EP0912912A1

EP0912912A1 - Electro-optical phase modulator with direction-independent pulse response

Info

Publication number: EP0912912A1
Application number: EP97938830A
Authority: EP
Inventors: Günter SPAHLINGER; Manjeet S. Ner
Original assignee: Litef GmbH
Current assignee: Northrop Grumman Litef GmbH
Priority date: 1996-07-19
Filing date: 1997-07-18
Publication date: 1999-05-06
Also published as: WO1998003895A1; EA199900133A1; CA2253271A1; DE19629260C1; JP2000515260A

Abstract

An electro-optical phase modulator, in particular for use in a fibre-optical gyroscope (FOG), is characterised, as long as the time it takes for the light to make one revolution is made to match the working cycle, in that the same pulse response is ensured in both directions of rotation of the light. For that purpose, the modulation electrodes (13, 14...) are arranged in relation to a common counter-electrode (12) in such a way that the propagation in space and time of the potentials on the modulation electrodes and of the electric fields between the electrodes generates a pulse response which is always symmetrically distributed.

Description

Electro-optical phase modulator with direction-independent

Impulse response

The invention relates to an electro-optical phase modulator with an integrated optical waveguide and at the field side at a constant mutual distance from the optical axis along the waveguide arranged modulation electrodes.

Phase modulators of this type are primarily used in fiber optic Sagnac interferometers, which form the actual yaw rate instrument in fiber optic gyroscopes (FOGs = Fiber Optic Gyroscopes), or as a core element in other interferometric measuring devices, such as Mach-Zehnder interferometers.

The problem and task on which the invention is based will, however, be explained below with reference to a fiber optic gyroscope (FOG).

In the case of modern fiber optic gyroscopes, an integrated optical chip (IO chip) is often used, which usually has an integrated polarizer on the input side, then a Y-branch and two equally spaced along the optical axes after the Y-branching arranged electrodes of two phase modulators in a certain configuration, which modulate the two light beams radiated into the ends of a fiber coil in the opposite direction in a certain manner, which will be explained in more detail below. Various versions of such phase modulators or digital phase shifters are described in US 5 137 359, US 5 237 629 and US 5 400 142. An FOG with this type of phase modulator is sensitive to interference signals interspersed in the phase modulator.

The coupling of such interference signals into the MIOC path (MIOC = modulating IO chip) containing the phase modulator can be done as explained below. analyze.

Interference signals that couple into the MIOC path can cause bias errors under certain circumstances. In the following, it will be examined how periodic interference signals have an effect if the gyro sampling clock is mis-matched with the transit time of the light through the fiber. In addition to an increased sensitivity to such couplings, a misalignment also leads to further interference effects, such as, for example, an increased random mandrel walk. However, these effects should not be examined here. To give the reader the opportunity to familiarize themselves with the mode of operation of Sagnac interferometers with random modulation and a closed, resetting control loop, reference is made to European patents EP 0 498 902 and EP 0 551 537.

In order to be able to record the effect of interference coupling in the event of a mis-tuning, it is sufficient to look at the Sagnac interferometer with the control loop open (cf. FIG. 1). Let T be the sampling clock of the system and at the same time the period of a coupling interference voltage, T ₀ be the deviating transit tent of light, φ (t) be the phase modulation produced by the modulator and φs (t) be the Sagnac phase. Neglecting DC components and amplification factors in the detector path, the following applies to the output signal y (t) of the interferometer:

y (t) = cos (φ (t) - φ (t - T ₀ ) + φ _s (t »(1) If one now assumes that a suitable modulation voltage acting in the cycle T and superimposed on the signal φ (t) controls the turning points of the interferometer characteristic curve in a known manner, and the respective effective sign of the slope of the characteristic curve is generated by a cycle clock T also acting Demodulator signal is compensated, then the interferometer can be approximated by a characteristic curve

y (t) = Sin (φ (t) - φ (t - T ₀ ) + φ ₃ (t)) (2)

be described without modulation and demodulation signals. Strictly speaking, the approximation only applies to T = T ₀ . For T ≠ T ₀ , additional transients occur in narrow transition areas, which are not taken into account in the above equation. Since these transients only contribute to an increase in the random walk and to simplify the calculation, the validity of (2) is also assumed for T ≠ T ₀ , provided that the mis-match is not too great. A further simplification results from linearization of the sine function:

y (t) = φ (t) - φ (t - T ₀ ) + φ _s (t) (3)

This signal is filtered by a filter arranged in the data path and then sampled, the nth sample value y _n being calculated by a weighted averaging in the interval ((n-1) T, nT]. Weighting function is the impulse response h reflected on the time axis (t) of the filter.No contributions occur outside the interval, even if the impulse response does not disappear there, because due to the statistical modulation, demodulated signal components outside the interval mentioned are uncorrelated

T y _n = 1 h (t) y (nT - t) dt (4)

0

The function h (t) is o.B.d.A. so standardized that

T

J h (t) dt = 1 (5)

0

apply. The average yaw rate results from lim - _ y - LV -r ι * - (6)

So that is

h (t) φ, 'nT - t) d _t

(7) With sufficiently stationary signals, the averaging over a sequence x _{n is} independent of an index shift, ie it is

In the second integral of (7), the index n can thus be replaced by n + 1. Let ΔT = T ₀ - T be the clock detuning. Then it will be

1 lim h (t) (φ (nT - t) - φ (nT - Δ7 ^" - ι)) dt + / (r) φ, {nT - t) dt v— IV -r I T-! (9)

If the ΔT is sufficiently small, φ '(t) ΔT = φ (t) - φ (t-ΔT). With this approximation, we finally get

Let φ (t) = φ (t + nT) be a periodic signal with T. Furthermore let φ _s (t) = φ _s = const. Then

y = | h (t) φ '(- t) ΔTdt + φ _s (1 1) 0

B e i s i e l

As an example, assume that h (t) = 2 / T for t <T / 2 and h (t) = 0 for t> T / 2. For φ (t) apply in the range te (0. T) φ (t) = φ ₀ for t 6 [0. T / 4) vt 6 [3T / 4. T) and φ (t) = -φ ₀ for te [T / 4. 3T / 4). Outside the range ts (0. T), φ (t) is periodically continued according to φ (t) φ (t + nT).

y - T ( ₁₂ )

The measured phase y is therefore proportional to the relative detuning ΔT / T and the amplitude of the interference φo. Assuming 100 ppm (ΔT / T = 10 " ⁴ ) as the relative detuning and φ ₀ = 2π- 10" ² as the amplitude of the interference. then for a gyro with a 2π rotation rate of 2000 ° / s the bias error caused by the scattering is:

O = ^2∞0 ° ⁴ ₆ ^36C ° = 28.8Vh (13) h - 10

The invention is based on the object of an electro-optical phase modulator for fiber-optic interferometers. to create in particular for fiber optic gyroscopes in which the sensitivity to interfering interference signals observed so far has been completely or at least largely eliminated.

The starting point for the invention is the knowledge that the sensitivity of electro-optical phase modulators of the type mentioned can theoretically be reduced to zero if the time of the working cycle of the interferometer or gyro with the orbital period of the light from the first phase modulator via the fiber spool to the opposite one Phase modulator is brought into line. It has been recognized that this is necessary. To take measures to ensure that the phase modulator has the same impulse response in both directions of light travel.

The technical teaching of the invention can thus be characterized for an electro-optical phase modulator with an integrated optical waveguide and on both sides at a constant mutual distance from the optical axis along the waveguide arranged modulation electrodes in that the electrodes are arranged so that the spatiotemporal spread of the potentials on the electrodes and the electric field between the electrodes produce a symmetrically distributed impulse response. This basic idea of the invention is suitable for both analog and digital phase modulators when used in FOGs.

The preferred embodiment for a digital phase modulator is then that a plurality of pairs of electrodes which can be driven in parallel and which are linearly graduated in terms of their longitudinal extent are provided with a counterelectrode arranged between these binary-graduated electrodes, each binary stage consisting of two sub-electrodes and the points of symmetry of all binary stages matching, such that that the complete electrode arrangement generates a symmetrically distributed impulse response.

In the following, the conditions are derived and individual designs for phase modulators are explained which, according to the invention, provide the same impulse response in both directions of passage of the light.

With the designs of such phase modulators on integrated optical chips used today, particularly with the digital variants (cf. US Pat. No. 5,137,359), the symmetry condition derived below is generally not met. In the case of high-precision fiber-optic measuring devices, in particular in the case of FOGs, the necessary adjustment of the sampling clock time to the light transit time is therefore not possible.

The conditions for an ideal modulated Sagnac interferometer are first described:

With the ideal modulated Sagnac interferometer, the readout function on the photodetector after demodulation is omitted, regardless of the modulation method, as described above:

y (t) = αu (t) - αu (t - T ₀ ) + φ ₃ (t) (14)

Here u (t) is the reset voltage (or interference voltage) acting on the phase modulator, α the electro-optical transmission factor, φ _s (t) the Sagnac phase and T ₀ the light transit time from the center to the center of the phase modulators through the coil of the FOG. In the following u (t) = u (t + T) (15)

a periodic interference voltage with the operating cycle T. With this, and with φ _s = 0

y (t) = α (u (t) - u (t - T ₀ + T)) (16)

If T = T ₀ is ideal, y (t) = s 0.

The construction of a real fiber optic interferometer is in Flg. 1 of the accompanying drawings.

The light coming from a light source D is split at a Y branch Y into two parts, which then pass through the modulators m _j and m2, then through the coil S and then again through the two modulators m _j , m ₂ . The light rays are under a mutual phase shift

φ = φ _m + φ _s (17)

reunited, where φ _{m is} the phase generated by the modulators and φ _{s is} the Sagnac phase. Both modulators are driven by the same voltage u (t). Let the transit time of the light from the center of the modulator m _j to the center of the modulator m ₂ be T _Q. Then for the phase φ _m with opposite polarity of the two modulators:

φ _m = αf (/) * w (+ ou (/) * «(/) - αf (* u (t - T ₀ ) - α ₂ ^" (* u (- T _Q ) (18)

Here, α _n ⁺ (t) is the electro-optic impulse response of the modulator m _n (n = 1, 2) in the direction from right to left, while α _n ~ (t) is the electro-optic impulse response of the modulators for the direction of passage from left to on the right side there is. The asterisk * indicates the folding:

α (t) ^• u (t) = J α (τ) u (t - τ) dτ (19)

If the interferometer is now operated with T = T _Q and a voltage u (t) which is periodic in T is obtained, the result is > ^• (/) = (α ^* (/) +; {t) - a; (r) -: (t)) < _u (t) (20)

The invention and advantageous details are explained in more detail below with reference to drawings. Show it:

1 shows the basic structure of a real Sagnac interferometer which has already been briefly explained above;

2 shows an electro-optical phase modulator, according to the invention as a system with a distributed impulse response;

3 shows a schematic representation of the electrode arrangement for a basic solution variant of an electro-optical digital phase modulator with an optimally distributed impulse response according to the invention;

4 shows another basic embodiment of an electrode arrangement for a digital phase modulator, which corresponds to the symmetry requirements according to the invention with regard to the spatiotemporal spread of applied potentials; and

Figure 5A. B shows a schematic representation of the electrode arrangement of analog phase modulators according to the invention, FIG. 5A illustrating a mirror-symmetrical and FIG. 5B a point-symmetrical electrode arrangement.

Phase modulators which correspond to the invention are in principle constructed as shown in FIG. 2. The optically active area, characterized by the illustrated rectangle R., runs along the x-axis extending from right to left. its extent being limited by the interval [- o- * _Q ]. The control voltage u (t) now couples with an impulse response h (x, t) that is dependent on x into each point of the active region and generates an incremental phase shift there. Now if v is the speed of light propagation in the active area. the electro-optical impulse responses: and

αa; -α (t>) = Y hh _n (--xx ,, tt-- -)) ddxx (22)

- / ^■

The aim is. to make y (t) zero in (20). According to the invention, there are two possibilities for this. In the first case you choose

; - (23)

This is fulfilled for (21) and (22)

h _n (x, t) = h _n (-x. t) (24)

The second option leads to

a; = cς (5)

It follows

h _x (x. t) = h ₂ (-x. t) (26)

which also automatically

h ₂ (x. t) = h ^ -x. t) (27)

and thus

: - a (28)

is fulfilled. From this it follows that y (t) = 0.

The symmetry requirements for an electrode layout that meets the conditions of the invention are explained below: The phase modulator of the type according to the invention, in particular for a fiber-optic gyroscope, can be produced as an integrated optical component (chip), an optical waveguide being diffused into a suitable material, in particular Ll bθ3 or LiTaθ3. This waveguide has an optical refractive index that is dependent on an applied electric field. The necessary electrical field is generated by the electrodes arranged on the surface of the module parallel to the waveguide.

FIG. 3 shows an electrode arrangement corresponding to the first case explained above (equation (23)) for an active channel 1 of the symmetrically constructed pair of phase modulators m ^ or m ₂ . 2 denotes electrode connections for the binary-controllable electrodes of the digital modulator. Reference note 4 identifies a counter electrode which is assigned to both modulators m _j , πi2.

The distributed impulse response word h (x. T) is therefore dependent on the spatiotemporal distribution of the generated electric field. The symmetry requirements for the distributed impulse responses h _n (x. T) derived in the previous section can be met by symmetrical electrode arrangements on the phase modulator, whereby the spatiotemporal spread of the potentials on the electrodes must also satisfy the symmetry requirements. This applies both to digital (FIGS. 3 and 4) and to analog modulators (FIG. 5A. B).

In the case of digital modulators (see Flg. 3), a separate electrode arrangement is provided for each bit value, the weightings being realized by corresponding area ratios. The symmetry conditions must be fulfilled for each individual bit, the symmetry points of all bits having to match, so that in the end the symmetry condition for the complete electrode arrangement is fulfilled.

The following arrangements therefore result for digital modulators for the two solution cases mentioned:

In the first case, the modulators must have an electrode layout that is mirror-symmetrical to the axis x = 0, so that a spatio-temporal symmetry to this axis is also guaranteed. For a digital mo- dulator for example that in Flg. 3 layout shown. The optical waveguide is represented by the arrow running in the middle between the electrodes (active channel 1). Only one of the two modulators is shown, the other must be designed accordingly.

The second case leads to an electrode layout in which the electrode geometries of the two modulators are separated by rotating them by 180 °. Flg. 4 shows the basic illustration of such an electrode layout. In the schematic diagram, the active channel of the modulator is m ^ with 10, and the active channel of the modulator is m ₂ with 1 1. with 12 the common counterelectrode, with 13 the binary electrode array of the modulator m _j and with 14 the binary electrode array of the modulator m ₂ rotated in its arrangement.

The corresponding relationships for the mirror-symmetrical or point-symmetrical design of the electrode layout for analog phase modulators of the type according to the invention can be recognized directly by the person skilled in the art from FIGS. 5A and 5B.

The invention allows at least two types of electrode layouts for Sagnac interferometers. both from analog and from digital phase modulators, so that the influence of periodic interference signals is reliably suppressed with ideal work cycle tuning.

Claims

1. Electro-optical phase modulator with integrated optical waveguide and bilaterally at a constant mutual distance from the optical axis along the waveguide arranged modulation electrodes, characterized in that the electrodes (2; 13, 14) are arranged so that the spatiotemporal spread of the potentials generates a symmetrically distributed impulse response on the electrodes and the electrical field between the electrodes.

2. Electro-optical phase modulator according to claim 1, characterized by a plurality of pairs of parallel controllable electrodes, graduated in terms of their area ratio, and a counter electrode arranged between these binary graduated electrodes, each binary stage consisting of two partial electrodes and the points of symmetry of all binary stages matching, such that the complete one Electrode arrangement generates a symmetrically distributed impulse response.

3. Arrangement with at least one electro-optical phase modulator according to claim 1 or 2, characterized in that two phase modulators are integrated together with a Y branch in an IO chip.

4. Electro-optical phase modulator according to claim 3, characterized in that the arrangement of the modulation electrodes of the one phase modulator (mi) is rotated by 180 ° relative to the arrangement of the modulation electrodes of the other phase modulator (m ₂ ).

5. Electro-optical phase modulator according to claim 3 or 4, characterized in that a polarizer is additionally integrated in the IO chip.

6. Electro-optical phase modulator according to claim 5, characterized in that the polarizer has been generated by proton exchange technology together with the integrated waveguide in a LiNbOß or LiTaθ3 substrate.

7. Use of an electro-optical phase modulator according to one of the preceding claims, characterized in that this is part of the optical structure of a fiber optic Sagnac interferometer.