Understanding Optical Power Measurements

As photonics technology becomes increasingly sophisticated, optical measurements grow increasingly complex. To make reliable measurements, one must consider the characteristics and interactions of light signals, as well as optical-to-electrical signal conversion, and the interpretation of electrical signals from the sensor. This article discusses light sources, detector types, calibration uncertainty, detector saturation and noise, attenuation, back reflection, interference, and beam divergence.

Light Sources

Optical measurements can be made with a wide variety of light sources. A light source can be continuous-wave (CW), modulated, pulsed, or even randomly fluctuating. A laser pulse can be as short as several femtoseconds (fs); can be modulated as high as hundreds of gigahertz (GHz); and can have kilowatts (kW) of average output or gigawatts (GW) of peak power. Most incoherent light sources, such as light bulbs or LEDs, have highly divergent beams. The wavelength can be anywhere between deep ultraviolet (UV) and far infrared (IR). Optical measurement techniques have been applied to non-optical, soft X-ray, and terahertz beams. Designers must consider various factors to ensure the measurement is accurate.

Absolute Versus Relative Measurements

Do you need to know the absolute optical power or energy? Or is a relative measurement (with respect to a reference, or simply a change) sufficient? Absolute measurements can be difficult or costly. For example, a Lambertian (highly diffuse) light source, or one with a large divergence angle, might not permit an absolute measurement, due to the difficulty of collecting all the light.

Optical Effects

In both experimental setups and commercial devices, optical elements such as lenses, filters, and beam splitters are used to process the optical signal and direct it to the detector. When selecting components for the test setup, the several factors need to be considered:

Optical attenuators use a variety of materials, such as colored glass, diffusers, or neutral density filters (absorptive and reflective). Each material has advantages and disadvantages that must be weighed. For example, absorptive neutral density filter may raise the temperature of the detector, causing errors in the measurement, when high power input light is incident on it. The high level of backreflection from a reflective neutral density filter can easily damage laser diode sources. Colored glasses or bandpass filters can be used only certain wavelength range needs to be transmitted.
If the optical train has two or more surfaces that reflect light, interference and/or back reflection can occur. Standing-wave interference between the incoming beam and the reflected beam can produce position-dependent measurement variations. Back reflection can damage certain sources (such as laser diodes).
There are no perfectly collimated light sources. All sources have some degree of divergence. And, few light sources provide perfectly even illumination. The “spatial distribution” of most varies.

Photon-To-Electron Conversion

An optical metering system consists of an optical detector, which converts an optical signal into an electrical signal, and a computing system that calculates the optical power or energy represented by the electrical signal. The measurements are displayed or stored in convenient formats, such as analog or digital output, entries in a data-collection file, or a graphical representation.

Newport’s 2936-R optical-power and energy meter (and its associated photodiode detectors) is such a system (Fig. 1). For details about the design of high-performance optical power meters, see “Optical Power Meters: Versatile and Economical."

1. Providing a digital display, an analog output, data-collection capability, and graphical presentations, the 2936-R optical power and energy meter comes with two 918D-series photodiode detectors.

Detector Types

The most common types of optical-signal detectors are photodiodes, thermopiles, and pyroelectric detectors. Photodiodes use the photon’s energy to create an electron-hole pair. The current created by the flow of these electrons is proportional to light intensity. Thermopile and pyroelectric detectors convert the photon’s energy into heat. The heat subsequently generates a voltage or a current.

Thermopile detectors are used for high-power laser sources, up to tens of kilowatts of optical power. Low-power laser diodes or LEDs are the most commonly measured light sources with photodiode detectors. A silicon photodiode detector is used for visible light, while a germanium or an indium-gallium-arsenide (InGaAs) detector is for infrared up to approximately 1.8 µm. With a properly calibrated optical attenuator, it can handle picowatts to several watts of optical power (Fig. 2). Pyroelectric sensors are popular for pulsed laser sources. Among these detector types, photodiode sensors are the most widely used.

2. A neutral-density filter widens the range over which power measurements can be made.

Photodiode Detector Designs

When a photodiode absorbs a photon, an electron-hole pair is created, generating a potential across the diode junction. Connecting the photodiode’s terminals to a load allows a current proportional to the light intensity to flow.

An op amp converts the photodiode’s current flow to a voltage. The magnitude of the voltage is the photodiode current multiplied by the value of the feedback resistor.

When low-noise continuous-wave (CW) measurements are required, the diode is operated in unbiased photovoltaic mode. The circuit in Figure 3 is a photodiode operating in photovoltaic mode.

3. When performing absolute power measurement, the photodiode detector operates in photovoltaic (unbiased) mode.

When fast response is needed to detect modulated light at narrow pulse widths or high repetition rates, reverse bias is applied to the photodiode to operate it in photoconductive mode. Changes in the number of electron-hole pairs change the photodiode’s conductivity. In this mode, the diode responds more rapidly to changes in light level.

NIST Traceability

To make an absolute measurement, a meter must be calibrated against an accepted reference. Detector manufacturers and calibration houses adhere to National Institute of Science and Technology (NIST) traceability standards and practices.

A gold-standard detector is calibrated by NIST at regular intervals, and all other detectors are calibrated against that standard. An extensive statistical analysis, gauge repeatability & reproducibility (GR&R), is required to determine calibration tolerances.

For absolute measurements, a detector is supplied with its own conversion table, which specifies the relationship between the optical input and the electrical output. This relationship is called the responsivity (R) and has units of amps per watt or volts per watt.

Many factors affect calibration tolerances: the calibration method; the calibration equipment; the facility at which calibration is performed; the skill of the people performing the calibration; and the processes that enforce repeatability. Each detector manufacturer and calibration house advertise different levels of calibration tolerance. It’s therefore critical to start with detectors that have the smallest claimed tolerances to achieve the most accurate results.

Even a carefully calibrated detector can be used incorrectly and produce invalid or misleading measurements. Any detector will saturate at some level of “too much” light. When there’s “too little” light, the signal might not be distinguishable from the noise. The power meter has its own uncertainties, which have to be considered.

Light Sources Revisited

Accurately measuring the power of a CW light source is challenging. Measuring a modulated or a pulsed light is even more difficult.

Due to the limitations of the detector’s response time, and the speed of the meter’s circuitry, it’s usually best to record the raw data, then process it with digital filtering or statistical averaging, rather than reading directly from the detector. This is particularly true when making peak-to-peak measurements, as waveforms can be significantly distorted.

An analog output port (such as those on most Newport optical power meters) allows a comparison of the waveform of the input signal with the waveform the detection system delivers.

Conclusion

Understanding the light source, the optical setup, the choice of detector materials, NIST traceability, and the capabilities of the detector manufacturers and calibration houses is important for getting accurate, repeatable measurements. As photonic components become more widely used in commercial and consumer devices, accurate yet economical methods of optical measurements will be increasingly important.

References

J. Jeong, “Optical Power Meters: Versatile and Economical,” Photonics Spectra.
Many textbooks discuss optical detector technologies. See, for example, C. L. Pollock, “Fundamentals of Optoelectronics,” Chapter 13.