WO2003098327A1

WO2003098327A1 - Method and apparatus for regulating electroabsorption modulators

Info

Publication number: WO2003098327A1
Application number: PCT/GB2002/004921
Authority: WO
Inventors: Andrew David Ellis; Gabriele Bonfrate
Original assignee: Corning Incorporated
Priority date: 2002-05-17
Filing date: 2002-10-30
Publication date: 2003-11-27
Also published as: AU2002337345A1

Abstract

An electroabsorption modulator (EAM) is regulated by obtaining a first signal representing the optical output power of the EAM (P, Do) and a second signal representing its photocurrent (Ip); combining the first and second signals to obtain a control function that has an identifiable value at the half-power point; and adjusting the bias value to bring the function to that identifiable value. The control function may be obtained by multiplying the first and second signals, in which case the bias voltage will be adjusted to maximise it; or, provided the signals are in such proportion to the variables they represent that their sum is substantially independent of the bias voltage, by computing the difference between the first and second signals, and adjusting the bias voltage until the difference is substantially zero. The second signal represents 1-x, where x is the optical output (P), and a correction factor varying with the bias voltage can be used to improve the accuracy with which it does so.

Description

Method and Apparatus for Regulating Electroabsorption

Modulators This invention relates to the field of optical communications, and in particular to a method and apparatus for regulating the operation of electroabsorption modulators (EAM's) .

Electroabsorption modulators are used to convert electrical digital data signals into optical pulses, and their efficient operation depends on regulating an electrical bias voltage to ensure operation on a part of the transfer characteristic curve of the device that is as nearly as possible linear. The curve is especially sensitive to the temperature of the device and the operating wavelength, and may also be affected by other environmental factors, such as the accumulation of electrostatic charges and aging, and so an active regulating circuit is required.

The optimum half-power bias point in an EAM does not, except in special cases, coincide with the maximum gradient of the curve or relate to it in any simple manner, and so it is not effective to seek the maximum gradient by injecting an audio frequency pilot signal and adjusting for the greatest resultant optical signal, as is the usual practice for Mach- Zehnder modulators.

The present invention is based on the recognition that the predominant absorption phenomenon is the production of a charge-carrier pair, and that the efficiency of this effect is relatively insensitive to bias voltage in the region of its optimum value. The photocurrent is proportional to the number of charge carriers available, and is thus almost directly proportional to (1-x) , where x is the optical output power. By obtaining also a signal proportional to x, it becomes possible by simple numerical manipulation to derive functions of x in which it is easy to identify the value x= 0.5 (in relation to its maximum value), which is a good approximation for the optimum bias point, at least when the modulator is not saturated by a high power optical input level, even for a realistic EAM. For better approximation and/or if operation at high saturation is required, corrections can be made for the efficiency and/or other characteristics of the EAM. The method of the invention thus comprises obtaining a first signal (Si = βx) representing the optical output power of the EAM and a second signal (S₂ = . (1-x) ) representing its photocurrent; combining the first and second signals to obtain a control function that has an identifiable value at the half-power point; and adjusting the bias value to bring the function to that identifiable value.

Depending on the nature of the function, the distinctive value may be a specific numerical value, such as zero, or it may be a value that is identifiable by reference to the function - for example a maximum or minimum value (where the first differential of the function is zero) or a well-marked inflection point (where its second differential is zero) .

Two specific functions are proposed, without intending to limit the invention.

The first specific function to be considered is x(l-x), which has a sharp maximum for x = 0.5, and when this function is chosen the method of the invention further comprises multiplying the first and second signals to obtain the control function; and adjusting the bias voltage to maximise that control function.

The second, and at present preferred, function is the difference between the first and second signals (Sι-S₂) , with the proviso that the proportionality constants (α and β) must in this case be equal when expressed in compatible units. It immediately follows, once this condition is satisfied, that

Sι+S₂ = α(l-x) + αx = a that is the sum of the two signals is constant, and that

S_x-S₂ = αx-α(l-x) = α(2x-l) which is zero for x = 0.5, irrespective of the value of . Effective equality of the proportionality constants is easily obtained by choosing appropriate component values in the sensing and computing circuits, and can be verified by checking the constancy of the sum of the two signals.

Preferably the first signal is the electrical current output of a photodiode that receives a fraction of the optical output power of the EAM; alternatively, a short section of the EAM may be provided with a separate electrical contact and the current flowing through that contact used as an indicator of the output power level. Preferably the second signal is a voltage derived (directly or indirectly) by causing the photocurrent of the EAM to flow through an impedance .

If found necessary or desirable, a correction factor may be applied to either or each of raw data (current or voltage, in the preferred forms) to obtain the first and/or second signal. The most substantial effect giving rise to a need to use a correction factor is the internal quantum efficiency of the EAM, which falls off steeply as the bias voltage approaches zero, because of increased opportunity for charge-carriers to recombine, with the result that the measured voltage will begin to deviate from strict proportionality to (1-x) . It is anticipated that the voltage will be relatively insensitive to other parameters such as temperature and wavelength, and on this basis it is a routine matter to determine a calibration curve and design an analogue computation circuit or a digital look-up table or formula to generate an appropriate correction factor for ^• S₂ according to the absolute bias voltage presently being applied. It is thought that there will seldom be a need to use a correction factor for Si.

The first and second signals may be digitised before manipulation, and this may be -advantageous if complex corrections are desired: but analogue multiplier circuits are generally also satisfactory and operational amplifiers can be used to obtain the difference of two electrical signals.

The reasoning underlying the invention is only strictly applicable to an EAM operating at a low optical input power level; nevertheless, it is our expectation (based on limited experience) that the invention will achieve a useful approximation to the optimum operating bias even for relatively high optical power levels. The reasoning remains valid even for relatively high electrical drive amplitudes.

The invention includes apparatus for regulating an EAM comprising a first sensor circuit for obtaining a first signal representing the optical output power of the EAM; a second sensor circuit for obtaining a second signal representing its photocurrent; and a computer for combining the first and second signals to obtain a control function that has an identifiable value at the half-power point; and for adjusting the bias value to bring the function to that identifiable value.

The first sensor circuit preferably comprises a photodiode outputting its photocurrent; and the second sensor circuit preferably comprises a resistor through which the photocurrent of the EAM flows, outputting the voltage across it or a voltage derived from it by applying a correction factor dependent on the present bias voltage.

The computer may comprise an analogue or digital computer for multiplying the first and second signals and a peak-seeking circuit for adjusting the bias voltage to maximise that control parameter. Alternatively it may comprise an analogue or digital computer or an operational amplifier for obtaining the difference between the two signals and a control circuit for adjusting the bias voltage to adjust the difference to zero. The computer may also be used to apply any correction factor that is needed or desired, in which case raw data (such as the actual voltage across the resistor in the preferred second sensor circuit) will be input to the computer.

It also includes an EAM in operative combination with that apparatus .

One of the advantages of the invention is that no radio frequency processing is needed in the regulating circuits.

The invention will be further described, by way of example, with reference to the accompanying drawings in which: Figure 1 is a graph showing optical output power and photocurrent as a function of bias voltage for an idealised

EAM;

Figure 2 is a graph showing the optical output power and the product of optical output power and photocurrent for this idealised EAM;

Figure 3 is a graph similar to Figure 1 for a realistic EAM, with the addition of an internal efficiency curve;

Figure 4 is a graph corresponding to Figure 2 for the EAM of

Figure 2; Figure 5 is an outline circuit diagram of an EAM with regulating apparatus in accordance with the invention using the first computation;

Figure 6 is a calculated eye diagram for an EAM regulated by the method of the invention; and Figure 7 is an outline circuit diagram of an EAM with regulating apparatus in accordance with the invention using the second computation.

Figure 1 is calculated for an idealised EAM that has a

20dB extinction ratio, perfect conversion efficiency and a known small dark current, and plots as curve 1 the optical output power (P) and as curve 2 the photocurrent (I_p) in arbitrary units against the negative bias voltage. It is clear from inspection of the figure that, even assuming the result would be correct, it would require complex arid difficult computation to establish the point of maximum gradient (or minimum curvature) of either of these curves.

In contrast, the product (P x I_p) shown as curve 3 (with

P) in Figure 2, shows a clearly defined peak that is easy to locate accurately and which coincides with -the ideal bias "half-power" point.

Figure 3 is calculated for a more realistic EAM with an extinction ratio of 12dB, background absorption of 2% and an internal efficiency η which increases to a limiting value of 0.7 according to the formula η_v = η_m(l-e^vb) where η_v is the value of η when the bias voltage is V; η_m is the limiting value of η, and b is a constant known as the characteristic voltage, as shown as an additional curve in the graph. The substantial effects of these adjustments are to increase the level of P for -V > 1 and to depress the value of I_p, almost uniformly except for -V < 0.5, where the depression increases rapidly. In the region of the half-power point, there is little effect except for a near-proportionate reduction in I_p.

Thus, as seen in Figure 4, the product (P x I_p) still shows a sharp peak which, at about 0.49, is sufficiently close to the half-power point to satisfy the practical regulating requirement.

Figure 5 is an outline circuit diagram of apparatus in accordance with the invention: for simplicity of description, a fully analogue implementation has been chosen for illustration. Electrical digital data input Di is applied to the control electrode 5 of an electroabsorption modulator EAM which receives as optical input a continuous wave CW. The resulting optical output signal D₀ is monitored by taking a small proportion of it to a photodiode PD, by any convenient sampling technique (for example by a 95:5 splitter or a regulated microbending technique) ^• to obtain as the photodiode output a first signal (Si) proportional to D₀. The electrode 5 is negatively biased by a voltage source VS through a resistor R, and the voltage developed across R is taken to an operational amplifier 6 which thus produces an output proportional to the current flowing through the resistor, which is substantially the same as the photocurrent of the EAM. In accordance with one of the preferred forms of the invention, the voltage being applied from VS is also taken to an analogue computer f_e(V) which is designed to produce a correction factor which is a predetermined function of the voltage and provides a correction for the known internal efficiency characteristic of the EAM. A first multiplier Ml multiples this correction factor by the output of the operational amplifier 6 to obtain a second signal (S₂) which is an idealised value for the EAM photocurrent I_p, and a second multiplier M2 multiplies that idealised value by the output (Si) of the photodiode to obtain a control parameter, which is supplied to a second analogue computer C^Λ? which operates a conventional "maximum hunt" algorithm to adjust the voltage supplied by the source VS to continually maximise the control parameter. In this way the EAM is constrained to operate close to its half-power point as already described. Figure 6 is a calculated "eye diagram" obtained by superimposing the calculated optical intensity curves for a series of "0" and "1" pulses with appreciable rise and fall times for the output of an EAM of the kind to which Figures 3 and 4^' relate, regulated in accordance with the invention: although a little unsymmetrical, the eye is satisfactorily open and the eye crossing is, as predicted, at the half-light level. Because the extinction ratio used in the calculation of this diagram was poor, the eye opening is somewhat asymmetric. A modulator with this extinction ratio is close to the limit of acceptance for most applications, according to ITU recommendations, and so this eye diagram may be considered near a worst case. It may be noted that the values of photocurrent and output power used in the calculation of this diagram took into account the data modulation.

Figure 7 is an experimental plot showing the control parameter used in the invention (P x I_p) as a function of negative bias voltage for low optical power level (Δ,read against the left-hand scale) and at saturation optical power level (A, read against the right-hand scale) . Also shown are the known optimum bias values of the EAM for low power level (7) and saturation level (8) as measured at the output of the EAM and (for completeness) as measured after 40ms/m (ps/n ) chromatic dispersion has occurred (9 and 10 respectively) . It will be observed that (for local measurement) the invention has given a substantially correct bias value at low power level and a value correct within about 0.2 volt even at saturation. It is believed that sufficiently correct values could be obtained for post-dispersion measurement either by appropriately modifying the chirp parameter of the EAM or by using an asymmetric electrical drive signal to modify the chirp of the optical pulse. The form of the invention illustrated by Figure 8 is similar to that of Figure 5, and only the parts that differ will be described. The signal (Si) from photodiode PD and the signal (S₂) from the multiplier Ml (or if no correction circuit f_e(v) is to be used, direct from the operational amplifier 6) are fed to respective inputs of a second operational amplifier 11 which computes the difference between them and supplies it to an integrator 11 which adjusts the voltage source VS to obtain a difference always substantially equal to zero. Any discussion of the background to the invention herein is included to explain the context of the invention . Where any document or information is referred to as "known" _F it is admitted only that it was known to at least one member of the public somewhere prior to the date of this application . Unless the content of the reference otherwise clearly indicates, no admission is made that such knowledge was available to the public or to experts in the art to which the invention relates in any particular country (whether a member-state of the PCT or not) , nor that it was known or disclosed before the invention was made or prior to any claimed date. Further, no admission is made that any document or information forms part of the common general knowledge of the 'art either on a world-wide basis or in any country and it is not believed that any of it does so .

Claims

CLAIMS :

1 A method of regulating an electroabsorption modulator comprising obtaining a first signal representing the optical output power of the EAM and a second signal representing its photocurrent; combining the first and second signals to obtain a control function that has an identifiable value at the half-power point; and adjusting the bias value to bring the function to that identifiable value.

2 A method as claimed in claim 1 comprising multiplying the first and second signals to obtain a control parameter; and adjusting the bias voltage to maximise that control parameter.

3 A method as claimed in claim 1 in which the first and second signals are, when expressed in compatible units, in the such proportion to the variables they represent that their sum is^' substantially independent of the bias voltage, comprising computing the difference between the first and second signals to obtain the control function, and adjusting the bias voltage until that function is substantially zero. 4 A method as claimed in any one of claims 1-3 in which the first signal is the electrical current output of a photodiode that receives a fraction of the optical output power of the EAM. 5 A method as claimed in any one of claims 1-4 in which the second signal is a voltage derived (directly or indirectly) by causing the photocurrent of the EAM to flow through an impedance. β A method as claimed in any one of claims 1-5 comprising digitising the first and second signals. 7 A method as claimed in any one of claims 1-6 comprising using a correction factor reflecting the internal quantum efficiency of the EAM in obtaining the second signal. 8 Apparatus for regulating an EAM comprising a first sensor circuit for obtaining a first signal representing the optical output power of the EAM; a second sensor circuit for obtaining a second signal representing its photocurrent; and a computer for combining the first and second signals to obtain a control function that has an identifiable value at the half-power point; and for adjusting the bias value to bring the function to that identifiable value. 9 Apparatus as claimed in claim 8 in which the first sensor circuit comprised a photodiode that receives a fraction of the optical output power of the EAM.

10 Apparatus as claimed in claim 8 or claim 9 in which the second sensor circuit comprises an impedance through which the photocurrent of the EAM is to flow and a detector for measuring the voltage dropped by that impedance.

11 Apparatus as claimed in any one of claims 8-10 comprising a digitiser for converting the first and second signals . 12 Apparatus as claimed in any one of claims 8-11 comprising a circuit for applying a correction factor reflecting the internal quantum efficiency of the EAM in obtaining the second signal.

13 Apparatus as claimed in any one of claims 8-12 comprising a computer for multiplying said first and second signals and a maximum-hunt circuit adjusting the bias voltage to maximise its value.

14 Apparatus as claimed in any one of claims 8-12 comprising a computer (such as an operational amplifier) for determining the difference between said first and second signals and for adjusting the bias voltage so that the difference is substantially zero.

15 An electroabsorption modulator in operative combination with the apparatus claimed in any one of claims 8-14.