Mach-Zehnder modulator bias and driver gain control mechanism

Details Mach-Zehnder modulator bias and driver gain control mechanism Mach-Zehnder modulator bias and driver gain control mechanism

2001-06-29

2004-03-02

Eckert, George (Department: 2815)

Coherent light generators

Particular beam control device

Modulation

C250S205000, C359S199200

Reexamination Certificate

active

06700907

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates in general to optical communication systems and components therefor, and is particularly directed to a microcontroller-based mechanism for performing closed loop control of both the bias and gain of a Mach-Zehnder (MZ) optical transmitter without the need for tuning or adjustment.
BACKGROUND OF THE INVENTION
FIG. 1
diagrammatically illustrates a typical external laser modulator architecture for a digital optical transmitter, such as that employed for use in a very high data rate (e.g., on the order of 10 Gb/s or higher) optical communication system. As shown therein the optical transmitter employs a laser modulator
10
, such as an X-cut lithium niobate (LN) Mach Zehnder (MZ) modulator, that is external to and disposed in the output beam path
12
of a continuous wave laser
14
. The LN MZ modulator
10
has a first, drive signal (data) port
11
to which a digital drive signal is applied, and a second, DC bias port
13
to which a DC bias voltage is coupled. The drive signal is supplied from the output
23
of an analog driver
20
, which has a signal input
21
coupled to a digital RF signal source
22
and a drive (gain) control port
24
coupled to receive a drive amplitude control voltage. The DC bias voltage is supplied from an output
33
of a bias controller
30
, which has a DC voltage input
31
coupled to receive a DC control voltage.
In an effort to sustain long-term, stable operation of the laser modulator
10
, compensation for environmental effects, such as temperature and aging (which is especially important for telecom and datacom applications), the DC control voltage to the DC bias controller
30
is coupled to a first reference oscillator tone-based closed loop control path
35
. This control path is used to monitor the beam output of the laser modulator, and to adjust the control voltage input to the bias controller
30
, as necessary, to ensure that the DC voltage bias necessary for proper modulator operation is coupled to DC bias port.
In addition, in order for the modulator drive signal to track changes in the modulation efficiency slope, a second, reference oscillator tone-based closed loop control path
25
is coupled to monitor the output of the laser modulator. This second closed loop control path is operative to controllably adjust the DC input to the analog driver
20
, in order to maintain a constant optical extinction ratio of the modulator.
Because this architecture employs tones/frequencies for signal drive control and DC bias adjustment, not only are different tuning channel circuits required, but they must include filter/isolation circuitry for reducing/minimizing interaction or cross-coupling of one control channel into the other.
For an illustration of non-limiting examples of patent literature describing prior art laser modulator architectures, including the tone-based optical transmitter control mechanism of the type employed in
FIG. 1
, attention may be directed to the following U.S. Pat. Nos.: 5,317,443; 5,742,268; 5,805,328; 5,917,637; 5,907,426; 5,400,417; 5,003,264; 5,343,324; 5,453,608; 5,900,621; 5,440,113; 5,170,274; 5,208,817; and 5,726,794.
SUMMARY OF THE INVENTION
In accordance with the present invention, shortcomings of conventional tone-based laser modulator control schemes, including those employed in the systems of the above-referenced patent literature, are effectively obviated by a microcontroller-based laser modulator control mechanism, that is operative to perform closed loop bias and gain control of the transmitter without tuning or adjustment. Independence from tuning is achieved by relying on the transfer function slope, or zero derivative at convergence, and the change in sign of the derivative for small perturbations in the control settings. Once converged, no further adjustment is required, and the control mechanism of the invention may be turned off for extended periods of time to save power. Advantageously, the period of time during which perturbations are created may be staggered in a random time fashion to achieve spectral dispersion of the control signals and provide insensitivity to periodic environmental noise.
As will be described, the invention executes a modulator bias and gain optimization routine that contains respective bias and gain control subroutines, that are executed in a time-interleaved manner, using feedback current extracted at the output of (a photodiode coupled to) the Mach-Zehnder waveguide. In accordance with this composite routine, once initial values for bias and gain have been set to ‘best guess’ parameters at the convergence point, a bias control subroutine is executed, followed by a gain control subroutine.
The bias control subroutine derives the peak of the sinusoidal Mach-Zehnder function, where the derivative is zero and the slope of an induced error signal has the correct sign. This bias control subroutine therefore depends on the shape of the transfer function and not on the absolute values of the control or feedback signals. The monitored feedback current from of a modulator output photosensor is sampled and stored as a “nominal” signal value, and the modulator gain is offset by a small percentage from its initial value, so as to increase the transfer function gain. By inducing an intentional error, the offset increases the sensitivity of the measurement.
The feedback photocurrent signal is then sampled again and stored as a “delta” signal. The previously derived “nominal” signal is subtracted from this “delta” signal in order to provide a derivative or Mach-Zehnder transfer function slope. If the difference value is less than or equal to a prescribed value (e.g., zero), the ‘bias’ is incremented. If the difference value is greater than the prescribed value (zero), the ‘bias’ parameter is decremented. The values by which the bias parameter is selectively modified can be fixed values, or a function of the magnitude of the “nominal” value.
Once the bias subroutine is completed, the gain value is restored to its initial setting and the monitored photodiode output signal is again sampled and stored as a “nominal” signal value. The modulator “bias” value is increased by a small percentage to increase the transfer function gain. Similar to the bias control loop, the purpose of the offset in the gain control subroutine is to increase the transfer function gain at the settling point. The photodiode output current signal is sampled and stored as a “delta” signal, and the stored “nominal” signal is then subtracted from the delta signal to derive a difference value.
Like the bias subroutine, the gain control loop uses the shape of the transfer function to settle to the location where the derivative is zero, and the slope of an induced error signal has the correct sign. If the difference value is less than or equal to the prescribed value (zero), the ‘gain’ parameter is incremented. If the difference value is greater than the prescribed value, the ‘gain’ parameter is decremented. Once the gain control loop is completed, the bias value is restored to its initial setting and the modulator control routine is then repeated indefinitely.
The offsets and measurements performed in the bias and gain control subroutines need not be repeated within the same period of time. Preferably, the repetition intervals are staggered relative to one another, in order to prevent the generation of a distinct modulation tone, with the energy being distributed over a larger range of frequencies, so as to minimize sensitivity to data harmonics of repetitive patterns and system noise. In order to reduce sensitivity to harmonic interferers in-band of the control loop bandwidth, and minimize impact to the modulated data stream for improved performance, the repetition intervals may be staggered in a pseudo random manner as employed in spread spectrum techniques. Once a control loop has settled, its associated subroutine may be interrupted, and placed in a power down mode; the subroutine may be periodically rerun, to compensate for aging and temperature variations,

Affiliated with

Also associated with

Rating

Mach-Zehnder modulator bias and driver gain control mechanism does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Mach-Zehnder modulator bias and driver gain control mechanism, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Mach-Zehnder modulator bias and driver gain control mechanism will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3257493