Page 1 of 3

A Cold Cathode Fluorescent Lamp (CCFL) Controller Used in Magnetic Transformer Application
Weiyun (Sophie) Chen

Download article in .pdf format (330 KB)

A cold cathode fluorescent lamp (CCFL) controller is designed for a flyback inverter with a piezoelectric transformer. But it can be easily used in magnetic transformer applications. This article describes the operational principle and design method of the UCC3975 CCFL controller with a magnetic transformer as a CCFL driver. The DC-characteristics of the series-parallel resonant converter are discussed in detail. The analog dimming of lamp and frequency range programming are also addressed. A complete design example also is presented.

INTRODUCTION

Portable devices, such as laptops, notebook computers, and personal digital assistants (PDA) are developed rapidly nowadays, which places an ever-increasing demand on display technology. The LCD with cold cathode fluorescent back lighting satisfies the requirements on display performance, size, and efficiency. The cold cathode fluorescent lamp (CCFL) requires two to three times the operating voltage to strike. A sinusoidal waveform is preferred to minimize RF interference and maximize the lamp efficiency. The resonant inverter shows exactly the characteristics that are needed in this application. Conventionally, a buck converter followed by a push-pull inverter (Royer circuit) is used in this application. The dc-to-dc buck converter regulates the lamp intensity and input voltage variation. The push-pull inverter transfers the dc voltage to a sinusoidal waveform to drive the lamp. The benefit of this two-stage approach is that it's easy to optimize each stage. The disadvantage is that the energy is processed twice, so more components are needed.

Figure 1. System implementation with the UCC3975 controller.

Figure 1 shows the single-stage CCFL driver with UCC3975 controller. UCC3975 is an 8-pin flyback piezoelectric transformer controller. It can be easily used as a CCFL controller for magnetic transformer applications. As shown in Figure 1, few components are used here. 50% duty cycle drive signals are generated at pin OUTP and OUTN to drive two external MOSFETs. To avoid both MOSFETs being on at the same time, a certain dead time is added between two drive signals. A square wave with almost 50% duty cycle is generated on the primary side of the transformer, which will be filtered by the LCC into a sinusoidal waveform to drive a CCFL. By varying the switching frequency, the output voltage can be controlled to compensate for both line and load variations. Regulating the voltage applied to pin FB through a resistor controls the lamp density.

OPERATIONAL PRINCIPLE

Figure 2 shows the schematic and typical waveforms of the power stage, the series-parallel resonant forward inverter. The inverter is composed of two bi-directional two-quadrant switches SP and SN and a resonant circuit L-Cs-Cp-R, where L is the resonant inductor; Cs and Cp are the series and parallel capacitors; and R is the load. Switches SP and SN are alternately turned on and off at the switching frequency with a duty cycle of 50%. A certain voltage whose amplitude is equal to input voltage is built up in the capacitor Creset. A square wave voltage with Vin amplitude is applied to the transformer primary side and is filtered by L-Cs-Cp-R a third order low-pass filter. So a sinusoidal voltage waveform is generated at R, which is the expected sinusoidal lamp voltage.
The detailed operation principle is discussed here and the analysis is based on the following assumptions:

1. Switching frequency is higher than resonant frequency, which is defined as the frequency at which phase shift is equal to zero. In this case, the inductor currents i_L1 and i_L2 lag behind the fundamental component of the primary voltage V_pri ;
2. The switches turn-on and turn-off instantly. So the switching losses are ignored;
3. The RDS_ON of the MOSFET is very small and voltage drop is very low compared to the voltage drop across the body diode. Most of the current flows through the MOSFET, so the current through the body diode is ignored.
4. The current through the resonant inductor is nearly sinusoidal.

(a) Schematic



(b) Typical waveforms

Figure 2. Schematic and typical waveforms.

[T1-T2]: At time T1, SN turns off and SP turns on. The primary current continues flowing in the same direction through L1. During this period, the current flows through SP and the reset capacitor Creset and the energy stored in the inductors charges the capacitor. The voltage across the capacitor increases. The primary current decreases according to the same rule as in period
[T0-T1]. At T2, the primary current decreases to zero and the negative half cycle starts. The period from T0 to T2 is determined by the time constant of the resonant tank and equals half of the resonant period
[T2-T3]: From T2, the primary current changes direction and flows through SP. During this period, the current discharges the capacitor Creset. The output power is supported by the stored energy in the reset capacitor, which was transferred from the magnetizing and leakage inductance during period [T1-T2].
[T3-T4]: At time T3, the GS signal of SP becomes high, SP turns off and primary current switches to SN. When the current reaches zero, this cycle is over and a new cycle starts. The operation repeats as described before.
When SP is on, the primary current flows through reset capacitor Creset and charges and discharges the capacitor. In order to keep the reset capacitor voltage constant, the charge and discharge energy should be the same over one cycle. This means that the two shaded areas should be the same. So a dc offset is created to compensate for any difference. A small air gap is needed in the magnetic core to store the energy, which is different from the traditional forward transformer.