Understanding how fluorescent and high-intensity discharge (HID) lamps work makes problem solving and corrective maintenance easier
To comply with new laws on reducing energy usage, fluorescent and HID lighting systems — both examples of arc discharge lamps — have become more complex and varied in their performance. To understand where we're going on the installation, maintenance, and troubleshooting fronts, it's important to understand where we've been.
Arc discharge lamps produce light by an electric arc struck between two main electrodes or cathodes. The arc is controlled by a ballast that allows the lamp to start reliably (provide the open circuit voltage), regulates the current (ampere flow) to the lamp, and maintains the proper electrical conditions during lamp operation. Although the basic concept behind these lamps hasn't changed, technology has evolved over time. Let's take a look at the latest updates in the arc discharge lamp arena to make troubleshooting easier.
A linear fluorescent lamp consists of a tube of glass with an electrode on each end. The sealed tube is filled with a very low pressure (atmosphere) of argon and a small amount of mercury, while a coating of phosphor lines the inside surface of the tube. The electrode, or filament, in a fluorescent lamp is similar to the filament of an incandescent lamp, except for a coating (called the emission mix) that increases the ability to emit electrons.
When proper voltage is applied between the electrodes, a plasma arc (the flow of electrical current) is created through the mercury vapor, thereby producing short-wave ultraviolet (UV) radiation. The UV radiation, or energy, absorbed by the phosphor coating is re-radiated as visible light energy.
Since their development in the 1940s, fluorescent lamps for area illumination in offices and industry generally have used three magnetic ballast constructions: The first is usually a 4-ft lamp; the other two are generally 8-ft lamps:
The rapid-start (RS) lamp preheats the lamp's filaments from a low-voltage transformer for about 1 sec before starting. After the lamp starts, the filaments are continuously heated with a reduced current during operation to provide optimal lamp life.
High-output (HO) lamps are rapid-start lamps designed to operate more efficiently at higher currents, thus producing approximately 40% more light than standard rapid-start-type ballasts.
Instant-start (Slimline) lamps do not have the cathodes preheated prior to starting; therefore, they require a starting voltage about 40% higher than if the cathodes were preheated.
While the 4-ft-long T12 RS 430mA, the HO 800mA, and the Slimline lamps have been the lighting industry's workhorses for years, Department of Energy (DOE) rulings are phasing out the magnetic ballasts for these types of lamps. Thus, only electronic ballasts will be available for these older lamp types. The industry is moving to electronic ballasts for the newer T8 lamp family, and the newest T5 family of lamps uses only electronic ballasts. Rather than delivering a 60-Hz current to a lamp, the electronic ballast provides an input current between 20,000 Hz and 40,000 Hz, resulting in greater phosphor excitation and thus a higher light output.
The DOE is also promoting the use of fluorescent dimming and bi-level load-shedding systems, which can reduce the light output by 30%. Fortunately, recently introduced fluorescent dimming ballasts (using 0V to 10V, 3-wire analog, 2-wire analog, and digital communication/addressable control schemes) are about as efficient as the standard ballasts.
The best way to begin a review of electronic ballasts is to study their starting methods.
An instant-start electronic ballast is similar to the older Slimline magnetic ballast because it creates a high voltage (typical 600V for F32T8 lamps) across the lamp electrodes to initiate lamp operation, without providing lamp electrode heating. This ballast uses 1.5W to 2W less energy per lamp than the rapid-start model. Typically providing a 10,000 to 15,000 switching cycle, an instant-start ballast generally has the lamps wired in parallel.
A rapid-start electronic ballast has a separate set of windings that provide a low voltage (approximately 3.5V) to the electrodes for 1 sec prior to lamp ignition. The starting voltage is between 450V to 550V for F32T8 lamps. Typically providing a 15,000 to 20,000 switching cycle, a rapid-start ballast generally has the lamps wired in series. Thus, if one lamp fails, the other lamps in the circuit will extinguish.
A programmed-start ballast also has a set of windings that provides precise heating of lamp filaments and controls the pre-heat time before the startup voltage, thereby reducing filament stress. The cathodes heat to about 700°C prior to ignition, allowing the programmed-start ballast to provide more than 50,000 starts. This extended lamp life is useful for applications where frequent switching shortens lamp life, such as spaces with occupancy sensors. Programmed-start ballasts are recommended for all types of automated switching strategies, including daylight compensation and demand response strategies. These ballasts generally provide a 20,000 to 40,000 switching cycle.
While the electronic ballast offers energy savings, controllability, and reduced size compared to the magnetic type, the layout and installation of linear fluorescent lighting systems requires care and attention to details, such as assuring that all connections in the branch circuit are secure and well made — especially the wires terminated at the ballast. Additionally, the case of an electronic ballast should be well grounded and securely mounted against a flat metal surface to achieve heat transfer. In general, excessive heat is the enemy of all electronic components, potentially shortening the life of the equipment and causing early component failure.
Fluorescent dimming systems, which will increasingly be specified in compliance with federal and state energy codes, require even more care. A dimming system involves the use of a central controller, a pair of low-voltage 18- to 22-gauge wires to provide control to each ballast, wall-mounted control stations, and remote hand-held controllers. Understandably, connecting all of these components means that the wiring is more complex. If you follow proper installation techniques, you can avoid the troubleshooting process later on. Here are some considerations to keep in mind when installing fluorescent dimming systems:
If 0V to 10V, powerline and digital electronic ballasts are retrofitted into existing fixtures, the sockets must be of the rapid-start type — not the jumpered (or shunted) instant-start type. Shunted socket terminations will damage the ballast and void the warranty.
The leads to the sockets from the ballast for linear lamps should not be more than 6 ft in length. With greater lengths, the ballast leads have increased capacitance, which lowers the voltage to the lamps. Never bundle the ballast leads.
Cap off the unterminated ends of all unused conductors. Never ground any of the control wires.
Flickering at low light levels can come from the lamps being too near or too far from the grounded surface of the fixture — the distance (or spacing) should be between ⅛ in. and ½ in. for linear lamps.
Don't install fluorescent fixtures in locations with considerable air movement, because the airflow can cause reduced light output and uneven dimming. All linear fluorescent lamps reach their maximum light output at a specific ambient temperature.
Compact fluorescent lamps in a dimming system should receive a 100-hr burn-in at full light output prior to dimming, or else the lamps can fail prematurely — and exhibit lamp striations, severe end blackening, and poor dimming performance at the low end.
Linear lamps should operate at full output for at least 12 hr before dimming for the first time.
Following are some situations you may encounter and the most likely reasons for them:
All fixtures stuck at full bright: control wires disconnected or incorrect ballast.
All fixtures stuck at full dim: shorted control wires, control wires at one or more devices may be cross-wired.
Lamps flicker at low light levels: long lead lengths, leads bundled too tightly, ballast not properly grounded, fixture not properly grounded, and lamps too close or too far from the ground fixture surface.
Three types of high-intensity discharge (HID) lamps in use for a number of years include: mercury vapor (MV), metal-halide (MH), and high pressure sodium (HPS). Effective Jan. 1, 2008, MV ballasts can no longer be manufactured or imported under the Energy Policy Act of 2005.
Dual-rated ballasts that listed MV and MH lamps will be changed so that MV lamps are no longer listed for use with these products. Use of MV lamps on a ballast not listed for MV lamps will violate the UL listing and the manufacturer's warranty. Thus, both the federal ruling and UL listing considerations focus attention on either MH or HPS lamps for specification.
The plasma arc within a MH tube contains a gas vapor much higher than that of a fluorescent lamp, and the spectral characteristic of the light produced also differs from a fluorescent source. The original MH lamp starting method placed an additional electrode, or probe, close to one of the electrodes to initiate a small arc at one electrode end to help the main arc to strike — thus the name “probe-start” lamp. The Security Act of 2007 halts the production of 150W to 500W probe-start MH magnetic ballasted fixtures, beginning this year.
The newest MH lamp design eliminates the starting electrode, using a higher voltage across the arc tube instead — thus the name “pulse-start” lamp. A user must now select either a quartz arc tube or a ceramic arc tube, pulse-start, metal halide. The quartz MH lamps are horizontal and universal-burning models in 175 to 450W ratings, with efficacies up to 70 lumens per watt, a CRI up to 96, and an operating life up to 20,000 hr. The slightly more expensive ceramic arc lamp offers improved color rendering and minimum color variation over an operating life beyond 20,000 hr.
As with fluorescent lamps, the industry trend is toward electronic ballasts (for lamps up to 450W) rather than the original core and coil magnetic type. Electronic HID ballasts with a continuously dimmable feature down to about 50% of full output power are available, and they have greater dimming efficacy than magnetic types.
Ballasts that are remotely mounted to reduce ballast hum should follow specific considerations to avoid early ballast failure and ensure proper starting. The spacing between ballast cases should be at least 12 in. in all directions, and they should not be directly mounted to a nonmetallic surface to ensure proper heat dissipation. Maximum ballast-to-lamp distances and minimum wire gauge size should be observed to ensure reliable starting. The fixture manufacturer generally offers tables for both conditions.
For ballasts using ignitors (HPS, low-wattage MH, and pulse-start MH), the distance is determined by the ignitor used. Increasing conductor size may not boost the allowable maximum distance, and might instead increase wire capacitance, which can further attenuate the ignition pulse. The ignitor provides a peak pulse of at least 2,500V to initiate the lamp arc.
An ignitor is designed to operate with specific ballasts and cannot be interchanged with other ignitors or different brands of ignitors and ballasts. The ignitor should be mounted near, but not on, the ballast. For ballasts without ignitors (mercury and probe-start MH), the ballast to lamp distance is determined by conductor size to keep any voltage drop below 1%.
For HID lamps, the usual end of life indications include: low light output, failure to start, or lamp cycling off and on. At the end of their lives, mercury and MH lamps will have low light output and/or intermittent starting. Visual indications include blackening at the ends of the arc tube and electrode deterioration. The only sure test of a defective lamp is to install a new one.
A voltmeter with true rms sine wave readings and an ammeter can be a help in troubleshooting an HID fixture. The line voltage should be checked at the fixture, not at the panel, to account for any voltage drop. For constant wattage ballasts (CWA, CWI) the voltage should be within ±10% of the nameplate rating. For reactor (R) or high-reactance (HX) models, it should be within ± 5% of the rating. High-, low-, or variable-voltage readings can be caused by load fluctuations. If the supply voltage is proper, measuring both the open-circuit voltage and short-circuit current can usually reveal the problem.
If the ballast has an ignitor, it must be disconnected or disabled by placing a capacitor (1000pF or larger) across the voltmeter input to protect the meter from the ignitor pulse. A ballast may have an integral ignitor, so the safest procedure is to insert a capacitor across the meter for all open-circuit voltage measurements.
CFLs are the mandated substitute for most incandescent lamps in applications with extended burning hours. They are available in a range of sizes from 5W to 80W.
A CFL should be matched to a specific ballast type from a particular manufacturer. A CFL that is incompatible with a particular ballast model may not properly start and operate or it could experience early failure.
Each CFL type has specific electrical characteristics, such as internal impedance, and every manufacturer develops its own ballast circuitry, which undergoes evolutionary changes over time.
For example, one ballast maker recently changed the design for its 24W CFL ballast, but before going into production, it did not test the new design with the PL-L (long) lamp, which has four pins in a line (2G11 base). It turns out that the new ballast design is not compatible with the PL-L lamp type, but it is suitable for the PL-C Cluster 4-pin lamp.
Generally, an inscription on the ballast lists all of the lamp types for which it is compatible. Installers should carefully check this list prior to starting work on a project.