Analysis of TFT-OLED pixel unit and driving circuit

1 Introduction

Organic electroluminescent devices (OLEDs) are all-solid devices that convert electrical energy directly into light energy. They have attracted great attention due to their advantages such as thinness and lightness, high contrast, fast response, wide viewing angle, and wide operating temperature range, and are considered to be a new generation of display devices. To truly realize its large-scale industrialization, it is necessary to improve the luminous efficiency and stability of the device and design an effective image display driving circuit. Recently, with the deepening of research, the luminous efficiency and stability of OLEDs have reached the requirements of certain applications, but its dedicated driving circuit technology is not yet very mature. At present, all flat panel display drivers use matrix driving mode, which is a matrix display screen composed of X and Y electrodes. According to whether switching components are introduced in each pixel or not, matrix display is divided into active matrix (AM) display and passive matrix (PM) display.

PM-OLED has the advantages of simple structure and low cost, and is mainly used in simple displays with low information content; AM-OLED is dominant in displays with large amounts of information, and generally uses amorphous silicon TFT (a-SiTFT) or polycrystalline silicon (poly-SiTFT) switching components. The input signal is stored on the storage capacitor, so that the pixel remains in the gated state during the frame period, so transient high brightness is not required, which overcomes the shortcomings of PM-OLED and is not limited by the duty cycle. Therefore, in order to achieve high-quality display, OLED must adopt an active matrix drive method. Starting from the TFT-OLED active matrix pixel unit circuit, this article focuses on the analysis of voltage-controlled and current-controlled pixel unit circuits, and briefly discusses the influence of control/drive IC on the TFT-OLED active drive circuit.

2 Analog pixel unit circuit

AM-OLED drive implementation schemes include analog and digital. In the digital drive scheme, each pixel is connected to a switch, and the TFT is only used as an analog switch. The grayscale generation method includes time ratio grayscale and area ratio grayscale, or a combination of the two. At present, analog pixel circuits are still the mainstream, but in the realization of grayscale, the combination of analog technology with time ratio grayscale and area ratio grayscale theory will be a development trend in the future. In the analog scheme, according to the type of input data signal, the unit pixel circuit can be divided into voltage control type and current control type.

2.1 Voltage-controlled pixel circuit

2.1.1 Two-tube TFT structure

The voltage-controlled unit pixel circuit uses the data voltage as the video signal. The simplest voltage-controlled two-tube TFT unit pixel circuit is shown in Figure 1.

Figure 1 Two-tube TFT drive circuit

Its working principle is as follows: when the scan line is selected, the switch tube T1 is turned on, and the data voltage charges the storage capacitor CS through the T1 tube. The voltage of CS controls the drain current of the driving tube T2; when the scan line is not selected, T1 is turned off, and the charge stored on CS continues to maintain the gate voltage of T2, and T2 remains in the on state. Therefore, during the entire frame period, the OLED is in constant current control.

Among them, (a) and (b) are called constant current source structure and source follower structure respectively. In the former, OLED is at the drain end of driving tube T2, which overcomes the influence of the change of OLED turn-on voltage on T2 tube current; the latter is easier to implement in terms of process. The disadvantage of the two-tube circuit structure is that the inconsistency of the threshold voltage of driving tube T2 will lead to uneven brightness of each display screen, and the current and data voltage of OLED are nonlinear, which is not conducive to grayscale adjustment.

2.1.2 Three-tube TFT structure

The voltage-controlled pixel unit circuit based on the second-generation current conveyor principle is shown in FIG2 . The left side of the dotted line can be regarded as an external driving circuit, and the right side is a unit pixel circuit.

Figure 2 Pixel circuit based on the second-generation current conveyor principle

In control mode, T2 and T3 are turned on, and T1 and the operational amplifier form a second-generation current conveyor. Since the amplification factor of the operational amplifier can be very large, the threshold voltage of T1 tube becomes insensitive to the current. At this time, the current flowing through T1 is:

IT1=Vin/Rin

And the source voltage of T1 tube should be lower than the turn-on voltage of OLED to prevent OLED from turning on. In the hold mode, T2 and T3 are turned off, the storage capacitor Cs maintains the gate voltage of T1 tube, and the current enters OLED through T1. The amplifier is implemented by COMS circuit, and all pixels in the same line can share one operational amplifier.

The simulation results show that although the T3 tube has charge injection and clock leakage effects, which makes the OLED current slightly smaller than the control current; when the nominal current of the OLED is 1μA and the threshold voltage drift exceeds 5V, the relative errors of the control current and OLED current are -0.18% and 5.2% respectively, which successfully compensates for the spatial unevenness and instability of the TFT.

2.1.3 Four-tube TFT structure

Dawson et al. first proposed a unit pixel circuit with a four-tube TFT structure. This circuit compares the data signal with the drive tube through automatic zeroing to eliminate the offset of the TFT gate voltage, and applies a priority zeroing signal (VAZB) before the data signal to release the charge accumulated in the parasitic capacitor, solving the problem of threshold voltage variation and not relying on the opening and charging time of the OLED. The defect of this circuit is that when the channel length becomes shorter, uneven light emission will occur.

Goh JC et al. proposed a voltage-controlled circuit that uses subthreshold current to compensate for threshold voltage changes. A compensation stage is added to the driving timing to make the driver work in the subthreshold region. At this time, the gate-source voltage of the driver, that is, the threshold voltage Vth, is stored in the storage capacitor. This voltage can compensate for the drift of the TFT threshold voltage during the data input stage. They also proposed a voltage-controlled driving circuit that uses discharge to compensate for threshold voltage changes. Unlike the former, this circuit uses discharge to make the driver enter the subthreshold region, obtain the superposition value of the data voltage and the threshold voltage, and effectively compensate for the threshold voltage change.

In addition to being able to effectively compensate for changes in threshold voltage, the voltage-controlled drive circuit also has the advantage of having a fast response characteristic. Because the voltage is directly applied to both ends of the storage capacitor CS, there will be a momentary large current at the beginning of the charging current to charge the capacitor, which greatly reduces the charging time.

2.2 Current-controlled pixel circuit

Although the voltage-controlled circuit has the characteristics of fast response speed, it cannot accurately adjust the gray scale of the display and is difficult to meet the display requirements, so people have proposed a current drive scheme. The current-controlled unit pixel circuit uses the data current as the video signal.

Generally speaking, current-controlled pixel circuits need to meet the following requirements:

1) Effectively compensate for the drift of threshold voltage,

2) It has good current following characteristics and good linearity.

3) The response speed is within an acceptable range,

4) Under permitted conditions, try to reduce the driving power supply voltage to reduce power consumption.

Therefore, most current-controlled pixel circuits receive input current signals and map them to the output terminal, while storing them in the storage capacitor inside the pixel to ensure stable output throughout the frame. The current-driven circuits reported so far mainly include three-tube TFT structure, four-tube TFT structure, five-tube or even more-tube TFT structure.

2.2.1 Three-tube TFT structure

Figure 3 shows a three-tube TFT current control circuit, which works in two stages: control and hold. In the control stage, the scan line is at a high level, T2 and T3 are turned on, a low level is applied to the drain of T1, the OLED is reverse biased, the input data current flows through T2, T1, and the gate-source voltage of T1 is stored in Cs. In the hold stage, the scan line is at a low level, T2 and T3 are turned off, and a high level is applied to the drain of T1. The current flows through T1 and OLED, and the gate-source voltage of T1 maintains the current of T1 unchanged. The circuit can effectively compensate for the change of threshold voltage. After working for 700 hours, the current decays by 11%, which can be improved by reducing the overlap capacitance of TFT.

Figure 3 Current-controlled 3-TFT pixel circuit

2.2.2 Four-tube TFT structure

The 4-TFT current control driving circuit with threshold voltage compensation reported earlier abroad is shown in Figure 4. In the addressing stage, the scanning voltage turns on T1 and T3, and the data current Idata flows through T4 into the light-emitting unit, and the gate-source voltage of T4 is stored in Cs; after the addressing is completed, T1 and T3 are turned off, and the introduction of VG can turn on T2. ​​At this time, T4 is connected to VDD as a current source, and it is only controlled by the voltage stored in Cs, which eliminates the influence of the threshold voltage change. However, the introduction of the VG line affects the aperture ratio of the display.

Figure 4 Current control analog drive circuit with threshold voltage compensation

Figure 5 Current-controlled current mirror pixel circuit

The current-controlled pixel unit circuit based on current mirroring is widely used. The working principle of this type of circuit is explained below with the structure shown in Figure 5. When the voltage on the scan line is at a high level, this pixel is selected, transistors T1 and T2 are turned on, and Idata first charges the capacitor Cs from the data line through the T1 tube. When the voltage across the capacitor Cs reaches a certain value, the entire Idata flows to the T3 tube through the T2 tube. At the same time, since the gate voltages of the T3 tube and the T4 tube are equal, the data current Idata is mirrored as the current flowing through the OLED. When this pixel is not selected, the gate voltage of the T4 tube is determined by the voltage stored across the capacitor Cs, maintaining the current driving the OLED.

The study found that the aging of the switch tube T2, the drift difference of the threshold voltage VT of T3 and T4, and the different initial values ​​of the threshold voltage VT of T3 and T4 are the main mechanisms that affect the driving current stability of the a-Si:H circuit based on the current mirror. Therefore, the basic requirements for the current mirror to accurately realize the current following function are that T2 has as low resistance as possible in the on state and low leakage current in the off state; the initial threshold voltages of T3 and T4 are equal and change consistently; T3 and T4 work in the saturation region. Guo Bin et al. simulated and analyzed the JV characteristics of the poly-SiTFT/OLED coupling pair as a pixel unit of the current-controlled polysilicon thin film transistor (poly-SiTFT) active matrix organic light-emitting diode (AM-LOED) and the IV characteristics of the poly-SiTFT current mirror. The results show that the driving voltage of the poly-SiTFT/OLED coupling pair is low, not exceeding 8V at 200A/m2; while the TFT current mirror has a good following ability, with a saturation voltage of only 1.5~2.5V at 0.0~2.5μA. Generally speaking, circuits based on current mirrors have good compensation characteristics, and current-controlled drive circuits similar to this type can also prove this point well. Experiments have shown that this circuit has good linear output and can accurately adjust the grayscale of the display.

The defect of the four-tube current-driven circuit is that the charging time is long and the signal delay is serious when the display is low-brightness. At present, the charging time between the data line and the pixel is reduced mainly by adjusting the reduction ratio of the OLED current and the input data current. There are two types of methods reported, one is based on the TFT geometric size, and the other is based on the storage capacitor size. The voltage-dividing current-controlled driving circuit belongs to the former. The relationship between the current flowing through the OLED and the data current in the circuit is:

Here μ is the field effect mobility, Cox is the insulating layer capacitance per unit area; W and L are the MOS tube channel width and length respectively. From the above relationship, it can be seen that using large data current charging can obtain a small IOLED and reduce the charging time, but this is at the expense of increased power consumption. The current control type circuit with a series storage capacitor structure belongs to the latter. In the gating stage, Idata=IOLED, and in the non-gating stage, the relationship between the current flowing through the OLED in the circuit and the data current is Idata=RSCALEIOLED, where RSCALE is the current reduction ratio, which is related to the storage capacitor CST2, the switch tube gate source/gate drain equivalent overlap capacitance COV-T2, and the change in the amplitude of the scanning signal when gating and non-gating △VSCAN, and as the above parameters increase, RSCALE increases accordingly. Compared with the former, the advantage of this circuit is that through the appropriate combination of RSCALE and IOLED, it can not only reduce the response time to a greater extent, but also meet the display needs of high and low gray levels without increasing power consumption.

2.2.3 Five-tube TFT structure

B. Mazhari et al. proposed a five-tube unit pixel circuit, which uses a TFT with a gate-source short circuit as a negative feedback resistor, effectively suppressing the kink effect of polysilicon TFT, achieving a data current of up to 20A, and the output characteristic curve still has good linearity, overcoming the defects of various previous circuits with a small current range under the premise of ensuring linearity. Epson-Cambridge Laboratory proposed an advanced self-adjusting voltage source technology, which is also a five-tube driving scheme. The circuit stores the source voltage of the driving tube TFT through a unit gain amplifier to ensure that the bias conditions of the driving tube are consistent in the gated and non-gated stages.

Although the current range is limited to 0.2A ~ 1A, it still effectively improves the disadvantage that the change of threshold voltage has a greater impact on the OLED current when the data current is small, but the circuit structure is complex and limits the duty cycle of the pixel.

3 Drive system

A complete active matrix OLED drive display system, in addition to the matrix display screen composed of pixel unit circuits, also includes a driver IC (row and column control/drive circuit), a single-chip microcomputer control circuit, etc. The typical block diagram of the OLED active drive system is shown in Figure 6.

Figure 6 Typical block diagram of OLED active drive system

The image data for display is stored in ROM or RAM. The CPU or MCU control circuit generates a general control signal. Under the general control signal, the row control circuit and the column drive circuit combine their internal functions to generate basic row signals and basic column signals. Under the general control signal, the basic row signal and the basic column signal, the row drive circuit and the column drive circuit combine their internal functions to generate row scan signals and column data signals, so that the OLED display screen displays the image information stored in ROM or RAM.

The driver IC is placed between the control circuit and the active glass plate, and is the core of the entire driver circuit. Many companies around the world have been engaged in the research of OLED driver ICs. So far, there is no fully commercialized AM-OLED driver IC. However, NextSierra has launched the integrated TFT-OLED row and column drivers NXS1008, NXS1009 and control chip NXS1010. Zhang Zhiwei and others used this series of chips to drive the 240×320×3 dot matrix TFT-OLED screen through the control of the MCS-51 microcontroller, realizing dynamic graphic display of large amounts of information.

Since the supporting driver chips of liquid crystal display devices have relatively complete functions and are inexpensive, transferring such chips to active matrix display screens (AM-OLED) has become the current research focus at home and abroad.

In order for the LCD driver chip to drive TFT-OLED, the key lies in two points:

1) The driving timing and display data of the liquid crystal driver chip meet the requirements of the TFT-OLED display;

2) The driving capability of the LCD chip driver meets the chip requirements of the OLED display.

The circuit of the display drive system is currently the weak link for TFT-OLED to display effectively. From the perspective of driver IC/control circuit, the most urgent task is to develop a general or dedicated driver IC and integrate the control circuit, which will greatly enhance the competitiveness of OLED in the flat panel display market.

4 Conclusion

For large-screen, high-resolution OLED flat panel display devices, active drive circuits have become an indispensable part. This article focuses on analyzing the working principles, advantages and disadvantages of voltage-controlled and current-controlled pixel unit circuits composed of different numbers of TFT tubes, and discusses the influence of control/drive IC on TFT-OLED active drive circuits, which can provide some basis for the design of OLED flat panel displays.