Fig. 1 Technician calibrates a test probe against a working-standard probe at the 514-nm argon-ion laser wavelength.
Laser-power-meter calibrations traceable to the National Institute of Standards and Technology (NIST; Boulder, CO) are becoming increasingly important (see "The value of traceability," p. 126). Laser manufacturers and end users need to quantify laser output by taking accurate and repeatable power or energy measurements to ensure that their equipment meets government and industry requirements. NIST can provide absolute calibrations at a limited number of wavelengths.
Today, as laser wavelengths proliferate, it is up to laser-power-measurement companies to develop techniques that create NIST-traceable standards for the wavelengths for which NIST does not provide calibration. Most important is the methodology that companies use to actually transfer the calibration from a NIST standard to the working standards for each wavelength. The working standards are then used on a daily basis in the actual calibration of unknown meters and detectors.
Measurement and metrology
Most power-meter manufacturers provide some type of calibration service to ensure that meters are maintained properly and that they perform at the highest possible level. In the process used at Molectron Detector Inc. (Portland, OR), the equipment is inspected and tested prior to calibration and any damage is repaired if necessary. During optical calibration, the product is closely tracked using a calibration database while being calibrated according to NIST-traceable, ISO-9002-certified procedures. All calibration information is recorded in a database. After a final inspection, the product is returned with a certificate verifying its calibration and NIST traceability.
The optical calibration system for the probes used to ensure NIST traceability stems from multiple NIST-calibrated "golden" or primary standard probes (see figure on p. 124). These golden-standard probes are very tightly controlled and monitored while in use. In addition, they are annually calibrated by NIST at the primary pulsed wavelengths of 193, 248, and 1064 nm, as well as at continuous-wave (CW) wavelengths of 514 and 1064 nm. These golden-standard probes are then used to produce multiple NIST-traceable "working-standard" probes for each of the calibration wavelengths offered. All other probes are calibrated against these NIST-traceable working-standard probes (see photo).
Calibration systems and procedures should, by design, provide the most accurate and repeatable calibration results possible. Calibrations can be automated using a graphical programming environment and thus benefit from computer-controlled data collection and calculation. For pulsed-laser calibrations, a ratiometric substitution method using a beamsplitter should be used to minimize the error caused by pulse-to-pulse fluctuations in the laser output. For CW laser calibrations, the inherent stability in the laser output permits the use of a straight single-channel substitution method.
Calibration of energy probes
Pulsed-laser systems can be used to calibrate energy probes and a limited number of power probes. A reference probe tracks laser stability and compensates for pulse-to-pulse variation in the laser source. This probe remains constant throughout the calibration process, and its output is measured each time the working-standard or test probes are measured. The working-standard probe is placed in the laser-beam path and its voltage output measured by a working-standard instrument, typically a dual-channel energy/power meter. A calibration program automatically gathers multiple data sets and calculates the average ratio of working-standard probe output to reference probe output. The working-standard probe is replaced by the probe to be calibrated. The same measurement repeats and the program again calculates the average test-to-reference-probe output ratio. The responsivity (Rv), in V/J, of the test probe is then calculated using
Calibration of power probes
Continuous-wave laser systems are used to calibrate power probes. The calibration process is very similar to that used for energy probes, except that no reference probe is used. First, a working-standard probe is placed in the laser-beam path and the output power of the laser is measured. Next, a test probe replaces the working-standard probe and its voltage output is measured. The responsivity, in V/W, is calculated as
Currently, NIST offers laser power and energy calibrations with uncertainties from 1% to 1.5%, depending on whether they are calibrated with its low- or high-power calorimeter standard. A full description of NIST capabilities can be found on its Web site at www.boulder.nist.gov/div815.htm.
The other half of the power and energy calibration is the calibration of the meter or display unit. Energy- and power-measurement instruments are calibrated using only NIST-traceable calibrated-measurement devices, and each instrument is calibrated over every usable range. All functions within the instrument are tested and verified. Strict quality-control inspections, both before and after calibration, ensure a finished product of the highest quality.
Continuous monitoring of all calibration systems, including both optical and electrical systems, helps to ensure accuracy and precision in the calibration processes and results. As with all measurement processes, calibrations have an associated uncertainty in the measurement result. Determining the uncertainty in product calibrations is done at our company by completing annual repeatability and reproducibility studies for all calibration systems. These studies repeat the calibration of a specified number of units multiple times to determine the repeatability of each calibration result if measured many times and the reproducibility of the measurement when completed by different technicians.
For each calibration system, two units of each type of probe or instrument are selected for the tests. These two samples are calibrated ten times each by two technicians, for a total of 20 calibration results for each unit. The standard deviations of a set of ten calibrations for each unit are combined in quadrature to determine the repeatability of the calibration. In addition, the variation in the average calibration result between each data set is calculated as the reproducibility of the calibration system.
For optical calibrations, these two contributions are added in quadrature to the uncertainty in the golden-standard probe calibration and the uncertainty in the working-standard probe calibration to determine the overall uncertainty in the calibration process. For electrical calibrations, these two terms are added in quadrature with the uncertainties in calibration of all equipment used to determine the overall electrical calibration uncertainty. The calibration uncertainty typically is ?2% for all instrument calibrations. For optical calibrations, the calculated uncertainties for each calibration system, listed by calibration wavelength, range from ?1% to ?3% (see table ).
Electrical substitution
The concept of electrical substitution involves embedding a heating element into the active area of the laser power sensor. When current is passed through the heater, an output signal can be measured from the probe. The relation between probe output and heater current is very repeatable. However, there are numerous factors that can cause the probe output to be different for the same number of watts of optical versus electrical input. These deviations contribute to what is referred to as the nonequivalence error.
Some examples are nonzero reflectivity of the optical absorber (less than 100% of the optical energy is converted to heat), heating of the leads connecting the heating element, and the fact that the electrically generated heat is typically produced in a location other than where optical absorption occurs. Electrical substitution was a major step forward in its time. It is, however, very difficult to obtain correct results with a disk-shaped absorber operating in air. Current work at NIST involves complex assemblies that are operated in high vacuum or cryogenic temperatures to adequately reduce nonequivalence errors. o
The International Vocabulary of Basic and General Terms in Metrology-available from the National Institute of Standards and Technology (NIST)-defines traceability as "the property of the result of a measurement or the value of a standard whereby it can be related to stated references, usually national or international standards, through an unbroken chain of comparisons all having stated uncertainties."
A statement from the NIST Web site (www.nist.gov) provides a clear outline of the purpose of NIST traceability: "Many government regulations and commercial contracts require regulated organizations or contractors to verify that the measurements they make are `traceable` and to support the claim of traceability by keeping records showing that their own measuring equipment has been calibrated by laboratories or testing facilities whose measurements are part of this `unbroken chain.` The purpose of requiring traceability is to ensure that measurements are accurate representations of the specific quantity subject to measurement, within the uncertainty of the measurement.
"The NIST Calibration Program often receives calls to verify the authenticity of a NIST Report of Test number appearing on another organization`s report. Although NIST can verify the authenticity of its report numbers, having an authenticity number does not provide complete assurance or evidence that the measurement value provided by another organization is traceable. Not only should there be an unbroken chain of comparisons, each provided measurement should be accompanied by a statement of uncertainty associated with the farthest link in the chain from NIST, that is, the last facility providing the measurement value. NIST does not have that information; only the facilities that provided the measurement values to the customer can provide the associated uncertainties and describe the traceability chain.
"To adequately establish an audit trail for traceability, a proper calibration result should include the assigned value, a stated uncertainty, identification of the standards used in the calibration, and the specification of any environmental conditions of the calibration where correction factors should be applied, if the standard or equipment were to be used under different environmental conditions."