HID light startup sequence

Infineon / Mitsubishi / Fuji / Semikron / Eupec / IXYS

HID light startup sequence

Posted Date: 2024-02-03

Ignition and takeover

Apply a voltage above 300V that is sufficient to take over (past the peak on the right in the diagram and start arcing). However, the HID lamp is still insulated, so the high-voltage pulse superposition causes the arc tube to electrically breakdown (over the peak on the left in the diagram).

After discharge begins, sufficient current needs to be provided (making the power higher than one-fourth of the rated value, but not more than twice the rated current). The lamp impedance during glow discharge is above 1kΩ, so a lamp voltage much higher than the rated voltage is required. When a hot spot forms on the electrode and arcing begins, the lamp impedance drops dramatically. If the takeover fails and the discharge goes out, the high voltage pulse is applied again. The process is almost instantaneous.


After the arc discharge begins, the lamp impedance is as low as a dozen ohms, and due to current limitations (twice the lamp rating or ballast capacity), the lamp cannot be supplied with rated power. As the temperature increases, the materials in the arc tube (mercury, metal halides, sodium amalgam, etc.) evaporate and begin to emit intense light from the excited metal atoms. As the vapor pressure increases, the lamp impedance increases and therefore the lamp power increases. After reaching the rated power, the control enters the constant power mode, but the lamp impedance continues to rise and the current decreases.


After a few minutes, the lamp will reach thermal equilibrium and the impedance of the lamp will become stable. Steady-state lamp voltage varies depending on specifications, generally around 85 to 115V.

When the HID lamp is working, the gas pressure can reach from 1 atom to several atoms. After the lamp is turned off, the gas pressure and breakdown voltage remain high for a period of time, so it cannot be restarted until the lamp temperature drops to a sufficient level to light it again. Relight time depends on the thermal design of the lamp and fixture, ranging from a few minutes to more than ten minutes for mercury and metal halide lamps, and less than a minute for high-pressure sodium lamps.

Lamp driver configuration

Figure 12. Lamp driver schematic

Figure 12 shows the schematic of the driver block. The red and blue symbols in the schematic correspond to the probe points in the screenshot shown in this article.

buck converter

Initially, I planned to use C-CCM, but the ripple current was large and a large-sized inductor was needed to meet the working conditions. So I used the spreadsheet buck.xlsx to calculate and decided to use CCM for the buck converter. Due to the hard switching of hundreds of volts, transistor Q2 dissipates a very large 3W.

When the low-side switch is a diode as in this configuration, the floating supply to the high-side driver cannot charge in the initial state and at no load. For this purpose, charge register R9 and bleed register R41 are placed to ensure initial charging at startup and freewheeling operation after startup (switching node drops to 0V), even without load.

Figure 13. Lamp control diagram

Figure 14. Waveform

Figure 13 and Figure 14 show the block diagrams of the lamp control and operating waveforms respectively (Ch1: point H, Ch2: point J, Ch4: point G). The buck converter is constant power controlled, so it not only feeds back the output current buf, but also feeds back the output voltage. The current control loop, the secondary loop, needs to be controlled responsively, so the PID parameters can be configured for each lamp configuration. The power control loop is the main loop and does not need to be controlled so fast (too fast will interfere with the current control and cause the light to flicker), so it is only controlled in item I. These control parameters can be preset and selected via thumbwheel switches.


The DC current output by the buck converter is converted into 200Hz AC current through the H-bridge inverter and provided to the HID lamp.


Figure 15. Ignition pulse

The output voltage remains at about 360V when no load (before discharge starts). Ignition capacitors C30 and C31 are charged in the voltage doubler. When the capacitor voltage reaches the breakdown voltage of GDT (SG1), approx. At 600V, the GDT is turned on, the pulse current flows out through the primary winding of the ignition transformer T1, and the high-voltage pulse on the secondary winding will be repeatedly applied to the lamp. The igniter has no on/off control, but the generation of pulses ceases when discharge begins and the lamp voltage drops.

T1 is an ignition transformer made from three junk toroidal cores with 4 turns (4.3μH) primary and 40 turns (430μH) secondary with Teflon wire. Figure 15 shows the ignition pulse waveform when no wires are connected. It reads peak voltages of 6kV and very fast transient voltages, 80kV/μs, so special attention needs to be paid to the traces on the circuit board. I haven't had any problems with the MCU, but some of the transistors and gate drivers in the H-bridge sometimes fail from the high voltage pulses of the discharge and go somewhere on the board. Since the pulse voltage is attenuated by the stray capacitance of the cable connected to the lamp, the cable length (usually 1.5 to several meters) is specified in the specifications of the electronic ballast.

State diagram

Figure 16. State diagram

HID lamps can exhibit unusual behavior at the end of their useful life. If this condition is entered, the ballast must be safely taken out of service. In order to ensure the start of discharge and monitor the working status of the lamp, a state diagram is defined, as shown in Figure 16.


Wait for the power supply to be good. Both converters are disabled. It always enters this state at power on or BOD. When V IN enters the operating voltage range, the boost converter starts, and when V BUS reaches the specified voltage, CT1 and CT2 are cleared and enter the IGNITION stage.


Start the buck converter and inverter and apply an ignition pulse to the lamp. If the ignition takeover is successful, clear CT1 and enter the RUNNING state. If unsuccessful and T1 time is exhausted, if CT1


After discharge begins, the lamp reaches thermal equilibrium within a few minutes. When the lamp voltage enters the rated range and after the T3 time, the lamp is considered to be in a stable state and CT2 is cleared. If it detects extinguishing, abnormal lamp voltage or short circuit, it will exit this state. Abnormal lamp voltage refers to a state in which the lamp voltage exceeds the rated range within the T2 time. This is due to the leakage of the luminous tube causing low lamp voltage, and the leakage of the outer tube causing preheating failure or end of life resulting in high lamp voltage. When exiting this state, CT2 is incremented. If CT2


The inverter is disabled to prevent continuous output of ignition pulses. Wait for T4 time and enter the IGNITION state.


Delirium detected. All functions are off and remain in this state until power outage or BOD.

#HID #light #startup #sequence