The choice of LED lamps. 2. About optimal power
The second part of the methodological material is devoted to the analysis of the optimal power of the LED lamps (SDL) of the “corn” type. Those. lamps which power is most profitable to buy.
Honestly, I first had some working hypothesis. However, the results did not confirm it.
I recall in the first part we chose a promising manufacturer. For this article I bought 4 identical lamps of different power from this seller (for attentive readers, I’ll specially note that the second cold light bulb on the left; but this is only because its warm copy illuminated this scene when shooting; it was the warm lamp that was used for real measurements) .
This is the top line of the manufacturer's SDL in aluminum cases with 5736SMD LEDs. All lamps have no ripples, color rendering index is ~ 84, operating voltage is 85-265V, T ~ 2900 (assessment method is described separately ).
')
The following table summarizes all the lamps required for analysis:
Analyzing data in tabular form is inconvenient, it is better to look at the graphs.
This is the dependence of the light output (lumens per 1 watt of power) for lamps of different values:
Red shows the promises of the manufacturer. Those. declared luminous flux related to the rated power. Everything is pretty decent, 90-100 lm / W. Black shows the actual data.
It can be seen that only a 7W lamp shows a good result.
And this is the dependence of price efficiency (lumens for 1 ruble) for lamps of different denominations:
Again, the dramatic difference between the statements of the manufacturer - and reality. And again, the most advantageous is the SDL at 7W.
Findings:
1. The higher the nominal power of the SDL, the lower its real efficiency. If we recall that powerful lamps heat up more, then their service life, respectively, will be lower.
2. Special mention deserves the lamp at 18 watts. If the rest of the lamps after heating the operating power is approximately equal to the stated nominal value - here it is much lower. Perhaps I got a bad copy. But it is more likely that the manufacture of powerful household SDLs in the “corn” construct is too complicated a task for producers at this stage.
3. The results obtained in these two articles should not be automatically transferred to other lamps. This shows a methodical approach. But the findings do not have the nature of laws for all occasions. Their findings need to be checked every time anew. For example, this manufacturer has another line on the 5736SMD, a little easier. So, even it is impossible to transfer the results obtained in this test. That is why I always emphasize that the emphasis should be on understanding the methods of testing and optimization, on the ability to perform such an assessment on my own.
Separately, it must be emphasized that the method of estimating the magnitude of the luminous flux used by me has a large error. And not only systematic, but also random. For example, a 18-watt lamp already stood for the diffuser of a desk lamp — that is, a part of the flow did not reflect downwards on the luxmeter sensor, but left “on the side”. Exact measurements are performed according to GOST in the integrating field.
4. It is more profitable to use lamps for the maximum number of lamps. Those. the total luminous flux to create a large (6-8) number of relatively low-power lamps instead of trying to put 2-3 supposedly powerful. For example, I have such a lamp in the kitchen:
8 lamps, of which 4 LN at 60 W (T = 2700) and 4 SDLs with a power of 7 W (T = 5600). The final color temperature is about 4100 K, and the illumination is 370 lux (the distance from the ceiling is about 160 cm).
PS This is exactly the kind of 9W lamp of warm light that was considered in the article “ Two approaches to the production of a maize-type NLL ” as a positive example.
Honestly, I first had some working hypothesis. However, the results did not confirm it.
I recall in the first part we chose a promising manufacturer. For this article I bought 4 identical lamps of different power from this seller (for attentive readers, I’ll specially note that the second cold light bulb on the left; but this is only because its warm copy illuminated this scene when shooting; it was the warm lamp that was used for real measurements) .
This is the top line of the manufacturer's SDL in aluminum cases with 5736SMD LEDs. All lamps have no ripples, color rendering index is ~ 84, operating voltage is 85-265V, T ~ 2900 (assessment method is described separately ).
')
The following table summarizes all the lamps required for analysis:
Analyzing data in tabular form is inconvenient, it is better to look at the graphs.
This is the dependence of the light output (lumens per 1 watt of power) for lamps of different values:
Red shows the promises of the manufacturer. Those. declared luminous flux related to the rated power. Everything is pretty decent, 90-100 lm / W. Black shows the actual data.
It can be seen that only a 7W lamp shows a good result.
And this is the dependence of price efficiency (lumens for 1 ruble) for lamps of different denominations:
Again, the dramatic difference between the statements of the manufacturer - and reality. And again, the most advantageous is the SDL at 7W.
Findings:
1. The higher the nominal power of the SDL, the lower its real efficiency. If we recall that powerful lamps heat up more, then their service life, respectively, will be lower.
2. Special mention deserves the lamp at 18 watts. If the rest of the lamps after heating the operating power is approximately equal to the stated nominal value - here it is much lower. Perhaps I got a bad copy. But it is more likely that the manufacture of powerful household SDLs in the “corn” construct is too complicated a task for producers at this stage.
3. The results obtained in these two articles should not be automatically transferred to other lamps. This shows a methodical approach. But the findings do not have the nature of laws for all occasions. Their findings need to be checked every time anew. For example, this manufacturer has another line on the 5736SMD, a little easier. So, even it is impossible to transfer the results obtained in this test. That is why I always emphasize that the emphasis should be on understanding the methods of testing and optimization, on the ability to perform such an assessment on my own.
Separately, it must be emphasized that the method of estimating the magnitude of the luminous flux used by me has a large error. And not only systematic, but also random. For example, a 18-watt lamp already stood for the diffuser of a desk lamp — that is, a part of the flow did not reflect downwards on the luxmeter sensor, but left “on the side”. Exact measurements are performed according to GOST in the integrating field.
4. It is more profitable to use lamps for the maximum number of lamps. Those. the total luminous flux to create a large (6-8) number of relatively low-power lamps instead of trying to put 2-3 supposedly powerful. For example, I have such a lamp in the kitchen:
8 lamps, of which 4 LN at 60 W (T = 2700) and 4 SDLs with a power of 7 W (T = 5600). The final color temperature is about 4100 K, and the illumination is 370 lux (the distance from the ceiling is about 160 cm).
PS This is exactly the kind of 9W lamp of warm light that was considered in the article “ Two approaches to the production of a maize-type NLL ” as a positive example.
Source: https://habr.com/ru/post/401213/
All Articles