The dimming profile describes how the light source behaves when dimming.
In other words:
What is the brightness of the illuminant related to the dimmer position?
The new index DPA in percent expresses the Dimming Profile Abberation in comparison to the incandescent light bulb.
At the bottom of this page, there is a list of illuminants tested for dimmability.
By dimming (classical leading edge or recent trailing edge), only a part of the sinusoidal mains voltage is supplied to the illuminant. In this case, each half-wave is connected to the illuminant at a particular point in time (leading edge) or truncated at a certain point in time (trailing edge).
In order to avoid large current pulses, the resulting greater load and premature aging of the illuminant with inductive loads (e.g. transformer) a leading edge dimming should be favoured. For dimming capacitive loads (e.g. LED illuminants) the trailing edge dimmer has advantages.
For pure resistive loads (e.g. the light bulb), the type of dimmer does not matter.
The diagram shows on the X axis the sinusoidal half wave of the mains voltage divided into 16 equidistant time sections for leading / trailing edge.
According to the partial voltage supplied, the incandescent lamp extracts electrical power, which is why the dimming is suitable for saving energy.
On the Y axis, this is shown normalized with reference to the nominal power of the illuminant.
It can be seen that the correlation between the energy supplied and the radiated light is not linear, the light output decreases with dimming. Incidentally in the same context, the light color temperature drops.
That is, when dimming, the light of the incandescent lamp becomes more reddish and dark as energy is saved because it converts the greater part of the electrical power into heat.
The course of the light efficiancy and thus the brightness of a light bulb during dimming is the same for all incandescent lamps. That's why we never had to ask ourselves what the course was; It has always been determined by the simple physics of the tungsten filament.
The correlation between the applied energy and the brightness of cold lamps (fluorescent tubes, energy saving lamps and LEDs) is determined by the electronic ballast. Since this can look totally different depending on the manufacturer, a cold lamp behaves correspondingly differently.
For this reason, there exists a distinction between dimmable and non-dimmable illuminants in contrast to the incandescent light bulb.
With the so-called dimmable LED lamps, however, the variance of the behavior during dimming as with light flickering has achieved such a degree that they largely do not correspond either to the behavior of a incandescent or to one another.
Or extremely: They call themselves dimmable, but sometimes they are not in practice, they may only survive operated on a dimmer.
In the meantime, it is generally known that a dimmer operated below its minimum load can cause problems, for example the light begins to flare like a candle light with low frequency and irregulary (not to be mixed up with light flicker at double mains frequency).
It is also known that the dimmability for other reasons fails with some dimmers and that there may be incompatibilities between dimmers and lamps.
This has also been stated by the NDR (a German radio broadcast service) in its program "Markt" series on 04-Apr-2016 (even if procedural errors were made, e.g. disregarding the dimmer minimum load).
However, this is not the subject of this page or of the dimming profile. The consideration of the minimum load and the basic compatibility of the dimmer with an LED illuminant is assumed for the following. Here with the dimmig profile it is about the illuminants unfortunate not ideal behaviour with an ideal dimmer, caused by the manufacturer's design of the electronic ballast.
By experience there are two main applications where the dimming behavior plays a role:
Many people replace a light bulb only with a defect against an LED illuminant. A mixing operation with LEDs is always possible without any technical problems.
However, if on dimmable chandelier you exchange the a candle incandescent bulb with an LED candle illuminant (same lumens and color temperature) the astonishment is perhaps big:
The LED illuminant dims differently, mostly it is significantly brighter than the light bulbs on the way to the dark.
In an existing apartment, all incandescents are exchanged for LED illuminants (no mixed operation):
Even if nothing flares or flickers, the LED bulbs can not be properly dimmed. Either the LED bulb is much too light (two to three times) when the dimmer is turned to the left. Or if at full brightness the dimmer is turned to the left to darken the light, nothing happens for long and suddenly within a small dimming range the light darkens much faster.
However, all this is not due to the dimmer but to the illuminant, which behaves completely differently under the same circumstances as the previously operated incandescent bulb.
In contrast to light flickering, which can be dangerous or unhealthy, the dimming behavior of a luminous means is a pure quality feature and when dimming the illuminant, a lower quality is immediately noticeable to everyone.
This raises two questions about the problem:
1. How bright is the LED bulb in certain dimmer settings compared to an incandescent lamp?
2. How do you express the partial large differences between leading edge and trailing egde?
The consumer would therefore be much clearer if the dimming behavior of an illuminant is measured, set as a measure compared to an incandescent lamp and specified in the technical data.
I've been thinking about this for a long time and found a solution.
The reference dimmer is a specially developed ideal dimmer:
It is processor-controlled, operates precisely according to the (dimming) educational book, and is independent of the load in its behavior.
This means that the measuring method is independent of dimmers on the market, which in turn behave differently.
The measuring method only reflects the dimming behavior of the illuminant.
The dimming behavior is measured as a dimming profile (relative brightness vs. dim position) in 16 dimpositions (evenly distributed over 360° of a circle, i.e. at distances of 22.5°). By means of the positions 4... 14 used for a real dimmer, the associated dimming profile abberation is determined using the abbreviation DPA as a percentage.
Strictly speaking, two percentages are given because, in contrast to the incandescent bulb, it can make a very large difference whether the light source is dimmed with a classical leading edge or a more trailing edge dimmer. This also clearly indicates at which type of dimming the LED illuminant runs better.
In the following diagrams, the light bulb is designated "Incandescent" as the reference of the dimming profile, for the device unter test it is "LE" for leading edge and "TE" for trailing edge.
Just as with light flickering: the smaller the deviation, the better.
The two figures as a pair of figures DPA/% = L##T## respectively indicate the percent deviations for leading edge and trailing edge.
With >99%, a lamp is not really dimmable in practice.
Values of <= 25% indicate that the light source can be adjusted properly.
At values of <= 10%, the illuminant is comparable with the behavior of incandescent lamps, that is, in mixed use, there will be hardly any difference noticable, with 0% there is no difference.
During the development of the DPA procedure I have noticed an additional horror:
It makes a difference with some lamps whether they are dimmed from light to dark or from dark to light (down-dimming or up-dimming):
For a particular dimming position (e.g. center), the brightness is different.
However, here the DPA specification no longer differentiates, all cases are measured, but only the more unfavorable is used as the final value.
Hereby I would like to clarify that the categorization and evaluation of the dimming properties of a product is not an overall assessment of the product but only the partial aspect of the dimming behavior. Nevertheless, I consider the DPA value to describe the dimming behavior for a useful quality criterion. Everyone makes their own decision, which criteria he or she uses to evaluate a product in its entirety.
The list on the test page for light flicker has already become very long.
From the presentation similar here the DPA list:
It should be noted that, particularly some filament lamps and all driverless lamps, which each contain only a small amount of electronics, perform relatively well in the dimming behavior, which is similar to the tungsten filament in the nature of the (missing) thing. However, these lamps have very high light flicker values, so a view at the light flicker test page is quite important.