Datasheet

ADN2872
Rev. 0 | Page 11 of 20
Operation with Lasers with Temperature-Dependent
Nonlinearity of Laser LI Curve
The ADN2872 ERCL extracts information from the monitor
photodiode signal relating to the slope of the LI characteristics
at the Optical 1 level (P1). For lasers with good linearity over
temperature, the slope measured by the ADN2872 at the Optical 1
level is representative of the slope anywhere on the LI curve.
This slope information is used to set the required modulation
current to achieve the required optical extinction ratio.
0
0.5
3.0
2.5
2.0
1.5
1.0
4.0
3.5
20 40 60
CURRENT (mA)
OPTICAL POWER (mW)
10080
08013-008
RELATIVELY LINEAR LI CURVE AT 25°C
NONLINEAR LI CURVE AT 80°C
0
Figure 25. Measurement of a Laser LI Curve Showing
Laser Nonlinearity at High Temperatures
Some types of lasers have LI curves that become progressively
more nonlinear with increasing temperature (see Figure 25). At
temperatures where the LI curve shows significant nonlinearity,
the LI curve slope measured by the ADN2872 at the Optical 1
level is no longer representative of the overall LI curve. It is
evident that applying a modulation current based on this slope
information cannot maintain a constant extinction ratio over
temperature.
However, the ADN2872 can be configured to maintain near
constant optical bias and an extinction ratio with a laser
exhibiting a monotonic temperature-dependent nonlinearity.
To implement this correction, it is necessary to characterize a
small sample of lasers for their typical nonlinearity by measur-
ing them at two temperature points, typically 25°C and 85°C.
The measured nonlinearity is used to determine the amount of
feedback to apply.
Typically, the user must characterize five to 10 lasers of a particular
model to obtain a good number. The product can then be cali-
brated at 25°C only, avoiding the expense of temperature
calibration. Typically, the microcontroller is used to measure
the laser and apply the feedback. This scheme is particularly
suitable for circuits that already use a microcontroller for control
and digital diagnostic monitoring.
The ER correction scheme, while using the average nonlinearity
for the laser population, supplies a corrective measurement
based on the actual performance of each laser as measured during
operation. The ER correction scheme corrects for errors due to
laser nonlinearity while the dual loop continues to adjust for
changes in the Laser LI.
For more details on maintaining average optical power and
extinction ratio over temperature when working with lasers
displaying a temperature-dependent nonlinearity of LI curve,
contact sales at Analog Devices.
CONTROL
The ADN2872 has two methods for setting the average power
(P
AV
) and extinction ratio (ER). The average power and extinc-
tion ratio can be voltage set using the voltage DAC outputs of a
microcontroller to provide controlled reference voltages to
PAVREF and ERREF. Alternatively, the average power and
extinction ratio can be resistor set using potentiometers at the
PAVSET and ERSET pins, respectively.
VOLTAGE SETPOINT CALIBRATION
The ADN2872 allows an interface to a microcontroller for both
control and monitoring (see Figure 26). The average power at
the PAVSET pin and extinction ratio at the ERSET pin can be
set using the DAC of the microcontroller to provide controlled
reference voltages to PAVREF and ERREF. Note that during
power-up, there is an internal sequence that allows 25 ms before
enabling the alarms; therefore, the user must ensure that the
voltage for PAVREF and ERREF are active within 20 ms.
PAVREF = P
AV
× R
SP
× RPAV (V)
AV
CW
CWMPD
ERSET
P
ER
ER
P
I
RERREF 



1
1
_
(V)
where:
P
AV
(mW) is the average power required.
ER is the desired extinction ratio (ER = P1/P0).
R
SP
(A/W) is the monitor photodiode responsivity.
I
MPD_CW
(mA) is the MPD current at that specified P
CW
.
P
CW
(mW) is the dc optical power specified on the laser data sheet.
In voltage setpoint, RPAV and R
ERSET
must be 1 kΩ resistors with
a 1% tolerance and a temperature coefficient of 50 ppm/°C.