THIS PROJECT HAS RISK OF SHOCK AND FIRE!
The HID lamp for automotive headlight has been developped in latter half of '90s. It has good luminous efficacy as several times better than halogen lamp used for the headlight until then, so it can obtain high luminous flux with low power consumption. The HID lamp is an ideal light source that can be considered nearly a point source and it makes the light distribution controls easy. The initial cost of the HID lamp system is higer than halogen lamp, however, thanks to the long life of HID lamp, it is more economical than halogen lamp in the long-term. Therefore the HID lamp has been widely used for the automotive headlight system in the end of '00s.
Now, it can be saied that the automotive HID lamp is an only HID lamp widely spread in the home, so every light bulb maniac will not able to help but light it with their DIY lamp driver. However, the automotive HID lamp has some special features required for automotive headlight and the HID lamp ballast for generic lighting cannot light it. Therefore I built another HID lamp ballast for the automotive HID lamps as a secondary project of Universal HID Lamp Ballst that I did it last year. Please refer it before read this article.
Figure 1 shows an automotive HID lamp. There are four type of lamp specifications, D1, D2, D3 and D4. The specifications of Dx lamps are defined in the ECE Regulation No.99, and also some types of HID lamp with hallogen lamp mount are available in the market for retrofitting. The automotive HID lamp is often called "xenon lamp" or "discharge lamp" and these are not that false, but the kind of automotive HID lamp is actually metal halide lamp. The specifications of automotive HID lamps are shown in Table 1. Only a significant electrical difference divides the types is the lamp voltage. The D1/D2 lamp contains mercury as typical metal halide lamp, while the D3/D4 lamp is in mercury-free. The difference between D1 and D2 (also D3 and D4) is in mechanical shape of the lamp base.
Item | D1/D2 | D3/D4 |
---|---|---|
Lamp Power | 35±3 W | |
Lamp Voltage (square wave) | 85±17 V | 42±9 V |
Luminous Flux | 2800±450 lm | |
Chromaticity Coordinate | x=0.375, y=0.375 (4000K) | |
Run-up (luminous flux) | ≧25 % at 1 sec., ≧80 % at 4 sec. | |
Hot Restrike | ≦10 sec. after power off |
The unique requirements for the automotive headlight are the bottm two items in the table, instantaneous rise of luminance and hot restrike. These are the weak points of the metal halide lamp. To meet the former one, the discharge tube is filled with high pressure xenon instead of argon as the buffer gas and it works as xenon discharge lamp until the mercury and the metal halide in the tube evaporates. The tiny discharge tube is driven in a special control profile to rise the temperature quickly. To meet the latter one, an igniter with very high output voltage is used for the hot restrike.
The automotive HID lamp ballast is in electronic control. Figure 2 shows a typical configuration of the automotive HID lamp drive system. It is almost same configuration as electronic HID ballast for generic lightings. The DC power supply from a battery is up-converted with a flyback converter upto 400 V and following DC-AC inverter drives the lamp at a frequency in range of 100 to 400 Hz.
The igniter is located in the ballast, on the cable or in the lamp holder depends on the system design. Because the discharge tube is filled with high pressure xenon gas, breakdown voltage is much higher than the generic HID lamps, several kV at DC and ten several kV at short pulse. Also the hot restrike is required. For this reasone, an igniter that can produce very high output voltage pulse, 20 - 30 kV of peak voltage, is used for the automotive HID lamp driver.
Then there is an idea to enable the generic HID lamps hot restrike with that igniter. This is not true, because every component, luminaier, lamp socket, lamp base, and etc, do not withstand such high voltage pulse, except for double end lamp with special luminaier.
HID lamp | D1/D2 automotive HID lamp |
Lamp power (Po) | 35/50 W |
Lamp current (Io) | 1 A max |
Ignition voltage (Vo) | 380 V |
Igniter | Embedded on board (>25 kVPK) |
Controller | STM32L010 |
Power input (Vi) | AC90 to 125 V, 50/60 Hz |
Efficiency (η) | ~93 % (Vi=100 V, Po=35 W, measured) |
Power factor | ~67 % (Vi=100 V, Po=35 W, measured) |
Table 2 shows the specifications of the built HID lamp ballast. It works in AC input while the automotive HID lamp ballasts need to work in DC input at 12 or 24 V. The reason why I tried to built it as AC powered lamp ballast is that the goal is simply to light the automotive HID lamps and there is no sense to make the same thing that already available.
The PFC feature was eliminated from the former design for the low lamp power, so that the input power factor is not good. The lamp power can be set 35 W or 50 W with a jumper switch. The 50 W mode is added with a design margin to support non-standard 55 W lamps sold for the replacement.
Figure 3 shows the circuit diagram of the built automotive HID ballast. There are some probing points indicated in red and blue letters that to be referred below.
The lamp voltage varies in wide range, 400 V on ignition, lowers to 20 V in warming-up and rises to 85 V in steady state. For this reason, flyback converter is usually used for the automotive HID lamp ballast. The turn ratio of transformer is set about 1:6 in typical case. I was going to design the DC-DC stage as a flyback converter in 1:1 of turn ratio, but I changed it to SEPIC converter in the end. The reason why I chose it is that I found that it can be in SEPIC converter for that turn ratio and it eliminates the snubber and the both windings are used efficiently. T1 is the transformer got from a junk power supply and I re-wound it as a 400 μH coupled inductor.
The C10 with resistor-diode network is a boost circuit to force an inrush current to the lamp on take-over. This ensures the hotspot on the electrodes be stable and the arc discharge starts at the first ignition.
The automotive HID lamp has symmetrical electrodes that the same as the HID lamps for generic lightings, so that it needs to be driven in AC. In order to avoid acoustic resonance phenomenon, the drive frequency below 1 kHz is chosen in most case. In this project, the lamp is driven at 250 Hz.
The T2 in series of the inverter output is an ignition transformer to produce the high voltage pulse. Figure 4 shows the built ignition transformer in 5 turns of copper strip as the primary winding and 240 turns of wrapping wire as secondary winding with a ferrite rod in D=6, L=30. When I checked the ignition pulse, I found an odd waveform in suppressed peak that was not expected. I unwound the secondary winding to see what is wrong and found stains on the wire. This is what caused by dielectric barrier discharge due to too high voltage between the layers in simple pallarell winding. The insulation was not that broken, however, it is an unwanted condition that attenuates the output voltage, so that I rewound the secondary winding with silicone sealant to prevent the dielectric barrier discharge. Therefore, high voltage transformers want a special care on the insulation and the unexpected discharge.
Figure 5 shows the control diagram of the SEPIC converter. The conduction mode of the inductor is in C-CCM which used in the PFC converter in the previous project, Universal HID Lamp Ballst. The output current is controlled in the same method that in only software estimated feedforward control without inductor current feedback. It is in a rough model and it resulted some gain error of 20% maximum. This is not a problem, because the current control is a minor control. The major control, the power control, operates in closed-loop that feedback the lamp voltage and the lamp current to control the lamp wattage correcly.
Figure 6 shows the waveforms in operation that captured in Ch1 as volgate A, Ch2 as voltage B and Ch4 as current C. The ripple voltage in cycle time of several milliseconds is from alternation cycle of the DC-AC inverter. From the magnified waveform at Ch1, the estimated C-CCM operation and valley switching seem to be working well as expected. As the result, the switching loss at the transistor was quite low as no heat sink is needed, but, on the ohter hand, the temperature rise at the transformer was significantly high, 38℃ at 35 W and 62℃ at 50W on the winding. The ripple current in C-CCM operation is higer than CCM. The windings of transformer in 0.5 UEW × 2 strands each is not proper for such high ripple current applications, so that litz wire should be used if available.
Figure 7 shows the waveform on the cold start captured at the inverter output (voltage D). The reason why it lowers the frequency briefly is that to drive each electrode in DC operation to stabilize the arc discharge.
To achieve the quick start-up, automotive HID lamp is over-driven in twice of lamp power until the lamp warms up. It enables a quick temperature rise of lamp and covers low luminous efficacy of xenon emission. The temperature is estimated in the lamp voltage and lamp power is adjusted higher as the voltage lowers when the lamp voltage is under 60 V. However it is limited to 75 W or 2.6 A to reduce the wear of electrodes. Note that this feature is not implemented in this project and the warm-up characteristics does not meet ECE99 regulation.
HID lamps can exhibit abnormal behavior at the end of life. If it get into this situataon, the lamp ballast must safely stop the operation. To ensure the start of discharge and monitor the lamp operating status, a state diagram is defined as shown in Figure 8.
Waiting for power-good. DC-DC converter is disabled. It always enters this state on power-on or BOD. When bus voltage gets in the working voltage range, the DC-DC converter starts and CT1,CT2 are cleared and go to IGNITION state.
Enable the DC-AC inverter and ignition pulses are applied to the lamp. If breakdown and take-over succeed, clear CT1 and go to RUNNING state. If it does not succeed and T1 time erapses, increase CT1 and go to WAIT state if CT1 < N1, or the ignition fails (end of lamp life or open circuit) and go to FAULT state.
After the discharge start, the lamp will reach thermal equilibrium within a minute. When the lamp voltage get into the rated range and T3 time elapse, the lamp is considered in stable state and CT2 is cleared. If an extinction, an abnormal lamp voltage or a short circuit is detected, exit this state. The abnormal lamp voltage is a state that the lamp voltage is out of rated range for T2 time. It is due to low lamp voltage caused by leakage of arc tube, warm-up failur caused by leakage of outer tube or high lamp voltage caused by end of life. When it exit this state, CT2 is increased. If CT2 < N2, go to IGNITION state, or it is considered that it is unable to get into stable state and go to FAULT state.
Inverter is disabled to prevent to output the ignition pulses continually. Wait for T4 time and go to IGNITION state.
An insanity is detected. All functions are shutdown and keep this state until power off or BOD.