GB1560124A - Optical fibre cable testing - Google Patents

Description

(54) OPTICAL FIBRE CABLE TESTING (71) We, STANDARD TELE PHONES AND CABLES LIMITED, a British Company of 190 Strand, London.

WC2 England, do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement: This invention relates to optical fibre cable testing, and is concerned in particular with examining an optical fibre in a cable consisting of a number of series connected sections in order to identify if there is a break in the fibre in any of the sections.

An optical fibre of an optical fibre cable can be examined for the presence of breaks by directing a light pulse down the fibre and examining the reflected light. In the special case, where there is perfect matching at both fibre ends and where there is no break or any other discontinuity, there will be no reflection. In general however some light is liable to be reflected at the launching end, some reflected at any break existing in the fibre, and, if there is no break, some reflected by the far end of the fibre. Each of these reflections will produce a seperate return pulse, and these will be seperated in time according to the different transit times for the light to propagate to the reflecting surfaces and back again.

Normally a break will reflect very little light back up the fibre in low loss modes.

Normally the majority of the light is divided between that which is scattered at the break and that which is launched into leaky or otherwise relatively highly attenuating modes. In a typpical glass fibre light will propagate at about 20 cm per nanosecond, and therefore a very short optical pulse and rapid response photodetector would be necessary to locate a fibre break with any great precision. If however the fibre under test contains splices or connectors there is the problem that the return signal from a fibre break close to a connector or splice may be masked by the potentially larger return signal from nearby connector or splice. In the case of multimode fibre the mere shortening of the pulse length and the response time of the detector is not the final solution to the problem of resolution because resolution will be limited by pulse spreading.Thus in a typical fibre a pulse will have spread to about 20 nanoseconds in propagating 1 km and back again. This corresponds to an uncertainty of 2 metres even for extremely high speed optical transmitters and receivers.

This invention is concerned with a method of break detection which still relies upon analysis of the reflected signal, but which looks for the absence of expected reflections rather than for the presence of extraneous reflections.

According to the present invention there is provided method of testing a cable containing at least one optical fibre, which cable is composed of a plurality of series connected sections, to determine whether there is a break in said fibre, and when there is such a break, to determine within which cable section it lies which method consists of directing a train of signal pulses of light into one end of said fibre and monitoring with a gated photodetector system the amplitude of the trains of return pulses produced by reflection at the joints between consecutive sections of the cable, wherein the time interval between consecutive signal pulses is longer than twice the transit time for light to propagate down the cable from one end to the other, and wherein the signal pulse width and gating width (the period within which the gate is open) are short enough to resolve return pulses reflected from the two joints in the fibre at the ends of the shortest section of cable, and wherein the timing of the gating is adjusted to bracket in succession the return pulses from the individual joints.

A feature of the invention is that no longer is it necessary to use extremely short duration light pulses, and hence a larger pulse energy is possible for peak-limited sources. Similarly the gating period does not have to be of extremely short duration, since all that is required is to be able to resolve between consecutive fibre sections.

Moreover since the resolution is relatively coarse it is possible within a reasonable time scale to average the response over a larger number pulses than would be possible with a finer resolution arrangement.

There follows a description of preferred embodiments of the invention. In this descritpiton reference is made to the accompanying drawings in which: Figure 1 depicts apparatus for detecting breaks in an optical fibre.

Figure 2 depicts the return signal produced when a pulse of light is directed into that fibre.

Fig 3 depicts how the gating of the apparatus of Figure 1 is adjusted in relation to the return signal of Figure 2, and Figure 4 depicts a modified version of the apparatus of Figure 1.

Referring to Figure 1, the output of a pulse generator 10 is used to drive a laser 11 and is also fed to a variable delay device 12.

The light from the laser is directed via a beam splitter 13 into one end of an optical fibre 14 of an optical fibre cable. Light reflected by the fibre is directed via the beam splitter 13 to a detector 15. The output of the photodetector is fed to an integrator 16 via a gate 17 that is triggered by the delayed signal from the delay device 12.

The optical fibre 14 is shown in Figure 1 as having a number of sections which are joined by connectors (or splices) 18a to 18d.

Also depicted is a break 19 between in the section between connectors 18c and 18d.

This break and the connectors will provide a return signal of the form depicted in Figure 2 with pulses 20a to 20d derived respectively from connectors 18a to 18d and pulse 21 derived from the break. Pulse 20b will typically be slightly smaller than pulse 20a on account of the extra attenuation associated with the extra path length. Similarly pulse 20c will typically be slightly smaller than pulse 20b. (In particular instances differences in coupling efficiency at the different connectors may mask this effect). In the absence of the break 19 the return pulse 20c, but because of the break it is very much smaller. This is because the break can be expected to attenuate severely the light reaching connector 18d and because the light that is reflected is again attenuated on its second passage through the break.

Referring now to Figure 3 the delay provided by the delay device 12 of Figure 1 is first adjusted to provide a gating period 30a bracketing the return pulse 20a. Then the delay is readjusted to provide a gating period 30b bracketing the return pulse 20b.

Within each of the gating periods 30a, 30b and 30c there is a relatively large return pulse, and hence the signal stored in the integrator 16 is signicicantly different when the delay is adjusted to provide the gating period 30d that brackets the much smaller return pulse 20d. In this way there is provided an indication that break exists between connectors 1 8c and 18d. The duration of the gating period is typically significantly longer than that of the light pulse in order to allow for the fact that the precise positions of the connectors may not be know. In order to improve signal to noise ratio it may be arranged that the integrator sums the return signals from a train of light pulses.Since the number of fating periods is usually relatively small, it may be convenient to shorten the measurement time by a distributive (box car) technique such as that dpicted in Figure 4. Here in place of the single delay device 12 single gate 17 and single integrator 16, there are sets of delay devices 12"122,123, gates 17l, 172, 173, ... and integrators 16l, 162, 163, so that the measurements of successive portions of the cable are made concurrently.

If desired, the gating periods bracketing the return pulses from the more remote connectors may be integrated over progressively more pulses in order to compensate for the increased attenuation suffered by these pulses. In one form of apparatus the outputs of the integrators are fed to individual threshold detectors and lamps which light when a predetermined threshold is exceeded thereby indicating the absence of a break in the associated section of cable.

For an optical cable composed of sections 500 metres long the light pulse typically has a duration of 100 nanoseconds and a gating period of typically between 1 and 3 microseconds. This gating period is short enough readily to resolve return pulses from consecurive connectors. The limiting pulse repetition frequency will however be determined by the length of cable under test. It will of course be necessary for the time interval between consecutive pulses to be greater than the time for light to travel from one end of the cable to the other and back again in order to avoid any ambiguity. For measurements made on cables with odd lengths between connectors the apparatus will need to be preset to meet the particular layout of the cable under test. In the case where the instrument is built into a system as a fault locator, the presetting may be carried out automatically, at an electronic circuit level, the equipment being used to detect changes in the pattern of received pulses.

Claims (4)

WHAT WE CLAIM IS

1. A method of testing a cable containing at least one optical fibre, which cable is composed of a plurality of series connected sections, to determine whether there is a break in said fibre, and when there is such a break, to determine within which cable section it lies which method consists of directing a train of signal pulses of light into one end of said fibre and monitoring with a gated photodetector system the amplitude of the trains of return pulses produced by reflection at the joints between consecutive sections of the cable, wherein the time interval between consecutive sections of the cable, wherein the time interval between consecutive signal pulses is longer than twice the transit time for light to propagate down the cable from one end to the other, wherein the signal pulse width and gating width (The period within which the gate is open) are short enough to resolve return pulses reflected from the two joints in the fibre at the ends of the shortest section of cable, and wherein the timing of the gating is adjusted to bracket in succession the return pulses from the individual joints.

2. A method as claimed in claim 1 wherein the return pulses are distributed between a set of integrators arranged such that each integrator receives return pulses from an associated one of the joints between consecutive sections.

3. A method as claimed in claim 2 wherein the integrators associated with more distant joints are fed with longer pulses of trains in order to compensate for the additional attenuation associated with a longer path length.

4. A method substantially as claimed in claim 1 and as hereinbefore described with reference to Figures 1, 2 and 3 or Figures 1, 2, 3 and 4 of the accompanying drawings.