequency f _L is on the order of KHz. In this example, the optical loss is measured to be large. Become. The relationship between the low cutoff frequency f _L and the error will be calculated below. For simplicity, consider a transient response to a step function. If the time constant of the CR differentiator circuit is τ, then the first
As shown in Figure 0, the step response is e ^-t/ 〓. The time constant of the circuit with a low cutoff frequency f _L is τ = 1/2πf _L. Considering the case of a 5 km fiber, the values at t = 50 μs are calculated as shown in the following table.

[Table] 10log ₁₀ e ^t/ 〓 is the error caused by cutting off the DC (round trip loss measurement error), so if the low cutoff frequency f _L is 10Hz or less, the error is 0.02dB or less, which is not practical for practical purposes. considered sufficient. In this study, the input waveform is a step function, but in reality it is an exponential function. In this case as well, the error is considered to be about the same. The high cutoff frequency f _H of the amplifier is determined by the conditions under which the transmitted pulse can pass through almost unchanged. Since the relationship between the high cutoff frequency f _H and the rise time t _r is t _r =0.35/f _H , when f _H = 10 MHz, τ _r = 35 nS; when f _H = 5 MHz, τ _r = 70 nS. On the other hand, since thermal noise included in a signal is proportional to the band, it is desirable that the band be as narrow as possible in order to improve the S/N ratio. Since there is no margin in the dynamic range for measuring backscattered light, the first priority is to improve the sensitivity, and from that point of view, we choose the one with less noise even if the rise is bad.
The high cutoff frequency _fH is approximately 5MHz. Therefore, τ _r is 70 nS, which is approximately the same value as the pulse half-width of 100 nS. As explained above, according to the present invention, it is possible to suppress high-level Fresnel reflections from being applied to the amplifier input by simply changing the voltage applied to the blanking electrode by -10V to -20V as a blanking pulse. , it is possible to prevent saturation of the amplifier with a simple configuration for controlling the voltage of the photomultiplier tube. Furthermore, since the zero point is determined by digital processing, it is possible to compensate for drift even if the amplification degree is extremely high. As a result, an optical pulse testing device was obtained that enables highly accurate measurements due to its high dynamic range. The results of using the apparatus of the present invention were very good, and it was possible to accurately observe not only the failure point of the cable but also the connection point and other conditions of the optical cable as it was laid.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of an apparatus according to an embodiment of the present invention.
FIG. 2 is a diagram showing an example of the structure of a photomultiplier tube (image disector). FIG. 3 is a received waveform diagram of reflected light.
FIG. 4 is a diagram showing the output voltage attenuation amount of the photomultiplier tube with respect to the blanking electrode voltage. Figure 5 shows the leakage voltage at the photomultiplier tube output terminal with respect to the blanking pulse rise time (the load on the photomultiplier tube is 10KΩ).
). FIG. 6 is a diagram showing the influence of amplifier drift on received waveforms. A is without drift, b is with drift. FIG. 7 is a diagram showing errors due to drift with respect to optical loss. FIG. 8 is a diagram showing sample points of zero level samples. FIG. 9 is a diagram for explaining the influence of the low cutoff frequency f _L of the amplifier on the received waveform. A is when the low cutoff frequency f _L is extremely low, and b is when f _L is several KHz. 1st
FIG. 0 is an explanatory diagram for calculating the amount of distortion of the received waveform due to the low cutoff frequency of the amplifier. 1...Pulse light source, 2...Directional coupler, 3...
...Optical connector, 4... Optical cable under test, 6...
Photoelectric converter, 7...Pulse generator, 9...Variable resistance attenuator, 10...Amplifier, 11...AD converter, 12...Averaging circuit, 14...Zero level detection circuit, 16...DA converter, 17...display device,
18... Timing circuit, 22... Photomultiplier tube, 23... Lens, 24... Photocathode, 25...
Deflection electrode, 27...aperture, 28...dynode, 30...anode.

Claims (1)

[Scope of Claims] 1. A first means for transmitting an optical pulse from one end of the optical cable under test, and a second means for receiving an optical signal generated when the optical pulse is reflected within the optical cable under test and converting it into an electrical signal. and a third means for analyzing the magnitude of the electric signal on the time axis, wherein a photoelectronic cable provided with a blanking electrode as the photoelectric conversion element of the second means is provided. A multiplier tube is used, and means for applying a pulsed voltage to the blanking electrode to reduce the output electric signal during a part of the time when the electric signal is being sent to the output of the photomultiplier tube. An optical cable pulse testing device characterized by: 2. A first means for transmitting an optical pulse from one end of the optical cable under test; a second means for receiving an optical signal generated by reflection of the optical pulse within the optical cable under test and converting it into an electrical signal; In an optical cable pulse testing device equipped with a third means for analyzing the magnitude of a signal on a time axis, a photomultiplier tube provided with a blanking electrode is used as a photoelectric conversion element of the second means. , comprising means for applying a pulse voltage to the blanking electrode to reduce the output electric signal during a part of the time when the output electric signal is being sent to the photomultiplier tube, and the third means and means for converting the electric signal into a digital signal, means for averaging the magnitude of the electric signal using the digital signal, and means for determining the zero level of the optical signal from the output of the means. Characteristics of optical cable pulse testing equipment.