The choice of LED lamps. 3. Spectra of light sources
Ideally, a spectrophotometer is required to assess the quality of the lamp's emission spectrum. As a last resort, you can use spectrophotometers for profiling / calibrating monitors (for example, ColorMunki) - if you have such a device. There is no point in buying spectrophotometers home for evaluating lamps, they cost from hundreds to tens of thousands of dollars.
However, for the needs of geologists and jewelers, they produce simple diffraction grating spectroscopes. Their cost is from 1200 to 2500 rubles. And this is a fun and useful thing.
The spectroscope looks like this:
')
You should look into the eyepiece (on the left, where the cone is), and the lens (on the right) should be directed towards the radiation source.
The diffraction grating decomposes the light into a spectrum (like a rainbow or an optical prism).
Before delving into the spectra of real lamps, recall the general information. (This is discussed in sufficient detail in the book in the “Quality of Light” chapter).
Here I will show two ranges of NLLs with an exceptionally high color rendering index of 97 (source here ):
Cold light:
You can see that the color temperature is 5401 K, index 97. The main thing is that you can see from what colors seen by the eyes the spectrum.
Warm light:
Temperature 3046 K, the index is also 97.
The spectrophotometer - in contrast to the spectroscope - shows not just which colors form the spectrum, but also gives their intensity. It is clearly seen that the spectra of both lamps have all the colors that make up white (“every hunter wants to know where the pheasant sits,” that is, red, orange, yellow, green, blue, blue, violet). The difference in color temperature is due to the relative contribution of cold (blue-blue) and warm (yellow-red) components.
I am forced to mention that this spectroscope is intended for mobile use with the help of eyes. It is extremely inconvenient to fix the picture, because the eyepiece is small and there are no devices for fixing on the camera. Therefore, you need to hold the camera, another spectroscope with one hand, and control the shooting with your voice. At the same time, you still need to keep the direction to the light source, small deviations from the normal lead to a distortion of the colors of the spectrum. Of the nearly a dozen different cameras that I have at home, the Samsung tablet turned out to be the best. The camera is only 5 megapixels there, but good software, and the size and position of the lens on the body of the device make it more or less convenient to attach a spectroscope. The white balance was fixed as “daytime”, ISO 400. The images were not processed, only leveled and cropped. The numbers on the right indicate the source color rendering index (100 - daylight in cloudy weather, 99 - incandescent lamp). The quality of the photos does not suit me very much - but I could not do better.
So, let's start from the top down and try to understand with concrete examples what to look for in such spectra.
• Daylight and incandescent lamp: the perfect spectrum, which presents all of the above colors.
• NLLs with color rendering indices 87 (reviewed here ) and 84 (discussed in the material chosen by the manufacturer) also show an almost complete spectrum. The red part usually becomes a problem - if yellow and orange are usually enough, then deep red hues are most often absent. You can not see them here. You can also assume (for example, by the amount of blue in the spectra) that manufacturers use different LEDs 5736SMD. Those. we are not dealing with the same lamp purchased from different vendors - but with different manufacturers.
• SDL with index 78 (its analysis is given in the chapter “Example of evaluation testing” in the book) along with the trimmed red part shows a small amount of blue. (It may seem that in comparison with the spectrum of the lamp with the index 84 it is not so. But here you need to remember that 84 is a warm lamp, T = 2900. And 78 is cold, T = 5750 K, there is much more blue by definition) . This is where the main drawbacks of simple low cost SDLs are, which supposedly form white light from the blue or purple light emitting diodes and the yellow-orange light of the phosphor. To the right of blue lies blue - but from the combination described it “does not work”. Therefore, in the spectrum of the NLL there is usually a failure. Due to this (plus the shortage of deep red) and the color rendering index falls.
• The lowest spectrum is a high-quality compact fluorescent lamp (CFL, T = 2700 K, 12000 hours life, declared color rendering index of at least 80). And here it is clearly seen, due to which this formally rather high value is achieved. The manufacturer himself calls this the “Tricolor system”. Those. it uses a phosphor of 3 components, each of which emits light in the form of a narrow band. (Of course, it is not at all easy to make such a lamp, since careful selection of a combination of phosphors is required.) It is the presence of such vertical bands (for example, purple, green, yellow) - a sign of low-quality light sources. The second consequence of the source line spectrum is the physical absence of certain colors in principle (in the figure, for example, there is almost no yellow and very little blue). It is obvious that the light of such lamps for the eyes is of little use despite the formally high enough indicators. Such lamps should be used in luminaires with high-quality diffusers (although, of course, this will not change the spectrum of the lamp).
Conclusion: in the spectra of light sources with a high color rendering index, all the colors of the spectrum should be present and there should be no intense narrow bands.
Separately, I want to warn against haste in the analysis of spectra. By the nature of my activities, I talked a lot with spectroscopists and noticed an iron pattern: the more qualified and professional the expert is, the more he is cautious and evasive in his conclusions. From the best of them, the professor, the head of the laboratory of spectroscopy in general, in principle, it was impossible to achieve a distinct conclusion (which at first I was wildly annoyed by my youth). Eye, of course, the best optical device of the existing. But the analysis and interpretation of spectra is an infinitely complex topic. There are a lot of different factors. Therefore, I strongly recommend only the simplest qualitative assessment of the spectra with the eyes, without attempts at clever conjectures and far-reaching conclusions. It is best to alternately look at the spectrum of the estimated lamp and the ideal spectrum of daylight or LN. Those. visual comparison between themselves.
However, for the needs of geologists and jewelers, they produce simple diffraction grating spectroscopes. Their cost is from 1200 to 2500 rubles. And this is a fun and useful thing.
The spectroscope looks like this:
')
You should look into the eyepiece (on the left, where the cone is), and the lens (on the right) should be directed towards the radiation source.
The diffraction grating decomposes the light into a spectrum (like a rainbow or an optical prism).
Before delving into the spectra of real lamps, recall the general information. (This is discussed in sufficient detail in the book in the “Quality of Light” chapter).
Here I will show two ranges of NLLs with an exceptionally high color rendering index of 97 (source here ):
Cold light:
You can see that the color temperature is 5401 K, index 97. The main thing is that you can see from what colors seen by the eyes the spectrum.
Warm light:
Temperature 3046 K, the index is also 97.
The spectrophotometer - in contrast to the spectroscope - shows not just which colors form the spectrum, but also gives their intensity. It is clearly seen that the spectra of both lamps have all the colors that make up white (“every hunter wants to know where the pheasant sits,” that is, red, orange, yellow, green, blue, blue, violet). The difference in color temperature is due to the relative contribution of cold (blue-blue) and warm (yellow-red) components.
I am forced to mention that this spectroscope is intended for mobile use with the help of eyes. It is extremely inconvenient to fix the picture, because the eyepiece is small and there are no devices for fixing on the camera. Therefore, you need to hold the camera, another spectroscope with one hand, and control the shooting with your voice. At the same time, you still need to keep the direction to the light source, small deviations from the normal lead to a distortion of the colors of the spectrum. Of the nearly a dozen different cameras that I have at home, the Samsung tablet turned out to be the best. The camera is only 5 megapixels there, but good software, and the size and position of the lens on the body of the device make it more or less convenient to attach a spectroscope. The white balance was fixed as “daytime”, ISO 400. The images were not processed, only leveled and cropped. The numbers on the right indicate the source color rendering index (100 - daylight in cloudy weather, 99 - incandescent lamp). The quality of the photos does not suit me very much - but I could not do better.
So, let's start from the top down and try to understand with concrete examples what to look for in such spectra.
• Daylight and incandescent lamp: the perfect spectrum, which presents all of the above colors.
• NLLs with color rendering indices 87 (reviewed here ) and 84 (discussed in the material chosen by the manufacturer) also show an almost complete spectrum. The red part usually becomes a problem - if yellow and orange are usually enough, then deep red hues are most often absent. You can not see them here. You can also assume (for example, by the amount of blue in the spectra) that manufacturers use different LEDs 5736SMD. Those. we are not dealing with the same lamp purchased from different vendors - but with different manufacturers.
• SDL with index 78 (its analysis is given in the chapter “Example of evaluation testing” in the book) along with the trimmed red part shows a small amount of blue. (It may seem that in comparison with the spectrum of the lamp with the index 84 it is not so. But here you need to remember that 84 is a warm lamp, T = 2900. And 78 is cold, T = 5750 K, there is much more blue by definition) . This is where the main drawbacks of simple low cost SDLs are, which supposedly form white light from the blue or purple light emitting diodes and the yellow-orange light of the phosphor. To the right of blue lies blue - but from the combination described it “does not work”. Therefore, in the spectrum of the NLL there is usually a failure. Due to this (plus the shortage of deep red) and the color rendering index falls.
• The lowest spectrum is a high-quality compact fluorescent lamp (CFL, T = 2700 K, 12000 hours life, declared color rendering index of at least 80). And here it is clearly seen, due to which this formally rather high value is achieved. The manufacturer himself calls this the “Tricolor system”. Those. it uses a phosphor of 3 components, each of which emits light in the form of a narrow band. (Of course, it is not at all easy to make such a lamp, since careful selection of a combination of phosphors is required.) It is the presence of such vertical bands (for example, purple, green, yellow) - a sign of low-quality light sources. The second consequence of the source line spectrum is the physical absence of certain colors in principle (in the figure, for example, there is almost no yellow and very little blue). It is obvious that the light of such lamps for the eyes is of little use despite the formally high enough indicators. Such lamps should be used in luminaires with high-quality diffusers (although, of course, this will not change the spectrum of the lamp).
Conclusion: in the spectra of light sources with a high color rendering index, all the colors of the spectrum should be present and there should be no intense narrow bands.
Separately, I want to warn against haste in the analysis of spectra. By the nature of my activities, I talked a lot with spectroscopists and noticed an iron pattern: the more qualified and professional the expert is, the more he is cautious and evasive in his conclusions. From the best of them, the professor, the head of the laboratory of spectroscopy in general, in principle, it was impossible to achieve a distinct conclusion (which at first I was wildly annoyed by my youth). Eye, of course, the best optical device of the existing. But the analysis and interpretation of spectra is an infinitely complex topic. There are a lot of different factors. Therefore, I strongly recommend only the simplest qualitative assessment of the spectra with the eyes, without attempts at clever conjectures and far-reaching conclusions. It is best to alternately look at the spectrum of the estimated lamp and the ideal spectrum of daylight or LN. Those. visual comparison between themselves.
Source: https://habr.com/ru/post/401447/
All Articles