Method and system for an optical device.
Technical field
The present invention relates to optical communications systems and particularly to a method for stabilising a carrier density in a passive section of an optical device as described in the preamble of claim 1.
The present invention further relates to a system for stabilising a carrier density in a passive section of an optical device as described in the preamble of claim 28.
Background of the invention
The rapidly increasing demands of data traffic have made conventional transmission networks insufficient . A way of dealing with these demands has been the introduction of optical communication systems. Today, optical communication systems are widely used where high bandwidths are required, for example in long distance communications between countries and cities.
In many of these communications systems single wavelength semiconductor lasers are used as light sources for the communication signals which are to be transmitted via optical fibres .
It is desired that an optical device operates reliably for several years in order to be a useful component of an optical communication system, and a very important factor of optical communication systems is the ability of keeping the wavelength of the optical device fixed. Due to ageing the optical device degrades, i.e. characteristics of the optical device changes with time and the same current applied to the optical device
when just manufactured may tune the optical device to a different wavelength after a period of ageing.
A common type of single wavelength semiconductor lasers are lasers with a Bragg diffraction. One of these types is the distributed feedback laser (DPB laser) . The DFB laser consists of a single active layer with a grating distributed over the entire cavity.
Another kind of lasers with a Bragg reflector is the distributed Bragg reflector laser (DBR laser) . In the DBR laser the waveguide is divided into different sections, an active gain section with a gain material, a passive Bragg section with a non-absorptive waveguide material and a Bragg grating, and an optional phase section.
The DBR laser has several advantages compared to the DFB laser. The DBR laser has better tunability, higher output power, better spectral purity and better modulation characteristics. Above all, the better modulation characteristics of the DBR laser allows for higher bit rates and thus a better utilisation of the bandwidth.
Shift of wavelength due to degradation is not a problem in a DFB laser. The carrier density in the grating section is kept constant by clamping and the carrier density is independent of increased recombination in the grating layer as long as the optical power is constant. The optical power is typically held constant by monitoring the back facet power and adjusting the bias current to maintain a constant optical back facet power. This is called automatic power control (APC) . Even without automatic power control, the DFB laser is very stable as the carrier density is nearly constant above lasing.
The DBR laser however, suffers from the problem of long-term stability even with the use of automatic power control. In a DBR laser a current is injected to the Bragg section in order to set the Bragg wavelength. The current changes the carrier density in the Bragg section, which in turn changes the refractive index and thus the wavelength of the grating. The carrier recombination in the Bragg section, however, changes in time due to degradation, and since the carrier density in the Bragg section is not clamped by stimulated recombination as in the active section, the carrier density will change for a constant current to the Bragg section. The changed carrier density will change the refractive index of the waveguide, which in turn changes the Bragg wavelength and completely different spectral and modulation characteristics may be the result .
Hence, a critical issue for a DBR laser is to control the carrier density in the passive section.
In prior art DBR lasers have been stabilised by applying a slight modulation to the passive section and monitoring of the front output power. This method, however, is complex and requires extra optical elements and cannot be used together with high-speed modulation (see for example Larry A. Coldren, Scott . Corzine, "Diode Lasers and Photonic Integrated Circuits", fig. 8.3, ISBN 0-471-11875-3). As a consequence of this DBR lasers are not used as much as high-speed modulated light sources as would be the case if they were not as sensitive to degradation as they are.
A DBR laser may also be realised using a single active layer and two electrodes. In this case the Bragg section must be biased exactly to the transparency point in order to keep it
passive. This is very difficult to obtain without control of the carrier density.
A problem with DBR lasers is thus that they experience degradation due to ageing and therefore they are of limited use in systems where stable operation for a long period of time is required.
Summary of the invention
The object of the present invention is to provide a method that stabilises an optical device comprising an active section and at least one passive section, which method solves the above-mentioned problem.
This object is achieved by a method as defined in the characterising part of claim 1.
Another object of the present invention is to provide a system for stabilising an optical device comprising an active section and at least one passive section.
This object is achieved by a system for stabilising an optical device as defined in the characterising part of claim 28.
An advantage with the present invention is that an optical device with an active section and a passive section can be stabilised by controlling the carrier density in the passive section by extracting a signal from the at least one passive section and a reference signal from a signal fed to the active section. A low-pass filtered product of the two extracted signals is then compared with a set value, and a current to the passive section is controlled based on the comparison.
It is also an advantage with the present invention that it can be implemented without the need for any other optical components .
Another advantage with the present invention is that it solves the problem of long-term stability for a DBR laser and makes it possible to use the advantages of a directly modulated DBR laser in a communication system where stable operation for a long period of time is required.
A further advantage relating to DBR lasers is that the present invention also enables the use of a DBR laser of single layer type in systems designed to have a stable operation for a long period of time.
An advantage with an embodiment of the present invention is that the carrier density can be controlled at a desirable level, thus enabling tuning of the optical device to several operation points.
An advantage with an embodiment of the present invention is that the optical device may be tuned to a point where the bandgap absorption is zero, i.e. the transparency point.
Another advantage with an embodiment of the present invention is that part of the data signal fed to the active section may be used as the reference signal, thereby omitting the need for extra signal sources that may increase cost and complexity of the system.
Yet another advantage with an embodiment the present invention is that the carrier density can be represented by a signal representing the bandgap absorption, which signal can be easily measured as a differential current (AC signal) created by optical power resulting from modulation.
A further advantage with an embodiment of the present invention is that the use of a signal representing the bandgap absorption when the data signal is used as reference signal is enabled by filtering part of the signal extracted from the passive section and/or part of the reference signal through band-pass filters with a pass-band well above the inverse of a differential lifetime of the carriers in the passive section.
An additional advantage with an embodiment of the present invention is that in an alternative embodiment a sinusoidal signal can be used as a reference signal, thus omitting the need for pass-band filtering of the extracted reference signal and/or the signal extracted from the at least one passive section.
Another advantage with an embodiment of the present invention is that an optical device acting as a continuous wave laser with a separately biased passive section can be stabilised to a preferred operating point independent of degradation in the optical device.
It is also an advantage with an embodiment of the present invention that an optical device with an external modulator, such as an Electro Absorption or a Mach-Zehnder modulator can be stabilised using the present invention.
Further objects and advantages of the present invention will be apparent to those skilled in the art from the following detailed description.
Brief description of the drawings
Fig. la shows an example of a Distributed Bragg Reflector Laser with a passive and an active section.
Fig. lb shows an example of a Distributed Bragg Reflector Laser with a passive, an active and a phase section.
Fig. 2 shows a system for stabilising an optical device according to a first embodiment of the invention.
Fig. 3 shows a system for stabilising an optical device according to a second embodiment of the invention.
Fig. 4 shows a system for stabilising an optical device according to a third embodiment of the invention.
Fig. 5 shows a system for stabilising an optical device according to a fourth embodiment of the invention.
Fig. 6 shows a system for stabilising an optical device according to a fifth embodiment of the invention.
Fig. 7 shows a system for stabilising an optical device according to a sixth embodiment of the invention.
Detailed description of preferred embodiments
In fig. la is shown a conventional two-section DBR laser, which is used in a preferred embodiment of the present invention. The laser comprises an active section 1 and a passive section 2, each with a cavity. The active section 1, constituting a gain section, has a waveguide consisting of a gain material 3, and the passive section 2, constituting a Bragg section, has a waveguide consisting of a non-absorbing material 4, and a Bragg grating 5. Both the gain section and the Bragg section comprise an electrode 6 and 7, thereby enabling the sections to be individually biased with currents Igain 10 and Iβragg 11. The laser also comprises a front facet 9, from which the generated light is collected, and a back facet 15. Igain 10 is a current injected into the active section and
is used to control the carrier density, and thereby the optical gain. Iβragg H is a current injected into the passive section and is used to control the Bragg wavelength through carrier induced changes in the refractive index of the Bragg grating 5. The active section 1 and the passive section 2 may be built upon a common substrate 8.
Fig. lb shows a conventional three-section DBR laser, which is used in a preferred embodiment of the invention. In addition to the parts of the two-section DBR the three section DBR has a phase control section 12, and an electrode 13 making it possible to also bias the phase section separately. The three- section DBR operates in the same manner as the two-section DBR with the difference that the additional I hase 14 is an injected current, which is used to change the index by carrier induction in the phase section. The use of the phase section enables tuning of the laser over a broader range of wavelengths than the two-section DBR.
The claimed invention will now be described according to a first embodiment in conjunction with fig. 2, which shows a system that is capable of stabilising an optical device, such as the DBR lasers described with reference to fig. la and fig. lb.
In a common system with a laser such the ones described above the biasing current Iβra g is a constant DC current intended to keep the wavelength of the Bragg grating at a desired value. Due to degradation, however, as explained above, the characteristics of the passive section changes in time due to ageing and consequently a constant current IBragg does not keep the laser at the desired wavelength as time goes on even if an automatic power control is used where the back facet optical power is maintained at a constant value.
This problem with ageing is solved by the system in fig. 2, by low-pass filtering a product of part of a signal fed to the active section, which signal functions as a reference signal, and part of a signal detected from the passive section, and then compare the low-pass filtered product with a set value and adjust the DC current to the passive section based on the comparison. This feedback loop can be combined with a conventional automatic power control where the back facet optical power is maintained at a constant value by controlling the bias current to the gain section.
The system in Fig. 2 comprises a laser 20 with an active section 21 and a passive section 22. The laser also includes connections 23 and 24 for connecting biasing signals to the active section 21 and the passive 22 section respectively. The system further includes a data signal 25, which is connected to a converter 26 where it is converted into a signal format suitable for feeding the active section 21. The converted signal 35 is connected to the connection 23 on the laser 20.
The converted data signal 35 corresponds to the gain signal Igain in fig. la and fig. lb and is used both to bias the active section 22 and to directly modulate the laser 20. The signal 35 turns the laser 20 on and off at a very high frequency, modulation frequencies up to 25 GHz are possible, thus enabling a very high bit-rate. The resulting modulated light signal is collected from the front facet 9 and usually then transmitted further via an optical fibre (not shown) .
The system also includes a device 31 for comparing a set value 39 with a signal obtained from a multiplier 28, which signal will be described more in detail below. The device 31 also includes means for applying a DC current 40 to the passive section through the connection 24.
Part of the data signal 25 is extracted as a reference signal 29 in the converter 26 and is fed via a band-pass filter 27 to the multiplier 28. Also fed to the multiplier 28 is a signal 30, which signal 30 is extracted from the passive section 22 of the laser 20.
The signal extracted from the passive section 22 is a differential current superposed on the DC signal 40 fed to the passive section 22 via the connection 24. The differential current is a consequence of the modulation optical power in the cavity resulting from the signal 35, that is, when there is a gain in the passive section 22 current is consumed from the connection 23, and,when there is absorption in the passive section 22 current is generated. Consequently, a differential current is generated in the passive section 22, which differential current is superposed on the DC signal 40 fed to the passive section 22 and detectable on the connection 24 to the passive section 22.
The signal 30 is obtained by blocking, e.g. with a capacitor 32, the DC component corresponding to the signal 40 and bandpass filter the resulting AC component through a band-pass filter 33. The signal 30 is a representation of the bandgap absorption, and also the carrier density as now will be described.
The carrier density is dependent on the current to the section, the optical power and the recombination in the section, which changes due to degradation. This is the reason why a constant current to the passive section is not sufficient to keep the carrier density constant .
The relationship between the carrier density n and the optical power p at an operational point is explained by the following
differential equation, which describes the impact on the carrier density of a small change in optical power:
dn , .. n
— = a(n0)Ap— (1) at τ
a is the bandgap absorption and τ is the differential lifetime .
Transformation of the differential equation into the frequency domain and solving it for the carrier density gives the solution:
ap n = (2)
- + jω τ
As can be seen from the equation (2) the impact of changed differential lifetime r due to ageing of the optical device will be negligible for frequencies well above the inverse of the differential lifetime and for these frequencies the carrier density can be controlled by measuring the bandgap absorption. As can be seen in equation (2) , as long as the optical power p is kept constant there will be a direct relation between the bandgap absorption a and the variation in carrier density n. The differential current extracted from the passive section is directly proportional to the variation in carrier density. This means that we can stabilise the bandgap absorption from the extracted differential current. As the bandgap absorption is a direct function of the operating point carrier density n0, we will also stabilise the carrier density and hence the refractive index of the passive section. To keep the optical power at a constant level, a standard automatic power control can be used where the back facet power is monitored and the bias to the gain section is controlled.
This differential current is, as described above, fed to the multiplier 28 via a band-pass filter. The band-pass filtering of the reference signal and the signal representing the differential current is a consequence of the condition above that the studied frequencies are frequencies well above the inverse of the differential lifetime, and due to the fact that the data signal 25 and thus the modulation signal 35 are square wave like with frequency spectrums reaching over a very wide bandwidth, including low-frequency components. Both signals are band-pass filtered through filters with at least partly overlapping pass-bands, though the generated signal is strongest when the pass-bands are equal, and which pass-bands are at frequencies well above the above mentioned inverse of the differential lifetime in order to avoid the impact of the differential lifetime. Best system performance is obtained when the envelope delay over each pass-band is approximately equal. It may be advantageous to amplify at least the signal from the passive section.
The product signal 38 of the two signals 29 and 30 is a signal that, opposed to the signals 29 and 30 and due to the multiplication with like frequencies, has a mean value component that, except for the transparency point where the carrier density is zero, is greater than or less than zero. The signal 38 is low-pass filtered through low-pass filter 41 to obtain the DC component (mean value component) and then fed to the device 31 for comparison with a set value 39. Depending on whether the low-pass filtered signal 38 is greater or smaller than the set value 39 the DC current 40 to the passive section can be adjusted until the set value 39 and the DC current are equal . In this way a desired carrier density in the passive section 22 can be obtained, and thereby it is possible to keep the laser 20 at a desired operating point.
The set value may be obtained in different ways . It may be obtained by using a microprocessor controlled measuring process that simultaneously measures the signal 38 and the wavelength for all wavelengths the optical device may be tuned into. These measurements are then stored in for example a Read Only Memory (ROM) and when the optical device is to be tuned into a certain wavelength the corresponding set value is read from the ROM and the optical device may be tuned into the desired wavelength since the stored value of the differential current always will represent the same wavelength independent of degradation in the optical device. The measuring process is preferably carried out during trimming of the optical device prior to delivery.
Fig.3 shows a system according to an alternative preferred embodiment of the present invention. Also in fig. 3 there is a laser 20 with an active section 21 and a passive section 22. There is also a device 31 for comparing a set value 39 with a signal 46 output from a multiplier 28. The system further includes a converter 43 for converting a data signal 25 into a signal format suitable for feeding the active section.
The system in fig. 3 differs from the system in fig. 2 in that instead of using part of the data signal as reference signal as in fig. 2, a sinusoidal generator 45 is used to generate a sinusoidal signal as reference signal 42. Part of the sinusoidal signal is also connected to the converter 43 where it is mixed with the data signal 25 for forming a signal 44 to be fed to the active section.
The sinusoidal generator 45 preferably generates a sinusoidal signal of a frequency well above the inverse of the differential lifetime of the carriers in the passive section. In this way the need for pass-band filtering the reference
signal 42 may be omitted and the sinusoidal signal 42 may be directly connected to the multiplier 28, as shown in fig. 3. However, the band-pass filter may be used anyway in order to reduce unwanted noise and avoid over-drive of amplifiers, if any.
The signal 30 fed to the multiplier 28 is obtained in the same manner as in the system described in fig. 2. Here there is still a need for band-pass filtering the signal extracted from the at least one passive section since there still is an impact on this signal from the data signal 25 part of the signal 44.
The product signal 46 from the multiplier 28 is low-pass filtered through a low-pass filter 41 and compared with a set value 39 in a device 31 in the same manner as described above.
Fig.4 shows a system according to yet another preferred embodiment of the present invention.
In fig. 4 the laser functions as a continuous wave laser (CW laser) . That is, the laser operates as a light source that emits unmodulated light continuously at a specific wavelength.
In fig. 4 there is a laser 20 with an active section 21 and a passive section 22. There is also a device 31 for comparing a set value 39 with a signal 50 output from a multiplier 28. A sinusoidal generator 45 generates a sinusoidal signal 51 as reference signal .
In fig. 4 the data signal is a CW signal 52 is a continuous signal. In order to be able to detect a differential current in the passive section 22 a slight modulation must be applied to the continuous signal . Therefore part of the signal generated in the sinusoidal generator 45 is mixed with the CW signal 52 in the converter 53 that converts the CW signal 52
to a signal of a format suitable for feeding the active section 21. This slight modulation of the signal gives rise to the differential current in the passive section 22 needed to control the carrier density in the passive section 22. Since the differential current that results from the slight modulation by the sinusoidal signal also is a sinusoidal signal the band-pass filtering of the signal from the passive section to the multiplier may be omitted as shown in the figure, but as was described in connection with fig. 3 the band-pass filters probably will be used anyway.
The system in fig. 4 thus makes it possible to tune a laser functioning as a CW laser to an arbitrary operating point and keep it there with a stable lasing independent of degradation in the passive section.
Fig . 5 shows a system according to yet another preferred embodiment of the present invention where the laser also functions as a CW laser.
The system in fig. 5 is similar to the system in fig. 4 except that in fig. 5 the laser has an external modulator 60 connected to it. A data signal 61 is connected to a converter 62 where it is converted to a format suitable for feeding the modulator 60. The modulator may be an Electro Absorption modulator or a Mach-Zender modulator. Thus, it is also an advantage with an embodiment of the present invention that an optical device with an external modulator can be stabilised.
The above-described examples describe stabilisation of a laser with an active section and a passive section. In a laser with an additional passive section, such as the laser described in fig. lb with a Bragg section and a phase section, the stabilisation of the additional section may be carried out with an extra control-loop similar to the one described above,
since the two sections are likely to degrade in different manners. An example of this is shown in fig. 6, which shows the embodiment of fig. 2 for a laser with two passive sections .
The system in Fig. 6 comprises a laser 20 with an active section 21 and a two passive sections 22, 70. The laser also includes connections 23, 24 and 71 for connecting biasing signals to the active section 21, the passive section 22 and the passive section 70 respectively.
As in fig. 2 the system further includes a data signal 25, which is connected to a converter 26 where it is converted into a signal format suitable for feeding the active section 21. The converted signal 35 is connected to the connection 23 of the laser 20.
Part of the data signal 25 is extracted as a reference signal 29 in the converter 26 and is fed via a band-pass filter 27 to the multiplier 28 and 'a multiplier 74.
The passive section 22 is stabilised as described above in conjunction with fig. 2.
The passive section 70 is stabilised in a like manner with a second control -loop . Device 72 compares a set value 73 with a signal obtained from a multiplier 74 and applies a DC current 76 to the passive section through the connection 71 based on the comparison. The multiplier 74 multiplies the above mentioned reference signal 29 with a signal 75 obtained from the passive section 70, which signal is obtained by blocking, e.g. with a capacitor 77, the DC component corresponding to the signal 76 and band-pass filter the resulting AC component through a band-pass filter 78.
Stabilisation of a laser with two passive sections has here been discussed with reference to the in fig. 2 described embodiment. Stabilisation of a laser with two passive sections is however naturally applicable to all of the above described examples and any others within the scope of the invention. Also, combinations of the above described embodiments are possible, where one passive section is stabilised using one of the above described control-loops and a second passive section is stabilised using another of the above described control- loops .
In the general case where the laser can be stabilised at an arbitrary point, as described above, the set value has to be measured in advance. There is however, one special case where the set value always is the same and that is at the transparency point, i.e. where the bandgap absorption and hence the carrier density is zero. In this case only the sign of the low-pass filtered signal from the multiplier is of interest since a positive sign indicates that there is a gain in the passive section and the DC current fed to the passive section consequently should be lowered, and a negative sign indicates that there is absorption in the passive section and the DC current fed to the passive section consequently should be increased. Accordingly no pre-measurement of the set value is necessary when tuning the laser into the transparency point .
Tuning into the transparency point is applicable on all the above-described embodiments of the present invention.
In the above-described embodiments of the invention it may be advantageous to complement the control-loop with automatic power control (APC) , which controls the optical power in the optical device, when the optical device is to be stabilised at
an arbitrary operating point . The APC maintains the back facet optical power at a constant value by controlling the bias current to the gain section. An example of this is shown in fig. 7, which corresponds to the system described in fig. 2. The system in fig. 7 is complemented with a back facet monitor 80, which is connected to the converter 26. The back facet monitor 80 monitors the back facet optical power and converts an optical measurement to an electrical representation, which is fed to the converter 26. The converter 26 compares the measured value with a set value 81 and adjusts the biasing 35 of the active section 21 based on the comparison.
APC is known in the art and it is of course possible to use other embodiments of APC, including passive components, in order to keep the back facet optical power constant.
The present invention is also applicable to a DBR laser of a single layer type. The advantage of being able to use such a laser is the considerably less complicated manufacturing process .
Although the description mostly refers to a DBR laser, the claims should not be interpreted in a limited sense, but also include other types of lasers consisting of an active section and a passive section.