Display with redundant light emitting devices

Abstract

An active matrix display where in one embodiment each cell comprises: a driving circuit for providing current to light emitting devices placed in the cell under the control of a data driver signal, a first light emitting device location connected to the driving circuit and a second light emitting device location connected in series to the first light emitting device location. A first thin-film transistor (TFT) is connected in parallel with the first light emitting device location and a second TFT is connected in parallel with the second light emitting device location, its gate node connected to the gate node of the first TFT. One terminal of a third TFT is connected to the gate nodes of the first and second TFTs and selectively connects a control signal to the first and second TFTs under the control of a scan driver signal. The control signal determines which of a first or second light emitting device placed in the cell emits light when the driving circuit provides current.

Claims

1. A display manufactured by a process, comprising: placing a plurality of first light emitting devices in a plurality of cells of the display; testing the display to determine whether one or more of the plurality of first light emitting devices in the plurality of cells are defective; placing a second light emitting device in each of the plurality of cells determined to contain a defective first light emitting device of the plurality of first light emitting devices; programming the display to emit light by the second light emitting device in each of the plurality of cells determined to contain the defective first light emitting device; subsequent to placing the second light emitting device, testing the display to determine whether the placed second light emitting device is defective; and placing a third light emitting device in a cell determined to contain the defective second light emitting device.

2. The display of claim 1, wherein the process further comprises: subsequent to placing the third light emitting device, testing the display to determine whether the placed third light emitting device is defective; and programming the display to disable the cell containing the third light emitting device determined to be defective.

3. The display of claim 1, wherein testing the display includes providing control signals of opposite polarity to the plurality of first light emitting devices and the second light emitting device.

4. The display of claim 1, wherein the process further comprises programming the display to emit light by the third light emitting device in the cell determined to contain the defective second light emitting device.

5. The display of claim 1, wherein: the cell operates in a plurality of modes; in a first mode of the plurality of modes, each of the plurality of first light emitting devices is activated; in a second mode of the plurality of modes, the second light emitting device is activated; in a third mode of the plurality of modes, the third light emitting device is activated; and in a fourth mode of the plurality of modes, at least one of the plurality of first light emitting devices, the second light emitting device, and the third light emitting device is deactivated.

6. The display of claim 5, wherein in the fourth mode, at least one of the plurality of first light emitting devices, the second light emitting device, and the third light emitting device is activated.

7. The display of claim 1, wherein the process further comprises: subsequent to placing the third light emitting device, testing the display to determine whether the placed third light emitting device is defective; and programming the display to disable the cell containing the third light emitting device determined to be defective.

8. The display of claim 1, wherein a first light emitting device of the plurality of first light emitting devices and the second light emitting device in each of the plurality of cells are connected in series.

9. The display of claim 1, wherein a first light emitting device of the plurality of first light emitting devices and the second light emitting device in each of the plurality of cells are connected in parallel.

10. The display of claim 1, wherein the process further comprises programming the display to disable a cell containing the defective second light emitting device.

11. The display of claim 1, wherein the plurality of first light emitting devices and the second light emitting device are inorganic light emitting devices.

12. The display of claim 1, wherein the process further comprises programming the display to emit light by a first light emitting device of the plurality of first light emitting devices in a cell other than the plurality cells determined to contain the defective first light emitting device.

13. The display of claim 1, wherein the second light emitting device is a known good light emitting device.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Embodiments of the invention will now be described, by way of example, with reference to the accompanying drawings, in which:

(2) FIG. 1 shows a conventional type active matrix display;

(3) FIG. 2 shows a placement method for light emitting devices within a display according to an embodiment of the present invention;

(4) FIG. 3 shows a cell design for an active matrix display according to an embodiment of the present invention;

(5) FIG. 4 shows a cell design with three light emitting devices connected in series and controlled by n-type TFTs according to another embodiment of the present invention;

(6) FIG. 5 shows a cell design with three light emitting devices connected in series and controlled by CMOS TFTs according to a further embodiment of the present invention;

(7) FIG. 6 shows a cell design with three light emitting devices connected in parallel according to a still further embodiment of the present invention;

(8) FIG. 7 shows a pixel design with three differently colored light emitting devices connected in series according to an embodiment of the present invention; and

(9) FIG. 8 shows three pixels such as shown in FIG. 7 connected in series according to a still further embodiment of the present invention.

DESCRIPTION OF THE EMBODIMENTS

(10) Referring now to FIG. 2, when light emitting devices such as the iLEDs referred to above are initially placed on a substrate, their yield is expected to be between 90-95% which is lower than the acceptable yield for a high performance display. This means that cells containing these devices need to be repaired if the display is to be shipped. One method for doing so comprises placing a second light emitting device in a cell containing a defective first device and then using laser to open or melt fuses to disconnect the first device from the cell driving circuit to which the second device is connected. The same procedure could be used if a third or fourth device were to be placed on the cell. As will be appreciated this “physical” solution has limited applicability (5-10 pixels per display), requires special equipment and increases the process time.

(11) In order to overcome these problems, embodiments of the present invention provide a cell comprising a plurality of locations in which light emitting devices can be placed. Examples of cell design will be described later, but we would first of all begin by describing a placement method for a display based on a cell design capable of receiving three light emitting devices.

(12) Initially, discrete light emitting devices such as iLEDs referred to above are placed in respective first locations of all cells of the display panel—this is typically performed using a pick-and-place technique where light emitting devices can be placed either in series or in parallel. When the pick-and-place is complete, the panel is tested through a visual inspection. Cells including devices that don't operate are recognized. The locations of these defective cells can be stored in a memory element available to the matrix controller (not shown) and in this way, a map of defective cells within the display can be produced. This panel map can be used later for programming the display.

(13) A second pick-and-place round now places devices in second locations of cells previously identified as containing defective devices. The second pick-and-place round can employ known good light emitting devices and it will be seen that where the natural yield of these devices is high, then only a small minority of such devices need to be tested in order that they would be known to be good. Nonetheless, especially where each cell includes locations for more than two devices, this testing may not be necessary.

(14) After the second pick-and-place, another visual inspection occurs. For any second placed device identified as working properly, the programming of the pixel can remain as before so that the second placed device can work according to the panel map. If any second placed device is identified as not working, a third light emitting device can be placed in the cell. Again, these device may be known good devices and again it will be seen that if only these devices were to be tested, only an extremely small number of such tests would need to be performed. The location of the second placed devices that don't work can again be stored in memory in a second panel map which can be used for programming the display.

(15) As will be appreciated, this process can be repeated for as many devices as any potentially defective cell may be able to accommodate. Depending on the reliability of devices being placed, placement of two, three or more device could be necessary in order to achieve acceptable yield.

(16) It can be understood, that as the number of devices which a cell might be able to accommodate is increased, more complex circuitry will be require to control them. A maximum number of locations is determined by the TFT process, the available pixel area and the light emitting device manufacture and assembly method.

(17) Turning now to examples of cell design which can be employed in displays being populated according to the method of FIG. 2.

(18) FIG. 3 shows a cell design which can be employed within an active drive matrix and where locations for up to two placed light emitting devices, in this case ILED1 and ILED2, are connected in series. The control circuit comprises 3 TFTs T1 . . . T3 and a charge storage capacitance. A data driver signal, such as D1 . . . Dm in FIG. 1 for the cell is provided by a data driver to the cell through an otherwise conventional voltage threshold compensation and driving circuit. Alternatively, a voltage threshold compensation and driving circuit such as disclosed in UK Patent Application No. 1606517.9 (Ref: I35-1702-02 GB) filed 14 Apr. 2016 can be employed. By comparison to a conventional 2T1C cell design, an additional control signal is needed and it is asserted (or not) in accordance with the value in the panel map for the display. Indeed the control signal state is essentially permanent as long as the value for the cell in the panel map does not change.

(19) If ILED1 has been checked as working normally, during the frame programming period when the scan driver signal, such as S1 . . . Sn in FIG. 1, is asserted, the control signal of the sub-pixel will have high “1” value, so turning ON T2 and connecting ILED1 to ground so that it will emit light in accordance with the data driver signal.

(20) Where ILED2 has been placed, because ILED1 has been identified as defective, the control signal will have low “0” value meaning that T2 will be turned OFF and T1 ON, so ILED1 will be shorted and ILED2 to be connected directly to the driving circuit. Therefore, ILED1 will not emit light and only ILED 2 will do so. The truth table for this situation is shown in Table 1.

(21) TABLE-US-00001 TABLE 1 Truth Table for two ILEDs connected in series configuration Control ILED 0 ILED2 1 ILED1

(22) In variants of the cell design of FIG. 3, a discrete storage capacitor may not be necessary, particularly for a display with a high refresh rate and where the voltage at the gates of T1 and T2 would remain stable enough without the need of the capacitance. In some implementations, a data driver signal (not shown) provided through the driving circuit might be asserted for a complete frame. In this case, the Scan signal could have the same timing as the data driver signal and again the capacitance would not be needed.

(23) As indicated above, if the desired yield can't be achieved with a cell design capable of locating two light emitting devices, a third one can be located. For the three devices, the complexity of the cell design is increased.

(24) Nonetheless, different configurations are possible depending on the TFT process being employed, for example, if only n-type TFTs are available (amorphous silicon TFTs and indium-Gallium-Zinc-Oxide—IGZO TFTs), or only p-type TFTs are available (Low Temperature polycrystalline Silicon—LTPS) or if CMOS devices are available (LTPS).

(25) Referring now to FIG. 4, there is shown a cell design implemented with n-type TFTs T1 . . . T4 only, whereas the cell design of FIG. 5 is implemented with CMOS TFTs T1 . . . T4. A configuration with only p-type TFTs would be the same as in FIG. 4, with all n-type TFTs being replaced with p-type TFTs.

(26) In both of these cases, either a constant supply voltage Vdd is provided through the driving circuit, or the driving circuit is switched using a PWM type pulsed signal.

(27) As it can be seen from FIGS. 4 and 5, three control signals (A, B and C) are required as well as additional circuitry to deliver these control signals. Thus instead of a single bit data line as for the cell design of FIG. 3, this might comprise 3 bits. However, when implementing a redundancy scheme for the display, the relative values of these 3 bits would remain constant through the life of the display, with each 3 bit combination being specific to a cell in accordance with the panel map values for the cell.

(28) In each of FIGS. 4 and 5 when a defective light emitting device is placed, it will be shorted by a corresponding TFT and at the same time a following functioning iLED should be connected to ground so it can emit light. Thus, when ILED1 should emit light, it is connected to ground through T3, while when it is not to emit light, it is shorted by T1. For each configuration of TFT process, a truth table exists which illustrates the way that the additional circuitry has to be programmed, so that a corresponding ILED will emit light. The truth tables for each case are shown below.

(29) TABLE-US-00002 TABLE 2 Truth Table for FIG. 4 implemented with only n-type TFTs. A B C ILED 0 0 0 All 0 0 1 ILED2 0 1 0 ILED1 0 1 1 None (Repair) 1 0 0 ILED3 1 0 1 None (Repair) 1 1 0 ILED3 1 1 1 None (Repair)

(30) TABLE-US-00003 TABLE 3 Truth Table for FIG. 4 implemented with only p-type TFTs. A B C ILED 0 0 0 None (Repair) 0 0 1 ILED3 0 1 0 None (Repair) 0 1 1 ILED3 1 0 0 None (Repair) 1 0 1 ILED1 1 1 0 ILED2 1 1 1 All

(31) TABLE-US-00004 TABLE 4 Truth Table for FIG. 5 implemented with CMOS TFTs. A B C ILED 0 0 0 ILED2 0 0 1 All 0 1 0 None (Repair) 0 1 1 ILED3 1 0 0 None (Repair) 1 0 1 ILED1 1 1 0 None (Repair) 1 1 1 ILED1

(32) As it is be seen from the truth tables, there are states where “All” emitting devices can be biased in order to emit light. Although there may be applications where this could be useful, it will be appreciated that applications which involve attempting to drive all placed devices could cause a cell potentially containing a malfunctioning device to be driven in an uncontrolled manner.

(33) There are also states where no devices will emit, since all of them are shorted by the TFTs, and this can be used as a repair method. Thus, if the third (or last) ILED which is placed doesn't work properly, it is preferable to treat the pixel comprising the cell as a “black” pixel and not to emit light, rather than driving the cell in an uncontrolled manner. Thus, in this case, the TFTs for a pixel are programmed in such way that all devices of the pixel are shorted to ensure that they will not emit light.

(34) In the circuits a FIGS. 4 and 5, a scan driver signal is not shown, however, it will be appreciated that each of control signals A, B and C can be connected to the gates of TFTs T1 . . . T4 via respective further TFTs (not shown) controlled by a scan driver signal as in the circuit of FIG. 3.

(35) The configurations of FIGS. 3-5 have light emitting devices connected in series to a common driving circuit and voltage threshold compensation circuit.

(36) FIG. 6 shows a cell design where three ILEDs are connected in parallel. The circuit of FIG. 6 can be implemented with a CMOS TFT process, as it requires oppositely doped TFTs. In FIG. 6, if only ILED1 is placed, control signal A has low “0” value resulting in T2 being turned ON. ILED1 is directly connected to the driving circuit so it emits light. When control signal A has high “1” value, T2 is turned ON and ILED1 is shorted and then control signal B will determine which from ILED2 or ILED3 will emit light. If control signal B has low value “0”, ILED2 will emit light and if control signal B has high “1” value, ILED3 will illuminate. The corresponding truth table is shown below.

(37) TABLE-US-00005 TABLE 5 Truth Table for FIG. 6 implemented with CMOS TFTs. A B ILED 0 X ILED1 1 0 ILED2 1 1 ILED3
X: either “1” or “0”

(38) One issue with the parallel configuration of FIG. 6, is the varying number of TFTs lying between the driving circuit and the selected light emitting device, especially the number of intermediate TFTs when ILED3 should emit light. This implies that the power supply voltage has to be increased to ensure that all TFTs operate in saturation region and the voltage at the ILED anode is more than its threshold voltage. Also, with the ILEDs connected in parallel and only the control signals A, B, C, one ILED will always emit light, meaning that no “Repair” configuration exists as in the implementations of FIGS. 4 and 5.

(39) In general in the above described examples, it is desirable to drive only a selected one light emitting device within a cell during operation of the display.

(40) However, there are applications where it can be useful to place more than one light emitting device within a cell and to selectively drive these.

(41) One such application provides a display with dynamic viewing angle. According to this application, one ILED with narrow and another ILED with wide beam are picked-and-placed within respective locations in every cell of the display (or at least those whose mode is to be changed). Depending on the required display mode, the corresponding ILED will be biased as explained above. For example, if only one user is viewing the display, the narrow beam (narrow viewing angle) devices would operate and if two or more users are viewing the display, the wide beam (wide viewing angle) would be used. (This technique could of course be extended to cover three different types of light emitting device being placed within each cell and being selected according to the required display mode.)

(42) In multi-display mode case, the value of the control signal or the relative values of signals A, B, C, rather than being permanent through the life of the display in accordance with the panel map, is switched dynamically according to the required display mode for the display. Also, rather than providing a per cell control signal or signals A, B, C, the control or A, B, C signals would be global, applying to the whole matrix. In a multi-display mode case where redundancy is not provided, there would be no need for a panel map.

(43) Swapping between modes can either be user driven; or can be automatic in response to viewing conditions (for example, if the display controller detects the number of people viewing the display).

(44) In a still further application, again two or more different sets of light emitting devices can be placed and these can be selectively driven to provide a display which can selectively operate in one of a 2D display mode and a 3D display mode.

(45) Referring now to FIG. 7, discrete light emitting devices, such as iLEDs, of differing colors behave differently than OLEDs in that there is a similar limited current range where the efficiency of red, green and blue discrete light emitting devices can be high and similar.

(46) In the cell design of FIG. 7, not only are light emitting devices of the same color or type connected in series, but light emitting devices for sub-pixels of a cell are connected in series and driven from a common voltage threshold compensation and driving circuit. Since the red, green and blue light emitting devices are connected in series, only one driving circuit is needed, meaning that the area efficiency of the pixel on the substrate is increased so enabling high display resolution.

(47) In FIG. 7 three identical circuits, each comprising two TFTs are applied to each ILED. One TFT (T2, T4 and T6) is connected in parallel to each ILED and a second TFT (T1, T3 and T5) is connected to the individual control signal Red Data (RD), Green Data (GD) and Blue Data (BD). During frame programming, a Scan driver pulse goes high “1” and T1, T3 and T5 are turned ON. Whether RGB ILEDs are to be turned ON or OFF is determined by the value of RD, BD, GD. If RD is high “1”, T2 will be turned ON and Red ILED will be shorted resulting in not emitting light. On the other hand, if RD is low “0”, T2 will remain OFF, so the Red ILED will be biased with the current produced by the driving circuit and it will emit light. The same procedure also occurs for Green ILED (T4 with GD) and Blue ILED (T6 and BD).

(48) In the circuit of FIG. 7, the ILEDS could be turned OFF, by asserting RD, BD and GD during a frame programming period. However, this would mean that power would be drawn through the transistors T2, T4 and T6 during this time. Thus, in the circuit of FIG. 7 an additional control signal Black data (BLD) controls power to the driving circuit. When BLD is asserted, no power flows through the pixel. The switching frequency of the BLD control line would typically be much less compared to the other data lines RD, BD, GD because this will only be activated for completely black pixels. Furthermore, using BLD improves the contrast ratio for a pixel because in this way, all current is blocked, resulting in a much deeper black colour.

(49) As will be appreciated, not all of the iLEDs of FIG. 7 should emit light contemporaneously or to the same extent, in order to produce the required grey-scale and color gamut for a frame. Ideally, the control signals driving the cell would be digital to ensure that each iLED would operate when switched on at its optimal operating current.

(50) Thus, displays incorporating the pixel design of FIG. 7 are either driven using a Pulse—Width Modulation (PWM) or color sequential scheme of the type referenced above.

(51) In this case, instead of a constant Vdd, PWM signals are provided from a PWM driver such as shown in FIG. 1. BLD, RD, GD and BD control signals are asserted in accordance with the required state of a pixel for a given sub-frame, each sub-frame varying in length. In both cases, if 8-bits grey-scale is desired, the frame time is divided into 8 sub-frames, each one having different duration. The Most significant Bit (MSB) of the grey-scale will have the longest pulse while the Least Significant Bit (LSB) will have the narrowest pulse. The difference between PWM and color sequence is that in the case of the PWM, all ILEDs emit light simultaneously during one sub-frame while for the color sequence each sub-frame is further divided into Red, Green and Blue periods meaning that for a given sub-frame duration only Red ILEDs over the whole panel emit light, then only Green ILEDs and finally only Blue ILEDs.

(52) The difference results in the colour sequential scheme requiring at least three times higher switching frequency since during the same sub-frame, it has to switch three times, once for each colour. Both driving schemes suffer from different visual artefacts that can be solved using special driving algorithms. In any case, the most important criterion for both schemes is that their frequency to be high enough, so that ILED ON/OFF transition is not noticeable by the human eye.

(53) In the circuit of FIG. 7, a maximum frequency can be achieved due to the lack of storage capacitance on all data paths; whereas in other designs a storage capacitance can be connected at the gate node of the driving TFT in order to keep the gate voltage constant during a sub-frame duration.

(54) In FIG. 7, the control signals determine if the parallel connected TFTs are ON or OFF. This means that their gate voltage may vary during the frame duration as long as it ensures that the TFT is turned ON. The lack of storage capacitance also reduces the data column power consumption which is among the highest in a display.

(55) On the other hand, if storage capacitance is provided for each sub-pixel and a data driver capable of providing analog or digital outputs were employed, then the hybrid driving scheme described in UK Patent Application No. 1604699.7 (Ref: I35-1702-01 GB) filed 21 Mar. 2016 could be used to limit the switching frequency required of the PWM driver.

(56) Finally, it will be appreciated that when a semiconductor light emitting device, such as an iLED emits light, there is a voltage drop between its terminals, referred to as threshold voltage. Where light emitting devices are connected in series, the total voltage difference between the output of the driving circuit and the ground should be at least the sum of their threshold voltages, for example, in FIG. 7 the threshold voltages for the red, green and blue ILEDS. This means that the high supply voltage (Vdd) (or pulsed PWM signal voltage) will be higher by comparison to circuits driving only one light emitting device in order to ensure all devices can emit light. Nonetheless, even though the supply voltage Vdd will be increased in FIG. 7, the total power consumption will be kept the same since the total ILEDs bias current is ⅓ of that for circuits driving only one light emitting device due to current sharing between the light emitting devices.

(57) Referring now to FIG. 8, in some applications, in which sufficiently high supply voltages are available, more than one pixel can share the driving circuit. Since the driving conditions of all ILEDs are the same, the driving circuit that produces the necessary bias current can be shared between more than one pixel. The only limitation for the number of the pixels that can share the driving circuit is how high the supply voltage is. The condition that has to be satisfied is:

(58) $V\mathrm{dd}>\Delta V_{\mathrm{driving},\mathrm{circuit}}+\underset{i=0}{\overset{n}{.Math.}}{V\mathrm{th}}_i$

(59) where ΔV.sub.driving,circuit is the voltage drop over the driving circuit while producing the necessary Σ.sub.i=0.sup.n Vth.sub.i, Vth.sub.i is the combined threshold voltage of the light emitting devices within a pixel i existing on the driving circuit to ground path and n is the maximum number of pixels.

(60) As mentioned above, the voltage threshold and driving circuit illustrated in FIGS. 3-8 can be any suitable design which produces an accurate bias current with immunity to TFT threshold voltage variations. Furthermore, the driving circuit can be either dynamic meaning that it can be programmed during every frame or sub-frame or DC (only programmed once) or even programmed periodically with a frequency much lower than the frame rate (FR). In the case of the DC or periodically programmed cases, any existing current source can be used if it can be implemented in the available pixel area.

(61) In the matrix of FIG. 1, the data driver is shown as a singular unit, however, it will be appreciated that in embodiments of the present invention the functionality of the data driver (and indeed the other peripheral components) can be divided among more than one unit, for example, with one unit providing data signals and another providing control signals.

(62) The features of the above described embodiments can be used either in combination or individually within a given display. In each case, embodiments are suitable for either wearable displays, such as smartwatches or large panel displays.