Something about LED lighting detection accuracy and consistency discussion

Illumination optical detection especially strategic emerging industries LED lighting is an optical detection involves a wide range, very practical, many aspects of teaching in schools that are not involved in the integrated technology. But also in the standards and testing methods, some such as an integrating sphere relative measurement of the LED lights and other standard reference optical detection problems have not been solved in the international arena. Therefore, to make LED highbay light optical detection accuracy and consistency also is indeed a problem. Many of our ongoing ability to verify the alignment is also limited attention and analysis submitted by the testing organization level than the results, there is no real mechanism for the detection of people, machines, materials, method, five aspects of the process of on-site inspection and monitoring. Therefore, some tests than on the activities of institutions to submit the results with greater than chance, and doubt the authenticity of the individual there. If the testing agency did not truly human, machine, material, method, the unity of these five areas as possible, it is impossible to truly accuracy and consistency and fundamentally improve their capabilities. This is a process requires effort, in order to minimize a variety of measurement error, inspection agencies should first do as much as possible from the five aspects of testing organizations own thing. This paper discusses the following, for reference.

A man

People are doing anything in the most critical factor. Trained team of excellent primary detection is very important. To make optical detection especially LED optical testing, test personnel should possess the necessary optical professional basis and cross-disciplinary knowledge, familiar with the standard, must consciously follow a strict testing system and uniform operating procedures, and have some practical experience .

Optical Inspection leader and even more so, you should be good at discovering even have reflexively ability to identify problems. Able to practice with excellent professional experience and profound theoretical basis of observation, analysis and judgment as well as to solve problems. Traditional lighting and LED illumination optical detection of the desired optical basic knowledge of Applied Optics, optics, photometry, colorimetry, and a variety of traditional lighting and LED lighting standards in optical definitions, terms and lighting measurement standards, as well as instrument The principle advantages and disadvantages of equipment and so on.

In practice, the optical measuring instruments and equipment such as an integrating sphere, the luminosity distribution meter, light meter, light meter, color measurement equipment, etc. should have practical experience in the operation of skilled can have a variety of lighting product development, the ability to better production practices (including traditional and LED, do have more to a differential).

Second, the machine

Illumination optical measuring equipment integrating sphere test system, light color distribution test system, according to the meter, luminance meter, optical radiation power meter, optical measurement systems biosafety. Testing organizations and businesses are commonly used or integrating sphere is the most used test system to measure the electrical parameters of LED highbay light color. It has a low acquisition cost, simple operation, test speed, etc., but there are many factors that may limit its measurement accuracy and consistency. Here, regardless of their effect size, the first major analysis of the test system integrating sphere potential factors that may affect.

(A) integrating sphere measurements of the basic working principle

Integrating sphere measurement method is based on the first use of the integrating sphere calibration standard reference light, and then measure the test light, and then processed by a computer to compare the relative measurement. When measured with a standard reference lamp lights are similar in all respects, the error will be reduced to a minimum. But even so, the actual device and the testing process to produce a number of factors will deviate, to mention most of the measured light and the difference between the standard reference light greatly. Although some factors less deviation from the ideal situation, a single factor has little effect on the measurement results. But too many factors deviate from the ideal, the possibility of cross-correlation effects will be greater, the error will be greater. So, should minimize equipment and testing process in a variety of deviations.

1 integrating sphere spectrometer measurement method works

Integrating sphere measurements have Photometry and spectroscopy measurements of the points. Early simply using a galvanometer pointer photoelectric photometer head of the table to read the photocurrent, then manual calculation method, the error is large, already no. After improving for the use of digital meter readings, but still outdated for a long time. Currently, much of the integrating sphere using spectral measurement. It can be divided into a photomultiplier tube and CCD array two. ‖

(A) photomultiplier tube inner workings of the spectrometer

Polychromatic light emitted measured in an integrating sphere uniform light mixing head after the optical input received by the optical fiber transmission into a spectrometer, and then the filter into the input slit, projected onto the grating on the spectral optical power signal decomposition.

Because as a photoelectric conversion of a photomultiplier tube itself can not distinguish spectrum, by a mechanical means so that rotation of the grating to the power of a monochromatic light signal bandwidth in accordance with the output wavelength of the order onto the slit, the slit close to a photomultiplier tube to receive and converting the optical power signal and the multi-stage amplification of electrical signals, and then further amplified by external circuitry for processing the output to the computer. In this series of process, technology is very complicated.

In addition, the measurement requires accurate wavelength scan (this continuous spectrum of light for measurement accuracy is not a problem, but the spectrum of the light separation optical parameters in particular, the color parameter measurement accuracy is important, such as incandescent color temperature measurement accuracy tricolor energy-saving lamps is high and the color temperature measurement accuracy is poor), the measurement resolution and accuracy but also with input and output slit width, wavelength positioning and scanning step great relationship. Therefore, the photomultiplier spectrometer measurements slower early spectrometer normally takes 2 to 3 minutes, the recent need about 10 seconds.



Photomultiplier spectrometer measurement accuracy can be high, but there are disadvantages photomultiplier: ① or because of the sensitivity due to light irradiation and reduce the irradiation time is too long, stopping irradiation partially restored, a phenomenon called "fatigue . " We used for many years for a photomultiplier tube spectrometer measurements of the same incandescent lamp 90 minutes was repeated trials, the unidirectional flux continuously decreased up to 3%, but little change in the color temperature within 5K (0.2%); ② light each point sensitivity uneven surface of the cathode; ③ applied in actual measurement voltage is too high will produce noise.

(2) CCD array spectrometer inner workings

Photomultiplier tube with a different place, the CCD array spectrometer measured by the grating of the polychromatic light is decomposed into order according to the size of the wavelength spectrum of the optical power signal, and can be distinguished at one time projected onto a CCD array spectral wavelength Such an image is received and obtained optical power of each wavelength spectrum mode signal in order to replace the need for scanning monochromatic light input to the photomultiplier tube in "time sharing" received signal power of each wavelength spectrum of light. Raster scanning is no longer needed and thus the mechanical rotating means. Therefore, the measurement speed is very fast, up to milliseconds. At present, the lighting of a better CCD converter 2048, the highest wavelength accuracy can be calculated as 0.2 nm. Overall, in all aspects, the current is less than the accuracy of the photomultiplier spectrometer.



CCD disadvantages are: ① substrate noisy; ② dark current and temperature are closely related, to be cooled, decreasing by 5 ~ 7 ℃, the undercurrent is halved, professional application of the CCD used liquid nitrogen cooling to a temperature below - 110 ℃; semiconductor cooling is generally -10 ℃ to -20 ℃, difficult to achieve a very high level; ③ CCD device quantum efficiency of each pixel is inconsistent, will result in the size of the optical power of each wavelength measurement error, which is more than the above-mentioned photomultiplier tube light sensitivity of the cathode surface uneven impact points more serious.



These two technically very complex inside the spectrometer, the links are likely to produce errors, the equipment cost different grades, the error level is different, and the purchase of equipment have been cured. Use of equipment testing organizations and businesses can not and will not allow casually move. In addition to the spectrometer, we should pay attention when in use to prevent or improve the measurement activities may be present in a number of other sources of error.

It is worth mentioning that: the earlier mechanical scanning the photomultiplier tube in the spectrum analyzer, when the calibration standard reference light, the photomultiplier should be adjusted to a negative pressure so that the maximum of the spectrum of optical power on the computer scan shows approximately 20% in height, then the measurement of the negative pressure adjustment knob would not have to move, or to be re-calibration.

Subsequent measurements simply adjust the magnification knob, because the standard reference light is generally inefficient incandescent glow, while the test lamp or LED highbay light efficiency may be several times higher than its ten times, not just flux close to the standard reference light; addition, many of the measured spectra and incandescent lamps vary greatly.

Therefore, the current amplification factor is often necessary to adjust the knob to prevent the measured spectral power signal to prevent overflow. In addition, each current amplification should compare each gear knob light in measuring the same test, the results are the same. When inconsistent, you need to use the standard reference light are detected in each gear about, select and measurement calibration values ​​closest stall.

(2) integrating sphere measurement error factors other

While integrating sphere measurement method is the use of standard reference lamps are compared with the relative measurement, but even with the standard reference test lamp lights identical or similar form, nor can therefore deviate from the ideal sphere that many factors some of it does not matter.

Some of the actual sphere single factor may deviate little effect, but deviate more factors, which affect the greater the likelihood of cross-correlation, the greater the likely overall error. So, should minimize the deviation of various aspects of the device.



‖

Ideal sphere radius around the inner spherical surface should be exactly equal throughout the high reflectivity spectrum was consistent and uniform neutral Lambertian diffuse reflection coating, the ball or other object has no internal standard reference light and the Measurement lights are just a luminous point of nothingness and so on. However, the actual integrating sphere is sure to have deformed; reflective coating unevenness, the reflectance is different (different coating materials), and a certain degree of non-spectral neutrality;, and built-in objects such as large lamp and bracket, block Guangping, the lamp and so on. In addition, although the testing organizations purchase specifications and diameter integrating sphere may be the same, but different brands of the internal structure of the integrating sphere may be quite different.

The form has been integrating sphere is fixed, there is light at the center of the sphere of 4Π law and the emergence of the rise of LED wall lighting of the ball 2Π method using 2Π measurement should compare and ensure horizontal glow with LED lights conventional down light 4Π method deviation between the measurement results within the allowable range. Recently also been a rotatable 2Π method integrating sphere, although it can be very easy to do down the light, but having regard to the exigencies of the standard reference lamp, still should compare it with conventional 4Π France between the two measurements Results deviation within the allowable range, no detailed description here.

(2) direct light shielding screen

Cover screen formula derivation principle is as follows:



However, the ideal integrating sphere should be no object. Therefore, blocking the screen should also be a test not only completely obscured lights direct light on the receiving probe, and its body and should be nothingness. Visible inside the sphere of physical shielding screen itself is a destructive ideal integrating sphere is the source of the error. Therefore, in addition to physical block screen wall with the ball should have the same reflective coating, but also should be measured not only completely obscured lights direct light on the receiving probe, and its shape and volume should be small and thin as possible, otherwise resulting error will increase.

In practice, the testing organizations and businesses purchase the same diameter integrating sphere block the screen in the shape (elongated, round, round and rectangular stack), the size, thickness and distance from the center varies. Therefore, although measurement of light before it is calibrated with a standard calibration reference, but the error of the size is still not the same.



Might be worse is that some testing organizations may not cover screen with a variety of specifications, or do not understand the principle of detection is not pressed in the shape and size of the measured object selection screen replacement cover and re-calibrated using standard reference lamp set standard. As well as in measuring double capped fluorescent lamps or LED lamps, you must use a long strip cover screen. But incandescent lamp with a standard reference test lamp shape that much difference, in principle, it is not suitable for long strip cover screen standard reference incandescent lamps, lamps for incandescent standard reference calibration accuracy greater impact. Therefore, the test results may produce small errors.

In addition, blocking the screen and the center of the sphere from the sphere of the radius R should be 1/3, the detection mechanism of the landing position of the integrating sphere block may vary. Despite being first with the standard reference lamp calibrated relative measurement, but deviate from the provisions will produce errors and consistency issues. Therefore, blocking the screen may be the larger error detection mechanism one of the main consistency. The following table results:



‖

(3) the standard reference light

Ideal standard reference shape of the lamp, the light-colored with a variety of parameters, and so should be possible to test the lights are the same, which requires numerous standard reference light, but in reality it is impossible to achieve. The current situation is that there are few standard incandescent or quartz halogen incandescent lamp as a standard reference.

Ideal standard reference quartz halogen incandescent or tungsten whole space should be uniformly distributed light, but in practice because they are blocking the light holder, and the incandescent filament is elongated, so that the whole space is not uniform light. But also the shape of the measured light, light color parameters vary. Some stray far from the standard reference lamps with similar demands for LED, shape, light uniformity, light angle and spectroscopy with quartz halogen incandescent or tungsten disparities. Therefore, the possibility of a relatively large error is inevitable.

(4) the standard reference light and the sample size of the lamp

Ideal ball inside the integrating sphere should not have any objects, including standard reference lamps and lights are just a test nothingness luminous point. Therefore, the actual measured standard reference lamps and lights are sources of error, the error must be minimized.

First, according to regulations, the maximum size of light integrating sphere radius R should not exceed 1/6 to 1/10. Visible, which for a long strip of tubular fluorescent and LED will cause large errors. Particularly the use of small power lamp measurement integrating sphere.

Secondly, the use of different diameter integrating sphere can also cause the same specifications measurement conditions inconsistent.

Third, the measure elongated tubular fluorescent lamps and LED highbay light,  you must use the long strip block Guangping, but with the standard reference incandescent calibration, the difference between the two types of shape of the lamp on the calibration block Guangping accuracy greater impact.

(5) the measured lamp warm-up stabilization time

Unlike a few incandescent, fluorescent twenty minutes, different LED lights settling time may take 30 minutes or one hour, whereas the standard requirements of light and color parameters are stable measurement results. So, when the testing organization for the same lamp warm-up time is not only the uniform regulations, it will produce accuracy and consistency errors.

(6) lamp pole and bracket

Ideal integrating sphere should be inside the ball without any objects, including lamp pole and bracket. Then, in the center of the sphere of 4Π measured lighting LED lamp holder and bracket will result in errors. 2Π measurement holder does not exist, but the ball and the hole wall of the hole size is fixed.

On the one hand destroyed the original spherical structure, while the measured light due to various different sizes, how to fill gaps and reflective coating material is a big problem; Second, the test lamp or lamps light the size of a standard reference that much difference; Third, because there is no light or lamp with the measured standard reference for other similar lamps or standard reference lamps, while the test lamp or lamps are tested cast light type of light distribution, and the whole space is basically a standard reference light emitting color Big difference distribution, which may cause large errors.

(7) the position of the probe, the fiber according to the pre-distortion and stability



① probe position, fiber twist: the spherical probe should be preceded by a thin flat matte film, and because the standard reference light and the measured light spectrum are often different, it should be free matte film selective absorption; the detection mechanism of the actual Some non-integrating sphere probe frosted films, some with Frosted Flakes, but whether there is selective absorption or absorption spectrum size is unknown; various testing organizations within the sphere of the actual location of the different probes, some flat, some Goals outer recess; but the test results show that the front position of the probe type, and non-destructive optical 10mm twisted little effect on the measurement results.

② Pre-Shot Stabilization: warm-up with electronic instruments need stability, like the photoelectric spectroscopy system should also be used with the measured light flux quite light for 20 minutes of Pre-shot stability, calibration and measurement was carried out, the actual This operation may be performed a lot of people did not pay attention.

(8) the position of the temperature measuring probe ‖

Temperature probe should avoid direct sunlight, close to the mask at the back of the block. And the distance from the measured light uniform, i.e., center of the sphere of radius R, the distance should be 1/3. For temperature-sensitive such as traditional fluorescent lamps, energy saving lamps and a variety of LED lights, ambient temperature measurement requirements of 25 ℃ (± 1 ℃).

In actual integrating sphere system, the detection mechanism sphere temperature probe position and distance may be different, even when measuring the same light only when the temperature of the probe light is not the same effect, making the need to adjust the ambient temperature, so measured light really felt ambient temperature may not be the same, lamp temperature will change. Therefore, there will be a light color difference between the measurement results, especially for high-power lamp or a small sphere case.

(9) temperature control system

Some temperature control thermostat can sphere, using the hair to adjust the ball and a constant temperature, which for incandescent lamps, high-intensity discharge lamp is not sensitive to the ambient temperature lamp has little effect on the measurement results. However, sensitive to environmental temperature of the LED light especially in the traditional double-ended tubular fluorescent lamps is not desirable, a greater impact, may have large errors. Not only to the ball hair hair between the double-shell to a temperature-controlled thermostat integrating sphere is a better way.



(10) reflective coating

Perfect reflection coating should have a high reflectivity Lambertian surface uniform and diffuse reflectance spectra were neutral.

① reflectance rate: There is barium sulfate reflective coating materials such as Si Beilun reflective coating, the reflection rate of 80%, 90%, 95%, and we hope to have a high reflectance coating. In addition, real integration sphere coated by light in particular ultraviolet radiation (e.g., long-term measurements of ultraviolet light leaking metal halide lamp and fluorescent lamps) will accelerate the aging of the coating produced; sphere lower reflective coating buildup of dirt, make up, lower reflectivity inconsistent, these differences will lead to quite different results.

② absorption spectrum selective coating: as compared with the high reflectivity in the visible spectral range were neutral uniform diffuse reflection may be more important. For the blue light reflective coating reflectance decrease, the overall reflectivity is higher, the formation of multiple reflections shade difference between the blue and the other will be. This standard reference light spectrum and the measured spectrum of the type of lighting, especially high color temperature fluorescent lamp of the three primary colors, high color temperature of the LED light will bring a large measurement errors, particularly for measurement of color parameters. But for low-pressure sodium lamp yellow singlet and blue incandescent few measurements have little effect.

(11) assist lamp system

Ideal integrating sphere should be inside the ball without any objects, including auxiliary lamp system block screen and lights. But in the actual integrating sphere have purchased, if the auxiliary lamp system already exists, then it is an additional source of error itself would, though it may be small errors caused. When measured with standard reference lamp lights look exactly the same or not very different, the absorption correction is not required, so when the lamp on the system is not supposed to exist. However, when measured with a standard reference lamp lights look big difference, so that the measured light absorption of light caused by measurement error is greater than the error caused by the system itself auxiliary lamp should be used when auxiliary lamp system absorption correction; For small sphere of its greater impact.

(12) the joint small sphere

Typically, the integrating sphere is divided into two semi-sphere opening and closing, in order to reduce collision impact, along the edge will have a semi-circular ball almost 1cm thick sponge cushion layer, which is not part of the sponge layer reflective coating (if sprayed after use is also vibration off), sponge layer reflectance spectral selectivity is tall and neutral, is not known. For small sphere of its greater impact. ‖

(13) lamp operating voltage sampling

In larger diameter integrating sphere system, electric control cabinet from the voltage meter to the holder or farther away from the terminals, power line voltage drop will be greater, especially when measuring high-power light is the need to consider this factor. Correct LED highbay light oltage sampling method is a four-wire connection method from the measured light ends of the lamp voltage sampling measurement. Otherwise, if the voltage electrical control cabinet meter readings will be measured lamp voltage and lamp power than the rated voltage and rated power is low. For example, a lamp voltage 220V power line voltage drop up some of 0.3V, about 0.14%, less than 0.2% like the allowable range, but this is caused by the power line, the voltage meter's accuracy when considering other factors, the voltage error may be exceeded; For low-voltage lights will become even more serious, such as some 24V 30W lamp up to 0.5 ~ 1V about 2% to 4%, will lead directly to large deviation optical parameters.

(14) built-in procedure

Most of the domestic sphere spectrum measurement system does not install the system in North America 8 diamond color tolerance system. Especially in North America system except step 7, the not given in the standard color tolerance 3,4,5,6 steps such as boundary conditions. Therefore, the need to use the words of the actual measurement, measurements must use manual calculation, accuracy and consistency vary, may be prone to problems.

(15) Power Supply

General test requirements of the standard mains frequency at 50Hz (± 0.5%) within; during stabilization requires preheating the lamp supply voltage ± 0.5% of the rated value, the measure requires the supply voltage rating of ± 0.2%, within Supply voltage total harmonic does not exceed 3% of the fundamental. Therefore, the power requirements are often ignored by some of the institutions and enterprises.

(16) voltage, current, power, accuracy of the meter

The voltage rating of the lamp to be used for measurement of the voltage meter, and for a current measurement of nominal light meter used, if they do not accurately measure the color of light leads directly to serious errors. Power meter is not accurate optical efficiency parameters will directly bring serious errors. Centralized installation in a control cabinet voltmeter, ammeter, power meter instrument should be removed and sent to a separate high-level metrology calibration. This is to improve the accuracy and consistency of the effective measures.

Third, the material

This is mainly used for detecting calibration traceable to standard reference lamps, integrating sphere spectral measurement system must also calibrated. Calibration traceable to improve the accuracy and consistency of the main requirements are:

1 to reduce calibration traceable links

Standard reference light money transfer will generate error and uncertainty, from the National Metrology Institute of measured value transfer to the provincial hospital, the provincial Institute of Metrology Metrology Institute and then passed to the city, and then passed to the inspection body; passing more times, the error and uncertainty is magnified more; therefore, as an important testing activities necessary to allow the participation of high-level inspection agency unified metrology institute to carry out money transfer calibration traceability.

(2) to establish various types of standard reference light

The principle of measurement is the relative standard reference lamps and lights are more similar when measured measurement error is smaller. In the international and domestic LED trichromatic fluorescent measurement accuracy is not enough there is the case of the long term research project should be carried out with the production of a number of commonly used test lamp shape, light and color parameters like standard reference lamp, the international and domestic authority After the calibration stoichiometry, to promote the use, both to ensure consistency, but also to ensure accuracy.

Fourth, France

It should be emphasized unity, only consistency:

1 standard understanding unity;

(2) develop a unified operating instructions, to avoid multiple testing agencies and the same detection mechanism when multiple inconsistencies between inspectors.

Five ring

Compared darkroom light distribution measurement method and the measured air temperature stability lights in both the upper and lower spaces giant slalom serious impact on the ambient temperature, the indoor air temperature stability of the measurement method for integrating sphere limited impact, but it should make the integrating sphere the temperature within the standards.

More info you can visit:www.lead-lighting.com