A vast array of LED products is available for a wide range of applications. When selecting LEDs and related components for an art application, a solid understanding of several key LED characteristics will help you find an optimal solution. This article describes the LED and LED subassembly types that are most commonly used in light art and sculptural art that incorporates colorful, dynamic LED lighting.
This article assumes a very basic understanding of what an LED is and what it does. This Wikipedia article can help fill in any gaps.
In the broader context, LED applications can be roughly divided into four main categories. The list below shows some of the specific applications within each category:
- Indicators and Signs
- Control panel indicators
- Multi-segment digital displays
- Traffic lights
- Automobile brake/turn lights
- Large-format video displays and TVs
- Holiday lights
- Light art
- Residential interior/exterior lighting
- Street lights
- Automobile headlights
- Handheld devices (flashlight, smart phone)
- Architectural lighting
- Art lighting
- Data Communication
- Fiber optic and free-space data links
- Infrared (IR) remote control
- Measurement and Detection
- Optical computer mouse
- Barcode scanner
The first two categories are of primary interest here. A key distinction between the two is that the LEDs in light art generally emit light that is directly viewed by the audience, while the LEDs in art lighting are used to throw light on an artwork such that the light is reflected back to the viewer’s eyes. This distinction will be explored further in another article in this series, Part 3: Lighting Techniques.
This article focuses on LEDs that can produce a wide range of colors using the combined effects of individual red, green and blue LEDs that are incorporated into a single package. This is commonly referred to as an RGB LED. The close proximity of the devices within the same package is especially important: from typical viewing distances the three single-color LEDs appear to be a single point source of light, improving the perception of a single blended color when two or more LEDs are illuminated.
Typically, an RGB LED is driven such that the brightness of each color can be controlled independently. Combining different brightness levels can produce “million of colors”, with the exact number depending on factors that are described in Part 4: LED Control Methods. Unfortunately, due to the relatively narrow spectrum of light frequencies emitted by the individual red, green and blue LEDs, RGB LED devices are not able to produce a pure white color that is comparable to the broad-spectrum light produced by typical incandescent or LED “light bulbs”.
The quality of white light is important in art lighting applications where light is projected on an artwork to illuminate pigmented or dyed colors in the art. The low-quality white light emitted by an RGB LED has a low Color Rendering Index (CRI) resulting in washed-out or dull colors. Fortunately, manufacturers have responded by developing new devices that incorporate one or more “white” LEDs in addition to the RGB LEDs. These “white” LEDs are actually blue LEDs with an overlay of yellow phosphor. The exact type of yellow phosphor determines the color temperature or Correlated Color Temperature (CCT) of the resulting light. The terms “warm”, “cool” and “daylight” are often used to describe different color temperatures of white light.
LED devices that incorporate a fourth (white) LED are referred to generically as RGBW LEDs. In lower-cost devices, the white LED color temperature may not be specified, or may have a wide range as a result of loose manufacturing tolerances or other factors. In cases where the color temperature is more tightly controlled, so-called “4-in-1” LEDs may be referred to as RGB+WW (warm white) or RGB+CW (cool white). To support applications in which it is desirable to achieve intermediate color temperature, or perhaps to dynamically tune the white color temperature, some newer devices include both a warm and cool white LED. These LEDs are referred to as RGB+WW+CW or RGB+CCT.
While RGBW LEDs provide additional flexibility in terms of the lighting results that can be achieved, their usage may introduce some additional control challenges. These issues will be discussed further in Part 4: LED Control Methods.
Each type of RGB LED (including the color variants described above) is designed to produce a specific maximum light intensity level, where the light intensity is roughly proportional to the amount of electrical power (in Watts) that is supplied to the device. While some of the electrical energy is converted to visible light energy, most of the power is converted to heat. Unlike incandescent lights that release heat in the form of invisible infrared waves, LEDs do not produce any infrared waves and the heat is released into the LED package. This heat must be conducted through the package to a printed circuit board and other substrate materials (including heat sinks) that release heat into the air via convection.
Effective thermal management is an important design consideration as it can have a significant impact on both the light output and the lifetime of an LED. Specifically, LED luminous flux (in lumens) and lifetime (often specified as 50,000 hours) are reduced as the LED junction temperature rises above a nominal rated level. Unfortunately, LED manufacturers rarely provide enough information to fully comprehend or compensate for these effects. An interesting note here is that LED lifetime is specified as the number of hours after which the luminous flux drops to 70% (sometime 50%) of its original value.
One of the most common packages for RGB LEDs is often referred to as a 5050 SMD (surface mount device) package. This is actually a plastic leaded chip carrier (PLCC) with dimensions 5.0mm x 5.0mm x 1.5mm and a clear epoxy lens exposing the three LED die mounted inside. In the simplest case, the package contains three LEDs with the anode and cathode of each LED brought out to pins on the 6-leaded package. These devices are typically rated for a maximum continuous forward current If of 20mA per LED. The maximum per-LED power is the product of If and the corresponding forward voltage drop Vf. Since the voltage drop depends on the detailed chemistry and physics of each LED, the formula for total device power dissipation is P = If x (Vf.red + Vf.green + Vf.blue). Since Vf varies by up to +/-20% based on the actual forward current, a nominal Vf value is specified at the rated value of If. For low-power LEDs the nominal Vf.red = 2.0V and Vf.green = Vf.blue = 3.2V. So P = 0.02A x (2.0 + 3.2 + 3.2) = 0.17W. That’s actually a pretty small amount of power given how bright these LEDs are, so thermal management isn’t too much of a concern unless they are used in high-density configurations such as LED strips, possibly coated with weatherproofing material that impairs convection cooling. For maximum performance and longevity it’s advisable to mount LED strips to aluminums channels or other metal substrate that will help dissipate heat.
Much higher-power RGB LEDs are available, ranging from 1W to more than 10W. Due to the thermal management challenges the packaging is somewhat less standardized than low-power LEDs. Shown here is an example of a commonly used package for LEDs in this power range. Operating current can range from 100mA to nearly 1A per color. Due to the risk of rapid thermal failure at higher power levels, careful provisions must be made for both heat sinking and for precise control of the forward current (the latter is discussed in Part 4).
Because these high-power LEDs are blindingly bright (literally) they are rarely used for direct-view light art. But they are perfect for use in flood- and spot-lights for reflected-view art lighting applications (see Part 2: Optics for more information).
Another package type that is useful for art applications is the 5mm round domed-top package, a through-hole device rather than SMD. This device has essentially the same electrical specifications as the 5050 RGB LED described above, with the key difference that this package has 4 terminals instead of 6. In the example shown here, the cathode (negative side) of each LED is internally tied together and brought out to a single “common cathode” terminal. Common anode devices are also available. The advantage of this package type is that it can be easily mounted in 5mm holes in a sheet of acrylic or metal.
While it is certainly possible to create amazing light art using only the foundational devices described in the sections above, LED manufacturers have made our lives easier by also offering LED devices and subassemblies that integrate additional capabilities. The most prevalent example is the WS2812b LED, which is basically a 5050 RGB LED with an “intelligent” control chip also incorporated into the package. This control chip adds the following capabilities:
- Constant-current drivers to precisely control the individual LED forward current levels (and therefore brightness) using a pulse-width modulation (PWM) technique.
- A voltage regulator that allows a commonly-available 5V power supply to be used
- A serial protocol decoder that allows an external controller to use a single serial data signal to set the brightness levels of the individual RGB LEDs with 8-bit resolution per LED. The multi-drop serial protocol allows an essentially unlimited number of WS2812b devices to be daisy-chained and individually controlled (addressed) with the same serial data signal.
- Serial data signal shaping circuits that prevent signal degradation through a long daisy chain of devices.
There are now many variants of this “smart LED” approach that offer one or more of the following:
- Use of a 12V power supply to reduce the total current that must be supplied to the device (maintaining the same total power), alleviating issues caused by voltage drop in long LED strips.
- Inclusion of a white LED to create addressable RGBW LEDs
- Alternate serial data protocols, some of which use 2 signals (clock, data)
- Redundant data signal paths that improve the reliability of LED strips in the event of a failure of a single daisy-chained LED
A good example of this type of improved smart LED is the WS2815.
The most common type of subassembly incorporating smart LEDs like the WS2812b and WS2815 is the ubiquitous addressable LED strip. Although a full survey of the bewildering selection of available LED strips is beyond the scope of this article, the list below describes some of the key LED strip characteristics to consider.
- Type of LED chip used. This will define most of the key characteristics of the strip, including:
- Color capability: RGB, RGBW, or (someday) RGB+CCT
- Power supply voltage: 5V, 12V
- Serial protocol and associated timing
- Dimming PWM frequency: an important consideration for video compatibility
- Power consumption: also depends on the colors/brightness levels used
- LED mounting density/spacing: usually 30, 60, 90, or 144 LEDs per meter
- Strip width: usually between 5mm and 12mm
- Strip substrate color: white or black
- PCB trace thickness/width: affects the voltage drop that occurs under high-current conditions
- Weatherproofing: additional coatings and enclosures that minimize dust and water intrusion
Another type of integration addresses thermal management for high-power LEDs. Many manufacturers offer these devices mounted to a small printed circuit board (PCB) that sometimes incorporates an aluminum substrate. This simplifies mounting, interconnects, and provides additional surface area for heat dissipation. In many applications, however, the metal substrate alone will not provide adequate cooling, and must be attached to a larger heat sink using thermal paste.
An interesting variation on the addressable LED strip concept is the “smart LED string” consisting of a number of individually-addressable LED modules connected by short lengths of (usually) 3-conductor cable. Each module includes a small number of RGB LEDs (sometimes only one) that are driven by a shared control/driver chip, often the venerable WS2811. These LED strings address some of the limitations of LED strips and open up new design possibilities. They are available for a variety of LED types, including high-power LEDs and the 5mm round package described above. The flexible wire interconnects enable non-linear mounting (e.g. curve following) without having to cut/solder sections of strip. However, one common issue is that the fixed interconnect cable lengths may be too short, requiring cable extensions to be inserted.
Since not all applications require the ability to individually control each LED on a strip, manufacturers also offer a wide range of non-addressable strips, sometimes referred to as “analog” LED strips. These strips consist of 5050 SMD LEDs (any the color types described previously) and associated current-limiting resistors that enable the strips to be driven in a common-anode configuration using a standard supply voltage of 5V, 12V or 24V. A simple circuit can be used to provide on/off or PWM dimming control for all the LEDs of each color on the strip. This type of strip is perfect (and less expensive) for diffused- or reflected-view application where a section of an art piece is to be lit with single (but still dynamically changeable) color. They are currently the only strips that are offered using RGB+CCT LEDs.
This article (Parts 1 of a series) has presented an overview of the characteristics of the LED types most commonly used in light art and art lighting applications. Part 2 expands on this material with a discussion of the optical characteristics of these device types, as well as an overview of the additional optical elements often used in conjunction with them: diffusers, reflectors, and lenses.