APPLICATION NOTE

THE DIFFERENCES BETWEEN THRESHOLD

CURRENT CALCULATION METHODS

This Application Note explains the four threshold

DISCUSSION
calculation algorithms used by ILX Lightwave® and why
each method will result in a slightly different threshold
The four methods of threshold calculations each act
value.
upon a different characteristic of the L/I curve, so they
will be described individually with ideal examples; Figure
BACKGROUND
1 shows an ideal L/I curve which will be used through
the rest of this application note to illustrate the different
Four different algorithms are used to calculate the laser
threshold calculation methods. For these ideal examples,
threshold current:
arbitrary current numbers are assigned so the methods
1. Linear line fit
can be compared.
2. Two-segment line fit
3. First derivative of light vs. current The most critical portion of the graph is the threshold
4. Second derivative of light vs. current knee region. This is the point where the laser switches
from strictly spontaneous emissions to lasing emissions,
These four methods are recognized and described and the slope efficiently increases dramatically.
in the Telcordia Technologies standard “Introduction
to Reliability of Laser Diodes and Modules” (SR-
TSY-001369). Each method, even when used on the
same data set, will typically generate slightly different
values for the threshold current.
Light
Some laser diode manufactures have a preferred
calculation method, or their customers require a Threshold
“Knee”
particular method is used. In other applications, such as
R&D or university environments, no particular method is
Current
prescribed. In these cases, the user needs to understand
the differences between the methods, and why they give Figure 1 – Ideal L/I Curve with Derivative Curves
different results.
Linear Line-Fit Threshold Calculation
Understanding the factors that affect threshold
The linear fit method is the simplest, but potentially the
calculation allows the user to choose the right method for
most unreliable. This method simply extends a straight
particular applications, and leads to more effective and
line down the lasing portion of the L/I curve until it
efficient use of test time and resources.
intersects the horizontal axis. The intercept point is
defined as the threshold current.
Figure 2 shows a close up of the threshold knee and light output powers and the threshold value will be
the single line Linear Fit method of calculating threshold shifted.
current.

Light
Linear Line Fit #1
Light
Linear Line Fit #2
Linear Line Fit

Light
Light

10 11 12 13 14
Current
11 12 13
Current
Figure 3 – First-Point Selection, Effect on Calculated Ith Value.

Figure 2 – Linear Line-Fit

Two-Segment Line-Fit Threshold Calculation
There are several serious disadvantages to the linear Figure 4 shows a close-up of the threshold knee region
line-fit method: and the two-segment line fits. A line is fitted to the pre-
threshold portion of the curve; a second line is fitted
1. The calculated Ith value is highly dependent on the
to the lasing portion in the same manner as the linear
slope efficiency of the laser. Less efficient lasers will
fit method. The point where the two lines intersect is
have a lower calculated threshold value, especially
projected down onto the current axis and labeled at Ith
when the threshold knee is rounded instead of
value.
very sharp. The slope efficiency may shift due to
internal laser module properties or even because of
improper light coupling to the photodetector used to
Light
measure light output. Two-Segment Line Fits

alue.
2. If the linear line is based on a linear regression,
then the proper start and stop points of the
Light
regression must be selected. If the regression
uses data too near the threshold knee then the
calculated threshold value will shift to a lower value.
Conversely, if the regression starts too far from the
10 11 12 13 14
knee, the line fit may be strongly influenced by non- Current
linearities at higher powers.
Figure 4 – Two-Segment Line Fit
3. If the linear line is based on a two-point fit, then the
selection of the first point is critical. A point too low
on the threshold knee and the x-intercept is shifted
to a much lower current value. Conversely, if the
point is too high on the light curve then the linear
line may be impacted by non-linearities at higher
First Derivative (dL/dl) Threshold Calculation shows the first derivative curve from a 918 nm pump
laser. Notice that measurement noise at the dL/dl
Figure 5 shows an ideal example of a first derivative
maximum level, as well as the lack of a clearly defined
threshold calculation. The threshold current is defined as
maximum point.
the current at which the first derivative curve reaches ½
of the maximum value.
Light and First Derivative
80 0.90

0.80
70
Light 0.70
dL/dI 60
0.60
50

First Derivative
0.50

Light
40 0.40
Light
0.30
30
½ of max dL/dI
0.20
20
light 0.10
10 dL/dI
0.00

0 -0.10
10 11 12 13 14 0 50 100 150 200 250 300 350
Current
Current
Figure 7 – Real First Derivative Curve with Noise
Figure 5 – Ideal First Derivative Threshold Calculation

In this case, the threshold calculation repeatability was

poor because the noise value changed slightly from
The first derivative method is straight forward in most
sweep to sweep. When the dL/dl maximum point is
cases, but problems arise when the maximum of the
not repeatable then, of course, neither is the calculated
dL/dl curve is not easily defined. Figure 6 shows an
threshold current.
example of a dL/dl curve that continues to increase
after the threshold knee, and does not have an apparent
maximum.
Second Derivative (d2L/dl2) Threshold
Calculation
Light
dL/dI The second derivative threshold calculation method is
recommended by Telcordia Technologies in the GR-
3013-CORE Generic Requirements document. The
Light
method is illustrated in Figure 8.
½ of max
dL/dI ?
The second derivative method locates the point of
maximum rate of change of the L/I curve, which is also
the inflection point of the first derivative curve. It is not
10 11 12 13 14
Current necessarily the same threshold point that is calculated
using the first derivative method, however.
Figure 6 – Continuously Increasing dL/d

The second derivative method is insensitive to the

The problem of poorly defined first derivative maximum
non-linearities before and after the threshold knee since
is exacerbated by measurement noise, which is a real
those portions of the curve are not considered in the
problem encountered on any test system. Figure 7
calculation.
 Automated Threshold Calculation
Light
d2L/dI2 The second derivative calculation method does
not guarantee reliable and repeatable threshold
measurements, especially when automated test
Light equipment is used to run the test. A number of test
parameters need to be adjusted in order to reduce
measurement noise, optimize test resolution, reduce test
time, and balance other factors. Contact your local ILX
10 11 12 13 14 Lightwave® representative on how to best address test
Current
configuration optimization issues, including the selection
Figure 8 – Ideal Second Derivative Threshold Calculation of Sentinel Laser Reliability and Burn-In Test Systems.

In testing real laser diodes, there can be a dual peak

in the second derivative curve, with the second peak
Conclusion
caused by a kink in the threshold knee (Figure 9, same
The threshold calculation method you choose for your
laser as used to generate Figure 7). This possibility is
test application will depend on a number of factors:
acknowledged by the Telcordia document, and in most
cases, it is easy to determine which d2L/dl2 peak is real, •• Does your customer require that you use a particular
method?
even when using an automated test program to
calculate Ith. •• Are you calculating the threshold manually or with an
automated system?
Light and Second Derivative

80 0.10 •• How linear are the pre- and post-threshold portions

70 0.08
of the L/I curve?
60
0.06 •• How repeatable does the threshold calculation have
Second Derivative

50 to be?
0.04
Light

40
0.02 Although all four calculation methods are recognized
30

0.00 in the Telcordia document, the derivative methods

20

light
are the most reliable and least affected by anomalous
10 -0.02
d2L/dI2
laser characteristics. The second derivative method is
0 -0.04
0 50 100 150 200 250 300 350 preferred and is recommended in the Telcordia CORE
Current document.
Figure 9 – Second Derivative, Double Peak
The following publications are available for download at www.newport.com/ilxlightwave.

White Papers • Repeatability of Wavelength and Power Measurements Using the OMM-6810B
• A Standard for Measuring Transient Suppression of Laser Diode Drivers Optical Multimeter
• Calibration and Traceability Ensure Measurement Accuracy • Square Wave Modulation of the LDX-3500B
• Degree of Polarization vs. Poincaré Sphere Coverage • Stability of the OMM-6810B Optical Multimeter and OMH-6727B InGaAs Power/Wavehead
• Laser Diode Burn-In and Reliability Testing • Temperature Control Range of the LDM-4409
• Power Supplies: Performance Factors Characterize High Power Laser Diode Drivers • Temperature Measurement Using a Linearized Thermistor Network
• Simplifying Parametric Analysis of Laser Diodes • Temperature Stability Using the LDT-5948 / LDT-5980
• Reliability Counts for Laser Diodes • Thermal Resistance of the LDM-4409
• Reducing the Cost of Test in Laser Diode Manufacturing • Thermistor Constant Conversions: Beta to Steinhart-Hart
• Triboelectric Effects in High Precision Temperature Measurements
Technical Notes
• Tuning the LDP-3840B for Optimum Pulse Response
• Accuracy and Repeatability of Power Measurements Using the FPM-8220
• Typical Long-Term Temperature Stability of a LDT-5525 TEC
• Automatic Wavelength Compensation of Photodiode Power Measurements Using
the OMM-6810B Optical Multimeter • Typical Output Drift / Noise of an LDX-3412
• Bandwidth of OMM-6810B Optical Multimeter Analog Output • Typical Temperature Stability of the LDT-5500B
• Broadband Noise Measurements for Laser Diode Current Sources • Using Status Event Registers for Event Monitoring
• Callendar-Van Dusen Equation and RTD Temperature Sensors • Using the Dual Modulation Inputs of the LDX-3620B
• Clamping Limit of an LDX-3525B Precision Current Source • Using the LDM-4984 with the LDP-3840B
• Connecting Your Laser to the LDP-3830 • Using Three-Wire RTDs with the LDT-5900 Series Temperature Controllers
• Determining the Polarization Response of the FPM-8220 • Voltage Drop Across High Current Laser Interconnect Cable
• Effects of Cabling and Inductance When Pulsing High Power Laser Diodes • Voltage Drop Across High Current TEC Interconnect Cable
• Facility Power Requirements for the LDX-36000 • Voltage Limit Protection of an LDC-3916 Laser Diode Controller
• Four-Wire TEC Voltage Measurement with the LDT-5900 Series Application Notes
• Guide to Selecting a Bias-T Laser Diode Mount • App Note 1: Controlling Temperatures of Diode Lasers Thermoelectrically
• High Power Linearity of the OMM-6810B and OMH-6790B Detector Heads • App Note 2: Selecting and Using Thermistors for Temperature Control
• Large-Signal Frequency Response of the 3916338 Current Source Module • App Note 3: Protecting Your Laser Diode
• Laser Wavelength Measuring Using a Colored Glass Filter • App Note 4: Thermistor Calibration and the Steinhart-Hart Equation
• LDC-3736 Laser Protection • App Note 5: An Overview of Laser Diode Characteristics
• LDM-4982 and 4984 Quick Setup Guide • App Note 6: Choosing the Right Laser Diode Mount for Your Application
• LDP-3830 Independent Current Limit • App Note 8: Mode Hopping in Semiconductor Lasers
• LDP-3830 Laser Protection • App Note 11: Pulsing a Laser Diode
• LDP-3830 Pulse Performance • App Note 12: The Differences between Threshold Current Calculation Methods
• LDT-5900C Temperature Stability • App Note 13: Testing Bond Quality by Measuring Thermal Resistance of Laser Diodes
• LDT-5910C PID Control Quick Start • App Note 14: Optimizing TEC Drive Current
• LDT-5940C Voltage Measurement Techniques • App Note 16: Measuring Wide Linewidth Source with the OMH-6700B Series Waveheads
• LDX-3232 Modulation Bandwidth • App Note 17: AD590 and LM335 Sensor Calibration
• LDX-36000 CQW Pulse Characteristics • App Note 18: Basic Test Methods for Passive Fiber Optic Components
• Long-Term Output Stability of an LDX-3620B Laser Diode Current Source • App Note 20: PID Control Loops in Thermoelectric Temperature Controllers
• Long-Term Output Stability of an LDX-3525B Precision Current Source • App Note 21: High Performance Temperature Control in Laser Diode Test Applications
• LRS-9434 Temperature Set Point Accuracy • App Note 22: Modulating Laser Diodes
• LRS-9434 Temperature Coefficient • App Note 23: Laser Diode Reliability and Burn-In Testing
• LRS-9434 Threshold Current Measurement Repeatability • App Note 25: Novel Power Meter Design Minimizes Fiber Power Measurement Inaccuracies
• LRS-9434 and LMS-9406 Transient Protection • App Note 26: ReliaTest L/I Threshold Calculations
• LRS-9550 Device Temperature Algorithm • App Note 27: Intensity Noise Performance of Semiconductor Lasers
• LRS-9550 Fixture Temperature Range • App Note 28: Characterization of High Power Laser Diode Bars
• LRS-9550 Laser Drive Current Setpoint Accuracy • App Note 29: Accelerated Aging Test of 1310 nm Laser Diodes
• LRS-9550 Laser Eye Safety Features • App Note 30: Measuring High Power Laser Diode Junction Temperature and Package
• LRS-9550 Water Quality Guidelines Thermal Impedance
• Measurement of 4-Wire Voltage Sense on an LDC-3916 Controller • App Note 31: Mounting Considerations for High Power Laser Diodes
• Measuring the Power and Wavelength of Pulsed Sources Using the • App Note 32: Using a Power / Wavehead for Emitter Level Screening of High Power Laser
OMM-6810B Optical Multimeter Diode Bars
• Measuring the Wavelength of Noisy Sources Using the OMM-6810B • App Note 33: Estimating Laser Diode Lifetimes and Activation Energy
• Minimum Temperature Range of the LDM-4405 • App Note 34: Using USB Through Virtual COM Ports
• Minimum Temperature Control Range of the LDM-4982M / LDM-4894T • App Note 37: Measuring and Reducing Noise Using an LDX-3620B Ultra Low Noise Laser
• Nominal PID Constants for the LDT-5900 Series Controller Diode Current Source
• Output Current Accuracy of an LDX-3525B Precision Current Source • App Note 38: Achieving Millikelvin Temperature Stability
• Paralleling Laser Diodes
• Pulse Parameters and LDP-3830 Control Modes
• Quick Start: Modulation a Laser Diode Driver
www.newport.com

DS-082003_08/20 MKS products provided subject to the US Export Regulations. Diversion or transfer contrary to US law is prohibited. mksinst™
©2020 MKS Instruments, Inc. is a trademark of MKS Instruments, Inc., Andover, MA. Swagelok® and VCR® are registered trademarks of Swagelok Marketing
Specifications are subject to change without notice. Co., Solon, OH. Viton® is a registered trademark of E.I. Dupont, Wilmington, DE.