AN10803 Triac dimmable CFL application using the UBA2028/UBA2014/UBA2027x Rev. 2 — 27 February 2012 Application note Document information Info Content Keywords UBA2028, UBA2014, UBA2027x, dimmable, CFL, Triac, dimmer, charge pump Abstract This application note describes how to design a dimmable CFL using the UBA2028, UBA2014 or UBA2027x. AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x Revision history Rev Date Description v.2 20120227 second issue • Modifications: v.1 20091009 Text and drawings updated throughout entire data sheet. first issue Contact information For more information, please visit: http://www.nxp.com For sales office addresses, please send an email to: [email protected] AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 2 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x 1. Introduction WARNING Lethal voltage and fire ignition hazard The non-insulated high voltages that are present when operating this product, constitute a risk of electric shock, personal injury, death and/or ignition of fire. This product is intended for evaluation purposes only. It shall be operated in a designated test area by personnel qualified according to local requirements and labor laws to work with non-insulated mains voltages and high-voltage circuits. This product shall never be operated unattended. This application note describes how to design a dimmable Compact Fluorescent Lamp (CFL) circuit with the UBA2028, UBA2014 or the new UBA2027x series. The differences between the versions are the ones that have integrated MOSFETs and controller versions that use external MOSFETs. The versions with internal MOSFETs can drive burners to 18 W. Use the versions with external MOSFETs for larger burners. The control parts of the different versions are identical, so the same calculation tools can be used. Remark: Unless otherwise stated all voltages are AC. The design of a Triac dimmable CFL is a complicated task. Interaction between the dimmer and CFL, tolerances, aging of components and, last but not least the circuit must not be expensive affects the dimming behavior. Most components serve multiple purposes, so it is not always possible to choose the optimum value for each purpose. There is no guarantee that the designed circuit works with all dimmers, because it is not known in advance which dimmer type the CFL is connected to. 2. Triac dimmers Virtually all domestic wall dimmers are phase cut dimmers. This means that the supply voltage to the lamp is reduced with cutting the phase see Figure 1. In practice, this condition is implemented by placing a switch. An example is a Triac in one of the supply lines that is not conducting during a certain part or angle (see Figure 1) of the mains period. Effectively, this reduces the RMS voltage supplied to the lamp. Vmax φ φ 019aab989 Fig 1. AN10803 Application note Phase cut sine wave All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 3 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x Phase cut dimming with a Triac works very well for dimming incandescent lamps. However, dimming CFL lamps with this principle is more complicated due to the interaction between the dimmer and the CFL. It is important to understand how a wall dimmer works before designing a dimmable CFL lamp. Figure 2 shows the circuit diagram of a basic wall dimmer with Triac. incandescent lamp 470 kΩ L 2.7 kΩ diac K2 triac 230 V AC K1 100 nF N 019aab960 Fig 2. Basic 230 V dimmer The capacitor is charged via the combination of a fixed and a variable resistor. The dimming angle is set with the variable resistor. When the resistance is low, the capacitor is quickly charged until the breakover voltage of the Diac. The Triac is then triggered and immediately a current starts to flow from k1 to k2. This current continues to flow until the Triac blocks when the current drops below the minimum hold current IH. As the incandescent lamp is a resistive load, this occurs at the zero crossing of the mains input. The Triac is a bidirectional device that works in two quadrants (see Figure 3), so the same process is repeated during the negative half cycle. I latching current lL breakover voltage minimum holding current lH gate triggering range -V V minimum holding current lH breakover voltage latching current lL -I Fig 3. AN10803 Application note 019aab982 Typical Triac VI characteristic curves All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 4 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x Normal operation of the dimmer requires the presence of a conducting or charging path for Cdim when the Triac in the dimmer is not conducting. It requires a low ohmic load for sustainable conduction of the Triac when the Triac is conducting. The incandescent light bulb see Figure 2 provides all of this. The situation is different when a CFL or other load that contains a bridge rectifier is connected. It depends on the moment of switch-on to determine if current flows through the Triac. This is only the case when the (momentary) input voltage is higher than the DC bus voltage. See Figure 5. DIMMER CFL 470 kΩ 2.7 kΩ VBUS Cdim L 100 nF load Cbus N 019aab961 Fig 4. Basic dimmer with CFL The Triac switches off when the current drops below the hold current IH. Here it is not the moment at which the input voltage crosses zero, but the moment where the momentary input voltage drops below the DC bus voltage. In practice, this moment is soon after the Triac is switched on. Most of the time, the bridge diodes are not conducting, so there is no conductive path to charge the capacitor in the dimmer. This is fine when the dimmer phase cut is set to a very small angle and Cdim charges enough to trigger the Triac within the period that the mains is higher than the bus voltage. However, when the phase cut is set to a larger angle the charge of Cdim is not completed in a single half cycle and triggering of the Triac is at irregular intervals as shown in Figure 5. AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 5 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x VBUS VMAINS time too short to charge Cdim ITRIAC Triac triggered VCdim 019aab974 Fig 5. Input voltage and current when large dimmer phase cut is set The result is that the Triac is triggered at irregular moments resulting in the input voltage of the CFL is irregular. The dimming level of the lamp is retrieved from the input voltage, so an irregular input voltage results in an irregular dimming level or lamp flicker. The latch current threshold of the Triac must be exceeded before the gate trigger current is removed. The current through the Triac must be maintained above the Hold current threshold IH, until the next zero crossing of the mains voltage to ensure a smooth operation of the dimmer. A hold current of 15 mA to 20 mA is sufficient for most dimmers. A simple solution that generates the hold current, is to place a resistance across the input. However, the hold current must also flow at low input voltages, so the resistance must be very low resulting in huge losses. For that reason, a current source behavior is required. This can be realized with a Power Factor Corrector (PFC). It can either be a boost converter or a charge pump-based PFC. In the latter solution, a capacitor is used as an energy storage element to draw current from the mains at low voltages. The practical implementation consists of two diodes and a capacitor and is discussed in detail in Section 3.3. This application note addresses the charge pump solution because it is the most cost effective solution. The charge pump circuit enables a conductive path as long as the lamp is on. However, if the phase cut angle is set too large, the supplied voltage becomes too low for the circuit to maintain the lamp current and the lamp turns off. Now there is no conductive path any more through the CFL to charge Cbus in Figure 4. In practice, there is some leakage through the filter components in the dimmer, so after some time the Triac is fired. Then it depends on the potentiometer setting in the dimmer if the CFL starts, or begins in a flash mode. This start-up behavior heavily depends on the design of the dimmer. AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 6 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x 3. Designing the application 3.1 Block diagram Figure 6 shows the block diagram of the dimmable CFL with charge pump. The phase cut mains voltage is supplied to the CFL. The input voltage is rectified and supplied to the inverter, and also used for detecting the phase cut adjustment of the dimmer. This is done by measuring the average input voltage and supplying it to the control input of the inverter control IC. That is, the CSP pin of the UBA2014 or UBA2028 and the DCI pin of the UBA2027x. The dimmer phase cut detection must be non-linear. That is, the lamp must be dimmed to 10 % before the bus voltage becomes too low and the inverter switches off due to low supply voltage. voltage source charge pump CFL INVERTER mains DIMMER FILTER DIMMER PHASE CUT DETECTOR 019aab963 Fig 6. Block diagram of the charge pump CFL dimmer In practice, the lamp must already be dimmed to 10 % at 120 degrees phase cut and preferably maintain the lowest dim level to 130 degrees. Beyond that value, the lamp is off, but the inverter is still switching and providing a charge pump current. Without the charge pump current, the CFL input is high (ohmic), and the input voltage is no longer a reliable representation of the dimmer position. 3.2 Resonant tank Once the gas in the CFL tube is ionized however, its impedance becomes negative. This means that the more current is flowing through the tube, the more conductive it becomes The increased current increases the degree of ionization of the gas. Therefore, add some form of current limiter to the lamp to prevent the current increasing to a level where the tube is destroyed. An inductor is placed in series with the tube therefore, the tube is supplied with a current, rather than with a voltage. A capacitor is placed in parallel to the tube that creates a resonant circuit or resonant tank in combination with the series inductor to ignite the tube. Before ignition, the lamp impedance is almost infinite, so the quality factor of the resonant tank is high. At start-up and before lamp ignition the switching frequency is above the resonance frequency and goes down in frequency until the tube ignites. AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 7 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x 019aaa341 3 Vlamp / Vhb (V) 2 (1) (1) 1 (2) 0 0 20 40 60 80 f (kHz) Fig 7. Transfer of resonant tank versus frequency before and after ignition For the calculation of the resonant tank, the first harmonic approach is used. This approach works well, especially near the resonant frequency. Equation 1 shows the RMS value of the first harmonic of the half-bridge voltage. 2 1 4 V BUS V I RMS = ------- --- ------------ = ------- V BUS 2 2 (1) The input parameters for the first estimation of the resonant tank are as follows: • • • • • Nominal lamp current Nominal lamp voltage Nominal operating frequency Minimum bus voltage The phase angle between the input voltage and current of the resonant tank VLR LR VI 1 ILamp VI CR REQ VLamp 2 ICR II ILamp VLamp 019aab980 Fig 8. Schematic resonant tank and vector diagram An equivalent resistor (REQ) is calculated from point 1 and 2. Equivalent resistance is only valid at nominal lamp power. The operating frequency (fnom) at nominal lamp power, is set between 40 kHz and 50 kHz. AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 8 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x It is important to guarantee Zero Voltage Switching (ZVS) of the half-bridge stage to have low losses in the switching devices in the half-bridge driving the resonant tank. Therefore the resonant tank must have an inductive behavior, so the input current must lag the input voltage. The larger the phase angle between the input current and voltage, the easier it is to maintain ZVS. However, when the phase angle is made too large, the losses become potentially high. In general a value between 30 and 45 degrees is chosen. The voltage transfer of the resonant tank is: V lamp 1 ----------------- = -------------------------------------------------------------------------L 2 V I RMS 1 – nom LC + j nom ---------- R EQ with nom = 2f nom (2) The tangent of the phase angle between the input voltage and current of the resonant tank, is as calculated in Equation 3: 2 L – tan = nom --------- nom R EQ C + nom R EQ LC R EQ (3) When these two equations are combined, the capacitance value can be calculated using Equation 4: 2 2 2 2 – V I RMS + V lamp + V lamp tan C = --------------------------------------------------------------------------------------V I RMS R EQ nom (4) The value of the inductor is as calculated in Equation 5: tan + nom R EQ C L = -------------------------------------------------- nom 2 3 ------------ + nom R EQ C R EQ (5) Now the inductor and capacitor values are fixed and the dimming behavior can be determined. The control loop of the inverter works by setting a reference value for the lamp current. It adapts the half-bridge frequency until the desired lamp current is reached. Because of the negative impedance of a fluorescent lamp, the lamp voltage decreases when the lamp current increases. The exact relation between lamp current and lamp power depends on the lamp type. The lamp type must be known to calculate the relation between lamp power and frequency. The relation between lamp power and operating frequency is given in Equation 6: V la I la AN10803 Application note V la V la 2 ------- 1 + ------C I la I la = ------------------------------------------------------------------------------------------------------V la 2 V la 2 2 V la 2 3 ------ + L – ------ C + L ------ C I la I la I la All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 (6) © NXP B.V. 2012. All rights reserved. 9 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x Solving Equation 6 yields: 1 2 2 2 3 2 2 4 4 2 2 2 2 f = ------------------------------ 2LCV la – L I la + – 4L CI la V la + L I la + 4L C V la V I 2 2LCV la (7) With this equation, the power can be plotted versus frequency. See Section 5. 3.3 Charge pump circuit The charge pump circuit is used in many applications to improve the power factor of the CFL. In this application, it is also used to draw hold current through the Triac to keep the Triac on, following ignition. In an ordinary rectifier with an inverter, current is only drawn from the mains at around the maximum input voltage. Figure 9 shows an inverter with a voltage source charge pump. Adding two Hi-Speed DX and DY diodes and a capacitor CX, makes it possible to draw current from the mains during the whole cycle. The CFL inverter with charge pump power feedback (see Figure 9) is replaced by its equivalent circuit (see Figure 10). A voltage source represents the lamp voltage, the rectified mains voltage and the voltage across the bus electrolytic capacitor. DX DY CX VMAINS LR CDC CBUS VI CR VBUS lamp 019aab964 Fig 9. CFL inverter with charge pump power feedback The lamp voltage or equally the voltage across CR (see Figure 9) can also be considered a high-frequency voltage source. Therefore Vla is equal to the lamp voltage and Cx can be connected to this point. The capacitance Cx is charged from VI via Dx and is discharged to VBUS via DY. A charge package is “pumped” from VI to VBUS in each inverter switching cycle of Vla. VI is the rectified mains voltage that has a ripple that is twice the mains frequency. The parameter of interest is the average current during one inverter switching cycle (II(ave)) that is drawn from the mains. In Ref. 1, it is shown that this current is calculated with the following equation: I I ave = C x f s V I + 2V lamp – V BUS (8) Where: • fs is the inverter switching frequency • Vlamp is the amplitude of the lamp voltage AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 10 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x Equation 8 shows that the input current is following the input voltage when the peak-to-peak lamp voltage (= 2 Vlamp) is equal to the bus voltage. The power factor is equal to 1 in this situation (unity power factor condition). DY II DX CX Vi Vla VB Req 019aab966 Fig 10. Equivalent circuit of a voltage source charge pump When Cbus Figure 9 is directly charged from the mains, both diodes Dx and DY are conducting and Cx is then directly in parallel with CR. Therefore by connecting Cx to the resonant tank, Cx highly modulates it during a switching cycle. The total capacitance that is in parallel to the lamp is given in Equation 9. V I + 2V lamp – V BUS C EQ = C R + C X ---------------------------------------------------2V lamp (9) Equation 9 reduces to Equation 10 when 2 Vlamp = VBUS: VI C EQ = C R + C X -----------V BUS (10) The use of a voltage source charge pump circuit has some drawbacks: Unity power factor is only obtained when Vbus = 2 Vlamp. However, this can only be fulfilled when the lamp is operating at a constant level and is not dimmed. When the lamp is dimmed, the frequency goes up and the tube voltage rises because of negative incremental impedance of the tube. The unity power factor condition is not fulfilled any more. The voltage across the tube is not only much higher when the lamp is dimmed, but also when the lamp is in the preheat or ignition phase. Both are considered as low-power situations and cause the charge pump to absorb a lot more energy from the mains than the load consumes. This extra energy is stored in the buffer capacitor Cbus. It causes the bus voltage to exceed 800 V which is much higher than the voltage rating of the bus capacitor and driver IC. Another drawback is the modulating effect of Cx on the resonant tank. This causes the crest factor of the lamp to become as high as 2.6, reducing lamp life and increasing the possibility of visible lamp flicker. A crest factor of 1.7 is preferred. A way to reduce these problems is to place an extra capacitor across DY. This capacitor improves the crest factor to a value of 1.6. It reduces bus voltage stress by 60 % while still providing a cost effective way to produce the necessary hold current for a dimmable CFL application. AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 11 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x This all comes at a price of 1.6 times higher current stress on the half-bridge switches so this technique is most suited for the drivers with external output devices. See Ref. 3 for more information on the voltage source charge pump with improved crest factor (VSCP with ICF). CY DY II DX CX Vla Vi VB aaa-002499 Fig 11. VSCP with ICF Another voltage source charge pump which can be considered is the voltage source charge pump with low-frequency second resonance (VSCP with LFSR). This topology uses more components and two inductors. However, it has a good power factor and a low bus voltage stress over the full dimming range: DX DY LR2 VMAINS CX CDC DR1 LR1 CBUS VI CR lamp Cp VBUS DR2 aaa-002500 Fig 12. VSCP with LFSR 3.4 Voltage doubler and charge pump for 120 V input When the mains voltage is 120 V, then the bus voltage is too low to supply the lamp. For that reason, a voltage doubler is used instead of a full bridge rectifier. Figure 13 shows the voltage doubler circuit. CH is charged during the positive half-cycle and during the negative half-cycle CL is charged. If there is no load, the bus voltage becomes equal to the mains peak-to-peak voltage. CH VMAINS VBUS CL 019aab984 Fig 13. Voltage doubler circuit AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 12 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x The electrolytic capacitors in this circuit have a double capacitance and half the voltage with reference to the 230 V circuit. Figure 14 shows the voltage doubler combined with the charge pump circuit. A symmetrical voltage source charge pump replaces both rectifier diodes. The two top left diodes charge energy into CH at the positive half cycle of the mains. The two left bottom diodes charge energy into CL during the negative half cycle of the mains (see Figure 14). Instead of using two high-voltage charge pump capacitors driving each charge pump from the resonant tank, a single high-voltage charge pump capacitor and an AC coupling capacitor between the two charge pumps are used. Because only one set of diodes is conducting during each half mains cycle, the energy stored in Cp is automatically stored to either CH or CL only, depending on the mains polarity. A coupling capacitor of 100 nF is more than sufficient to have a neglecting influence compared to the value of Cp. By using only a single capacitor CY a charge pump with improved crest factor is created. CX CY CH CAC VMAINS VBUS CL 019aab983 Fig 14. Voltage doubler with charge pump circuit 3.5 Current feedback control loop With the UBA2014 or the UBA2028, it is easy to control the lamp current as a controller is implemented in the IC. VBUS UBA2014 UBA2028 UBA2027x VCONTROL DCI VBUS VCONTROL CSP VCO DCI CDC LR CSN lamp CSW CR OTA VCO CDC LR lamp CSI VSENSE RSENSE CR VSENSE RSENSE 019aab967 Fig 15. Current feedback control loop AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 13 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x The lamp current is measured across a sense resistor. The UBA2014 and UBA2028 are single-sided rectified externally. The UBA2027x family is double-sided rectified internally. The measured value is compared with a control voltage that is applied to the CSP pin. The control loop changes the half-bridge frequency until the measured voltage is equal to the control voltage, so the RMS lamp current can be calculated with: For UBA2014 and UBA2028: V control I lamp RMS = ------- -----------------2 R SENSE (11) Due to UBA2027x families' overpower protection, the maximum RMS lamp current is limited to: 1 I lamp RMS = -------------R sense (12) In the Triac dimmable CFL application, the control voltage is derived from the average mains voltage in the dimmer position detector. 3.6 Dimmer phase cut detector The control voltage for the current feedback control loop is derived in the dimmer phase cut detector. The dimmer phase cut detector is based on average voltage detection of the phase cut mains voltage and converted to a lower voltage average reference voltage. The average reference voltage is supplied to the CSP pin in case of UBA20214/UBA2028 and the DCI pin in case of the UBA2027x family. Ideally, the lamp lumen is linear with the position of the dimmer control. Figure 6. The human eye is more sensitive to small light changes at low lumen level, than to large light changes at high lumen levels. The lamp current has a direct linear relation to the lamp lumen output. When the phase cut dimming angle is detected by using an average voltage detection, the result is that the average control voltage curve follows a cosine function. The cosine slope has a small level change at the high lumen levels and a large change at mid to low lumens. Typically the UBA2014 and UBA2028 have a linear control voltage to lamp current transfer (see Figure 16 left image). The response to a dimmer is observed as less natural because the dimmer must be set halfway to notice response in the lumen output of the lamp. Also, the lower dim levels are more difficult to control. A way to overcome this problem is using extra diodes in the lamp current sense which helps at the lower dim rage. However, a better solution is to make the control voltage to lamp current transfer function non-linear. This is what has been implemented in the UBA2027x family, its transfer function is shown on the right of Figure 16. It is called the natural dimming transfer function and makes better use of the control rage of the dimmer. AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 14 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x light output light output max. max. phase cut angle min. phase cut angle min. max. max. 019aab968 Fig 16. Dimming curve UBA2028 versus UBA2027 Figure 17 shows the phase cut angle detection for a double sided rectified mains voltage, typically used in 230 V applications and 120 V applications that sweep up the lamp voltage in the resonant tank. 470 kΩ 470 kΩ 100 kΩ BAS20 100 kΩ CSP/DCI 33 kΩ 1.0 μF 39 kΩ 470 nF 019aab965 Fig 17. Dimmer phase cut detector circuit for rectified mains supply Figure 18 is a typical circuit that is used for single side rectified 120 V voltage doubler applications. VDC CB1 33 nF 220 kΩ 220 kΩ 39 kΩ VMAINS 100 kΩ BAS20 100 kΩ CSP/DCI CB2 0 VMAINS + VDC / 2 100 kΩ 100 kΩ 1.0 μF 39 kΩ 470 nF BAS20 019aab975 Fig 18. Input voltage measurement in voltage doubler AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 15 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x 3.7 Supply of the IC and ACM for UBA2028 and UBA2014 Before start-up the supply electrolytic capacitor is charged via the start-up resistor to the DC bus voltage. The IC starts when the supply voltage exceeds VDD(start). Then the half-bridge starts to switch and the IC is supplied via the dV/dt capacitor that is connected to the half-bridge. The supply current of the IC and the charge required to drive larger MOSFETs determine the value of this capacitor. The IC needs more supply current which requires a larger dV/dt capacitor. On the other hand, if the dV/dt capacitor is too large, the half-bridge can be hard switching at the higher frequencies. In most cases, a value of 470 pF is a good compromise between these two mechanisms. A larger CdV/dt also has a positive influence on the EMI as it slows down the rise and fall times of the half-bridge switching pulses. A small capacitor is placed in parallel with the 12 V Zener diode, so it does not overload if CdV/dt is larger than required to supply energy to the IC. This capacitor also determines at which bus voltage level the IC stops oscillating. This to create a flicker free operation of the lamp by using the hysteresis on the VDD pin of the IC. A Zener diode clamps the voltage, which is supplied to the supply electrolytic capacitor via a diode. This must be a diode with a fast recovery. VBUS GH SH TH RSTARTUP resonant tank and lamp CdVdt GL VDD TL Coff 12 V RPCS RACM ACM PCS 019aab970 Fig 19. IC supply circuit The UBA2014 and UBA2028 have an adaptive non-overlap function and a capacitive mode detection that measure the charging current of the dV/dt capacitor by measuring the voltage across RACM. This voltage is supplied to the capacitive mode detection input. For more information on this function, consult the respective data sheet. 3.8 Preheating Before ignition, the tube resistance is infinite. However, the resonance inductor has two auxiliary windings that are used for heating the filaments, so the quality factor of the resonant tank is high, but not infinite. The half-bridge frequency starts at fmax and is swept down until the voltage on the PCS/SLS pin reaches the VPH level of 0.6 V. The frequency sweep stops. The current in the resonant tank is kept constant for the duration of the preheat time, which is defined with the value of the preheat capacitor connected to the CT/CP pin. AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 16 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x freq (kHz) fmax fprh fnom fmin t (ms) tpreheat 019aab969 Fig 20. Frequency versus time during start-up During the preheat time, the current is controlled by regulating the half-bridge frequency so that the voltage on the PCS/SLS pin stays constant. This means that the half-bridge current is kept constant. The value of this current can be adapted by changing the vale of the PCS/SLS resistor. Take care that the tube voltage does not become too high during preheat time. Generally, it must be less than half the ignition voltage. Tube voltage that is too high can cause an early uncontrolled ignition in the preheat phase. 10000 Vlamp(off) V 1000 lgnition Vlamp(on) 100 10 150 mA Ilamp 100 50 0 30 40 50 60 70 80 90 100 110 120 130 kHz 019aab981 Fig 21. Transfer of the resonant tank AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 17 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x After preheating, the frequency sweeps down and during this part of the sweep the voltage across the lamp rises further and the lamp ignites. The UBA2014 and UBA2028 always sweep down to fmin. The UBA2027 detects a lamp current and stops the sweep and immediately goes to fnom (see Figure 20). The tube is now a resistive element in parallel to the capacitor in the resonant tank. This changes the quality factor of the resonant tank. Figure 21 shows this situation (blue line (Vlamp(on))). After the frequency reaches fmin or the current in the tube has been detected, the lamp current control loop takes over the tube current control. When inductive preheat is used (the only option for the UBA2027), the resonance inductor is in fact a three winding coupled inductor. A series capacitor can be used for tuning the preheat current into constant power to the filament during preheat. A resistor and a capacitor in parallel to the resonance inductor can represent the preheat circuit. The coupling between the windings is assumed to be perfect. Itot CFIL 3 La 1 2 Lr La Vin Cr RFIL 4 5 6 Ifp Ifp = 2 × CF = 2 × La Lr La Lr × IFIL × CFIL Lr CFIL RF = 1 Lr × × RFIL 2 La RFIL Cr 019aab979 Fig 22. Preheat circuit The equation of interest is the relation between the half-bridge current (Itot) and the preheat current (Ifil), because the value of Itot during preheat is fixed with the value of the PCS/SLS resistor. The resulting value of Ifil can be calculated with Equation 13: 1 Lr 1 Lr --- -----i fp --- -----jL r I fil 2 La 2 La ------ = -------------- = ----------------------------------------I tot I tot 1 jL r + ----------- + R f jC f (13) Equation 13 leads to: 4 4 2 Lr Cf I fil 1 - ----------------------------------------------------------------------------------------------------------- = -2 2 2 4 2 2 2 2 2 I tot L a L r C f + R f C r – 2L r C r + 1 (14) This equation does not contain Cr. The total current that is drawn from the half-bridge is: 2 3 Cr Cf Rf + j Lr Cr Cf – Cr I tot = --------------------------------------------------------------------------------------------------------------------V I 2 2 3 Lr Cr + Lr Cf – 1 + j Lr Cr Rf Cf – Rf Cf AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 (15) © NXP B.V. 2012. All rights reserved. 18 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x The resistance of the filament is not constant. During preheating, the resistance of the filament becomes ideally 4.75 times higher than the cold resistance. Changing the value of the resistance in the equations above shows that the filament current is almost independent of the filament resistance. However, the filament current strongly depends on the frequency, and the voltage across the inductor is almost a square wave. This means that the preheat current contains numerous higher harmonics that are present in the half-bridge voltage. High dV/dt gives a high preheat current. So if the bridge is hard switching, then there is a large preheat current flowing through the filaments. It is more practical to simulate the preheat circuit in spice than to calculate it in a spreadsheet application. A simulation also shows the switching behavior of the half-bridge. During dimming, the lamp current decreases and subsequently the current through the filaments also decreases. The result is that the filaments become colder and the emission of electrons becomes more difficult. For that reason additional heating of the filaments is required, which the auxiliary windings provide. The way of connecting the auxiliary windings to the lamp is important: depending on the connection the additional current is added or subtracted from the lamp current that also flows through the filaments (see Ref. 2.). This mechanism can be used to adapt the heating current to the specification of the lamp. 3.9 Input filtering The half-bridge, and to a larger extent the charge pump, are drawing a high frequency current from the mains. Filter out the HF current because it causes EMI problems. Figure 23 shows a basic input filter. The inductor is blocking the high frequent current the charge pump circuit is drawing. The capacitor is providing a low ohmic path for this current. II (t) L VI (t) R C ICP 019aab962 Fig 23. Basic input filter The damping of the HF components is calculated using Equation 16 II 1 Damping = 20 log ---------------- = 20 log -------------------------------------------------------2 2 2 I CP 1 + LC + RC (16) Assuming that: • L = 10 mH • C = 47 nF • R = 10 For 50 Hz, the damping is 0 dB, for 50 kHz the damping is 66 dB. For CFL lamps that are not placed behind a dimmer, this circuit works well. However due to the voltage steps in the phase cut input voltage, there is ringing of the input current. When the input current drops below the hold current during this ringing (or even crosses AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 19 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x zero), then the Triac is turned off. After some time, the Triac is triggered again and switches off again. This process repeats until the load current is so large that there is no more zero crossing of the mains current. This phenomenon is called multiple triggering or firing and it is shown in Figure 24 VMAINS 100 V/div RMS 122.293 V IMAINS 0.5 A/div RMS 245.947 V 5 ms/div 019aab978 Fig 24. Mains voltage and mains current with multiple firing of the Triac The mains current when a voltage step Vstep is applied to the system is calculated as shown in Equation 17: and Equation 18 V step i t = -------------- e res L R – --------2 2L sin t + I CP I (17) with res = 2 1 R ------- – --------2- LC 4L (18) If the charge pump circuit current (ICP) rises very quickly after the voltage step and ICP is larger than the negative amplitude of the ringing, there is no zero crossing and hence no multiple firing. AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 20 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x I (mA) ICP IRING 3T/4 0 t (ms) T/4 019aab971 Fig 25. Input current ringing Example: • • • • • • • • Vstep = 40 V L = 10 mH C = 47 nF R = 10 res = 46126 fres = 7.3 kHz T = 136 s 3T/4 = 102 s The peak current at 3T/4 is: 40 2.16 10 3 0.95 = 82 mA. In order to prevent ringing, the charge pump must switch on immediately and draw a current larger than 82 mA after 102 m. If the charge pump circuit does not rise fast enough, some other measures can be taken to reduce multiple firing. One solution is to place an RC combination can be placed across the input. In that way, a current is added to II (see Figure 26). L II (t) R RI VI (t) RDC C ICP CI 019aab972 Fig 26. Input filter with AC load AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 21 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x The drawback of the circuit in Figure 26 is that the filter capacitor CI is very large, as it must be a high-voltage type. The resistor is large too because it must handle a power of 1 W. Furthermore, it adds some reactive power slightly decreasing the power factor. LFILTER VMAINS CFILTER CChargepump 019aab973 Fig 27. Input filter placed after rectifier bridge creates the Constant Input Current (CIC) topology In Figure 27, resistor RDC is in parallel to the filter capacitor. This resistor can be added to provide a DC path for charging the capacitor or supplying current to electronics inside a dimmer that fires the Triac. 4. Circuit diagrams The next three paragraphs show the circuit diagrams for the dimmable CFL applications for 230 V and 120 V mains voltage. These circuit diagrams were tested with one burner type and with different dimmers. So with a different burner or specific dimmers some components have to be adapted. In most cases, it means modifying the resonant tank and the charge pump capacitor. Section 5 shows a calculation example for the resonant tank. In this example, the lamp voltage is assumed to be constant during dimming. Take the lamp characteristics into account when more accurate results are required. AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 22 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x 4.1 Circuit diagram for the UBA2027 application D5 D7 D6 FS C3 CCB CB C9 L2 C4 R5 GHS C6 R1 D1 L1 C20 C1 C16 HBO D2 CFL C10 C2 C19 R11 UBA20270 GLS D4 D8 C7 R2 CI VDD D9 C11 D3 CCP CP R6 MDL RMDL C15 TR1 SLS C5 C12 D11 C14 RREF RREF C8 R9 C17 RSLS R3 CSI R7 RCSI CCF DCI R10 D10 R4 CF C13 R8 C18 PGND SGND 001aam672 Fig 28. UBA2027 circuit diagram AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 23 of 30 xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxx x x x xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xx xx xxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxx x x xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxx xxx NXP Semiconductors AN10803 Application note 4.2 Circuit diagram 230 V input for the UBA2014 application C12 3.3 nF 1000 V 230 V AC 50 Hz L_L L2 10 mH R1 fusistor 4.7 Ω 1W R3 560 Ω C1 220 nF 250 V L_N D1 1N4007 D2 1N4007 D5 D6 1N4937 1N4937 C9 100 nF 400 V R30 220 kΩ R31 220 kΩ C3 22 nF 400 V R2 100 kΩ C3 22 nF 400 V L1 1.5 mH C20 D4 1N4007 D3 1N4007 R20 T1 3N60 220 nF 1 kΩ C21 C21 220 nF VDD CT CSW 10 nF C23 100 pF CF R21 33 kΩ IREF VREF CS+ R41 220 kΩ CSD40 C31 2.2 μF 7 1 9 FVDD L10 10 GH 2 11 3 6 UBA2014 4 8 14 12 15 13 16 5 D30 1N4148 C25 100 nF SH GL T2 3N60 PCS C32 470 pF 400 V C33 1 nF 3 mH C11 L10A L10B 20 μH 20 μH C14 22 nF 68 nF C10 1.5 nF 1000 V 21 W D31 12 V ACM LVS GND R27 1Ω R34 2.7 Ω D41 D51 R51 C40 4.7 μF 4.3 V R42 12 kΩ C15 22 nF CFL 1N4148 R43 47 kΩ C41 4.7 μF 1 kΩ C51 1 nF R50 33 Ω C52 1N4935 D50 1 μF 1N4935 019aab959 Fig 29. 230 V circuit diagram AN10803 24 of 30 © NXP B.V. 2012. All rights reserved. Triac dimmable CFL using UBA2028/UBA2014/UBA2027x Rev. 2 — 27 February 2012 All information provided in this document is subject to legal disclaimers. R40 220 kΩ C30 100 nF xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxx x x x xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xx xx xxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx xxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxx x x xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxx xxx NXP Semiconductors AN10803 Application note 4.3 Circuit diagram 120 V input for the UBA2014 application C12 120 V AC 60 Hz D2 1N4937 L_L R1 fusistor 4.7 Ω 1W R3 560 Ω L_N C1 220 nF 250 V 3.3 nF 1000 V R2 100 kΩ C2 47 nF 400 V D3 1N4937 R31 220 kΩ C3 100 nF 400 V C7 47 μF 400 V L1 10 mH C8 47 μF 400 V D4 1N4937 C30 100 nF C20 R20 1 kΩ VDD C21 CT 220 nF C23 CF 100 pF R21 IREF 33 kΩ R40 68 kΩ VREF CS+ R41 68 kΩ D43 1N4148 D40 D41 4.3 V 1N4148 R42 22 kΩ R44 CS- R45 33 kΩ 33 kΩ R43 22 kΩ C41 4.7 μF 7 1 9 FVDD 11 3 6 UBA2014 4 8 14 12 15 13 16 5 D30 1N4148 L10 10 GH 2 T1 3N60 C25 100 nF SH GL T2 3N60 PCS C32 470 pF 400 V C33 1 nF 3 mH C11 68 nF L10A L10B 20 μH 20 μH C14 22 nF C10 1.5 nF 1000 V 21 W D31 12 V ACM LVS GND R27 1Ω R34 2.7 Ω D51 R51 C42 4.7 μF C15 22 nF CFL 1 kΩ C51 1 nF R50 33 Ω C52 1N4935 D50 1 μF 1N4935 019aab958 Fig 30. 120 V circuit diagram AN10803 25 of 30 © NXP B.V. 2012. All rights reserved. Triac dimmable CFL using UBA2028/UBA2014/UBA2027x Rev. 2 — 27 February 2012 All information provided in this document is subject to legal disclaimers. CSW 10 nF C45 100 nF 400 V C31 2.2 μF 220 nF C21 C40 100 nF C9 100 nF 400 V R30 220 kΩ D1 1N4937 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x 5. Resonant tank calculation example Typical lamp power versus frequency calculations as used in a mathematical design application are as follows: Input data: Vbus = 300, fop = 45 103, Ilamp_nom = 140 10-3, Vlamp_nom = 130, = 35 1 4 V bus V 1_RMS = ------- --- ---------2 2 (19) V1_RMS = 135.047, = --------- , = 0.611 , op = 2 fop, op = 2.827 105 180 V lamp_nom Plamp = Ilamp_nom Vlamp_nom, Plamp = 18.2, R eq = ----------------------- , Req = 928.571 I lamp_nom 2 2 2 2 – V 1_RMS + V lamp_nom + V lamp_nom tan C = --------------------------------------------------------------------------------------------------------------------- , C = 2.351 10-9 V 1_RMS R eq op tan + op R eq C L = ---------------------------------------------- , L = 3.133 10-3, op 2 3 --------- + op R eq C R eq 1 1 f res = ------ ------2 LC , fres = 58.64 103 The lamp voltage is assumed to be constant: Vla = 130, Ila = 0.001, 0.011. 0.3 1 f I la = ------------------------------------2 2 L C V la 2 2 2 3 2 2 4 4 2 2 2 2 L C V la – L I la + – 4 L C I la V la + L I la + 4 L C V la V l_RMS 2 P la I la = V la I la AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 26 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x 019aab977 25 P1a (W) 20 P1a I1a 15 10 5 0 40 50 60 70 80 90 100 f I1a (kHz) Fig 31. Lamp power versus frequency graph 6. Abbreviations Table 1. Abbreviations Acronym Description CFL Compact Fluorescent Lamp CMP Capacitive Mode Protection EMI ElectroMagnetic Interference MDL Minimum Dimming Level PFC Power Factor Correction RMS Root Mean Square ZVS Zero Voltage Switching 7. References AN10803 Application note [1] Charge Pump Power-Factor-Correction Dimming Electronic Ballast Jinrong Qian, Fred C. Lee, Tokushi Yamauchi. IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL.14, NO. 3, MAY 1999 [2] Standardized data for dimming of fluorescent lamps L.H. Goud, J.W.F. Dorleijn IEEE-IAS Conference 2002 [3] Application note: AN10962 All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 27 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x 8. Legal information 8.1 Definitions Draft — The document is a draft version only. The content is still under internal review and subject to formal approval, which may result in modifications or additions. NXP Semiconductors does not give any representations or warranties as to the accuracy or completeness of information included herein and shall have no liability for the consequences of use of such information. 8.2 Disclaimers Limited warranty and liability — Information in this document is believed to be accurate and reliable. However, NXP Semiconductors does not give any representations or warranties, expressed or implied, as to the accuracy or completeness of such information and shall have no liability for the consequences of use of such information. NXP Semiconductors takes no responsibility for the content in this document if provided by an information source outside of NXP Semiconductors. In no event shall NXP Semiconductors be liable for any indirect, incidental, punitive, special or consequential damages (including - without limitation - lost profits, lost savings, business interruption, costs related to the removal or replacement of any products or rework charges) whether or not such damages are based on tort (including negligence), warranty, breach of contract or any other legal theory. Notwithstanding any damages that customer might incur for any reason whatsoever, NXP Semiconductors’ aggregate and cumulative liability towards customer for the products described herein shall be limited in accordance with the Terms and conditions of commercial sale of NXP Semiconductors. Right to make changes — NXP Semiconductors reserves the right to make changes to information published in this document, including without limitation specifications and product descriptions, at any time and without notice. This document supersedes and replaces all information supplied prior to the publication hereof. Suitability for use — NXP Semiconductors products are not designed, authorized or warranted to be suitable for use in life support, life-critical or safety-critical systems or equipment, nor in applications where failure or malfunction of an NXP Semiconductors product can reasonably be expected to result in personal injury, death or severe property or environmental damage. NXP Semiconductors and its suppliers accept no liability for inclusion and/or use of NXP Semiconductors products in such equipment or applications and therefore such inclusion and/or use is at the customer’s own risk. Applications — Applications that are described herein for any of these products are for illustrative purposes only. NXP Semiconductors makes no representation or warranty that such applications will be suitable for the specified use without further testing or modification. AN10803 Application note Customers are responsible for the design and operation of their applications and products using NXP Semiconductors products, and NXP Semiconductors accepts no liability for any assistance with applications or customer product design. It is customer’s sole responsibility to determine whether the NXP Semiconductors product is suitable and fit for the customer’s applications and products planned, as well as for the planned application and use of customer’s third party customer(s). Customers should provide appropriate design and operating safeguards to minimize the risks associated with their applications and products. NXP Semiconductors does not accept any liability related to any default, damage, costs or problem which is based on any weakness or default in the customer’s applications or products, or the application or use by customer’s third party customer(s). Customer is responsible for doing all necessary testing for the customer’s applications and products using NXP Semiconductors products in order to avoid a default of the applications and the products or of the application or use by customer’s third party customer(s). NXP does not accept any liability in this respect. Export control — This document as well as the item(s) described herein may be subject to export control regulations. Export might require a prior authorization from competent authorities. Evaluation products — This product is provided on an “as is” and “with all faults” basis for evaluation purposes only. NXP Semiconductors, its affiliates and their suppliers expressly disclaim all warranties, whether express, implied or statutory, including but not limited to the implied warranties of non-infringement, merchantability and fitness for a particular purpose. The entire risk as to the quality, or arising out of the use or performance, of this product remains with customer. In no event shall NXP Semiconductors, its affiliates or their suppliers be liable to customer for any special, indirect, consequential, punitive or incidental damages (including without limitation damages for loss of business, business interruption, loss of use, loss of data or information, and the like) arising out the use of or inability to use the product, whether or not based on tort (including negligence), strict liability, breach of contract, breach of warranty or any other theory, even if advised of the possibility of such damages. Notwithstanding any damages that customer might incur for any reason whatsoever (including without limitation, all damages referenced above and all direct or general damages), the entire liability of NXP Semiconductors, its affiliates and their suppliers and customer’s exclusive remedy for all of the foregoing shall be limited to actual damages incurred by customer based on reasonable reliance up to the greater of the amount actually paid by customer for the product or five dollars (US$5.00). The foregoing limitations, exclusions and disclaimers shall apply to the maximum extent permitted by applicable law, even if any remedy fails of its essential purpose. 8.3 Trademarks Notice: All referenced brands, product names, service names and trademarks are the property of their respective owners. All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 28 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x 9. Figures Fig 1. Fig 2. Fig 3. Fig 4. Fig 5. Fig 6. Fig 7. Fig 8. Fig 9. Fig 10. Fig 11. Fig 12. Fig 13. Fig 14. Fig 15. Fig 16. Fig 17. Fig 18. Fig 19. Fig 20. Fig 21. Fig 22. Fig 23. Fig 24. Fig 25. Fig 26. Fig 27. Fig 28. Fig 29. Fig 30. Fig 31. Phase cut sine wave . . . . . . . . . . . . . . . . . . . . . . .3 Basic 230 V dimmer. . . . . . . . . . . . . . . . . . . . . . . .4 Typical Triac VI characteristic curves. . . . . . . . . . .4 Basic dimmer with CFL . . . . . . . . . . . . . . . . . . . . .5 Input voltage and current when large dimmer phase cut is set . . . . . . . . . . . . . . . . . . . . . . . . . . .6 Block diagram of the charge pump CFL dimmer . .7 Transfer of resonant tank versus frequency before and after ignition . . . . . . . . . . . . . . . . . . . . .8 Schematic resonant tank and vector diagram . . . .8 CFL inverter with charge pump power feedback .10 Equivalent circuit of a voltage source charge pump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 VSCP with ICF . . . . . . . . . . . . . . . . . . . . . . . . . . .12 VSCP with LFSR . . . . . . . . . . . . . . . . . . . . . . . . .12 Voltage doubler circuit . . . . . . . . . . . . . . . . . . . . .12 Voltage doubler with charge pump circuit . . . . . .13 Current feedback control loop . . . . . . . . . . . . . . .13 Dimming curve UBA2028 versus UBA2027. . . . .15 Dimmer phase cut detector circuit for rectified mains supply . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 Input voltage measurement in voltage doubler . .15 IC supply circuit . . . . . . . . . . . . . . . . . . . . . . . . . .16 Frequency versus time during start-up . . . . . . . .17 Transfer of the resonant tank. . . . . . . . . . . . . . . .17 Preheat circuit . . . . . . . . . . . . . . . . . . . . . . . . . . .18 Basic input filter . . . . . . . . . . . . . . . . . . . . . . . . . .19 Mains voltage and mains current with multiple firing of the Triac . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20 Input current ringing . . . . . . . . . . . . . . . . . . . . . . .21 Input filter with AC load . . . . . . . . . . . . . . . . . . . .21 Input filter placed after rectifier bridge creates the Constant Input Current (CIC) topology . . . . .22 UBA2027 circuit diagram . . . . . . . . . . . . . . . . . . .23 230 V circuit diagram . . . . . . . . . . . . . . . . . . . . . .24 120 V circuit diagram . . . . . . . . . . . . . . . . . . . . . .25 Lamp power versus frequency graph. . . . . . . . . .27 AN10803 Application note All information provided in this document is subject to legal disclaimers. Rev. 2 — 27 February 2012 © NXP B.V. 2012. All rights reserved. 29 of 30 AN10803 NXP Semiconductors Triac dimmable CFL using UBA2028/UBA2014/UBA2027x 10. Contents 1 2 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 4.1 4.2 4.3 5 6 7 8 8.1 8.2 8.3 9 10 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Triac dimmers . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Designing the application . . . . . . . . . . . . . . . . . 7 Block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Resonant tank. . . . . . . . . . . . . . . . . . . . . . . . . . 7 Charge pump circuit . . . . . . . . . . . . . . . . . . . . 10 Voltage doubler and charge pump for 120 V input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Current feedback control loop . . . . . . . . . . . . 13 Dimmer phase cut detector . . . . . . . . . . . . . . 14 Supply of the IC and ACM for UBA2028 and UBA2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Preheating . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Input filtering . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Circuit diagrams . . . . . . . . . . . . . . . . . . . . . . . 22 Circuit diagram for the UBA2027 application . 23 Circuit diagram 230 V input for the UBA2014 application . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Circuit diagram 120 V input for the UBA2014 application . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Resonant tank calculation example . . . . . . . . 26 Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . 27 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Legal information. . . . . . . . . . . . . . . . . . . . . . . 28 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Disclaimers . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Trademarks. . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Please be aware that important notices concerning this document and the product(s) described herein, have been included in section ‘Legal information’. © NXP B.V. 2012. All rights reserved. For more information, please visit: http://www.nxp.com For sales office addresses, please send an email to: [email protected] Date of release: 27 February 2012 Document identifier: AN10803