Liquid crystal device

0001The invention relates to a liquid crystal device.

0002A conventional active matrix liquid crystal panel as disclosed in "SID-83 digest pp. 156 - 157, B/W and colour LC video displays addressed by polysilicon TFTs" (Morozumi et al) is illustrated in Figure 19. Such an active matrix panel 1 comprises a picture element matrix 2, which utilises thin film transistors and which is formed on a transparent substrate. Gate line drive circuits and source line drive circuits 4, in the form of MOS integrated circuits composed of mono-crystal silicon, are attached to the edges of the active matrix panel 1 as shown, each drive circuit 4 being mounted on a respective flexible substrate 3 and being connected to the picture element matrix 2 of the active matrix panel 1 at pads 5. A mounting substrate 6 provides both mechanical support for the flexible substrates 3 and the active matrix panel 1 and electrical connection between the drive circuits 4 and external circuitry (not shown).

0003This conventional arrangement has the following disadvantages:

(1) High resolution is prevented.

0004In the prior art, each source or gate line of the active matrix panel 1 is connected to the associated drive circuit 4 at a respective pad 5 so that close positioning of the picture elements is limited by the pad interval required for forming the connections during mounting. Therefore, mass production of the active matrix panel with a picture element spacing of 100 microns or less is very difficult and so high resolution cannot be obtained.

(2) Miniaturisation of the display device is prevented.

0005In the conventional arrangement shown in Figure 19, each drive circuit 4 is mounted outside the active matrix panel 1 so that the external dimensions of the mounting substrate 6 have to be a fourth or fifth as large again as the dimensions of the picture element matrix 2. Therefore, the dimensions of the display device employing the active matrix panel 1 must inevitably be larger than the dimensions of the picture element matrix 2, which is the portion contributing to the display, and this limits the application of the active matrix panel in a micro-monitor, such as a view finder for a video camera.

(3) The production cost is expensive.

0006The production of a display device employing the arrangement shown in Figure 19 involves the process of connecting the flexible substrates 3 with the active matrix panel 1, the process of connecting the drive circuits 4 with the flexible substrates 3, and the process of mounting the flexible substrates 3 and the active matrix panel 1 on the mounting substrate 6, which renders production expensive.

(4) The reliability is low.

0007Due to the many connections, for example between the active matrix panel 1 and the flexible substrates 3 and between the drive circuits 4 and the flexible substrates 3, arid the stresses applied to the connecting portions in use, the connecting strength has been found to be insufficient. Thus, the reliability of the whole display device is inherently low, whereas the expense of ensuring reliability is great.

0008According to a first aspect of the present invention, there is provided a liquid crystal device for an electro-optical device, the liquid crystal device including a plurality of liquid crystal cells wherein data signals are supplied to the liquid crystal cells through a plurality of field effect transistors, arranged in a plurality of picture elements, the liquid crystal device comprising: <ul id="ul0001" list-style="none" compact="compact"><li>a substrate having a pixel region and at least one driver region;</li><li>driving means formed in the at least one driver region for driving said field effect transistors,</li><li>the driving means comprising at least a shift register having CMOS thin film transistors and sample and hold circuits, the CMOS thin film transistors including first conductive type thin film transistors and second conductive type thin film transistors, and the sample and hold circuits including at least first conductive type thin film transistors, the gate length of the first conductive type thin film transistors of CMOS thin film transistors being greater than the gate length of the first conductive type thin film transistors of the sample and hold circuit.</li></ul>

0009According to a second aspect of the present invention, there is provided a projection type colour display device including a light source, at least one light valve, and at least one projecting optical system for projecting an image, the at least one light valve including a plurality of liquid crystal cells, wherein signals are supplied to the plurality of liquid crystal cells, through a plurality of field effect transistors, arranged in a plurality of picture elements, the light valve comprising: <ul id="ul0002" list-style="none" compact="compact"><li>a substrate having a pixel region and at least one driver region;</li><li>driving means formed in the at least one driver region for driving said field effect transistors,</li><li>the driving means comprising at least a shift register having CMOS thin film transistors and sample and hold circuits, the CMOS thin film transistors including first conductive type thin film transistors and second conductive type thin film transistors, and the sample and hold circuits including at least first conductive type thin film transistors, the gate length of the first conductive type thin film transistors of CMOS thin film transistors being greater than the gate length of the first conductive type thin film transistors of the sample and hold circuit.</li></ul>

0010A third aspect of the invention provides a driving circuit applied to a liquid crystal device for an electro-optical device, the liquid crystal device including a plurality of liquid crystal cells wherein data signals are supplied to the liquid crystal cells through a plurality of field effect transistors arranged in a plurality of picture elements, the driving circuit formed in a substrate for driving said field effect transistors, and comprising at least a shift register having CMOS thin film transistors and sample and hold circuits, the CMOS thin film transistors including first conductive type thin film transistors and second type thin film transistors, and the sample and hold circuits including at least first conductive type thin film transistors, the gate length of the first conductive type thin film transistors of CMOS thin film transistors being greater than the gate length of the first conductive type thin film transistors of the sample and hold circuit.

0011The invention will be described further, by way of example, with reference to the accompanying drawings, in which: <ul id="ul0003" list-style="none" compact="compact"><li>Figure 1 is a circuit diagram of an active matrix panel employed in the present invention;</li><li>Figures 2(a) to 2(f) show examples of the detailed structure of a drive circuit shown in Figure 1;</li><li>Figures 3(a) and 3(b) are cross sections through parts of the active matrix panel showing Figure 1;</li><li>Figures 4(a) to 4(d) illustrate stages in the method of manufacturing an active matrix panel;</li><li>Figure 5 is a graph showing the characteristic of a thin film transistor employed in the invention in comparison with that of a mono-crystalline silicon MOSFET;</li><li>Figure 6 is a diagram for illustrating the terms "gate length" and "gate width";</li><li>Figure 7 is a cross sectional view through a thin film transistor employed in the invention showing depletion layer width and film thickness in a layer of silicon thin film;</li><li>Figures 8, 9, 10(a) and 10(b) illustrate a first modification of the active matrix panel;</li><li>Figures 11(a) and 11(b) illustrate a second modification of the active matrix panel;</li><li>Figure 12 illustrates a third modification of the active matrix panel;</li><li>Figures 13(a) and 13(b) illustrate a fourth modification of the active matrix panel;</li><li>Figure 14 illustrates a fifth modification of the active matrix panel;</li><li>Figures 15(a) and 15(b) illustrate a sixth modification of the active matrix panel;</li><li>Figures 16(a) and 16(b) illustrate a seventh modification of the active matrix panel;</li><li>Figure 17 is a diagram illustrating a first application of the active matrix panel;</li><li>Figure 18 is a diagram illustrating a second application of the active matrix panel; and</li><li>Figure 19 is a circuit diagram of an active matrix panel according to the prior art.</li></ul>

0012Figure 1 shows an embodiment of the invention, wherein an active matrix panel 11 comprises a source line drive circuit 12, a gate line drive circuit 21, and a picture element matrix 22, all being of complementary metal oxide semi-conductor structure (hereinafter referred to as a CMOS structure) made of silicon thin film and formed on the same transparent substrate. The source line drive circuit 12 includes a shift register 13, a sample and hold circuit comprising thin film transistors 17, 18 and 19 (hereinafter referred to as TFTs) and video signal buses 14, 15 and 16. The gate line drive circuit 21 includes a shift register 20, and it may also include a buffer 23 if occasion demands. The picture element matrix 22 includes a plurality of source lines 26, 27, and 28 connected to the source line drive circuit 12, a plurality of gate lines 24 and 25 connected to the gate line drive circuit 21, and a plurality of picture elements 32 and 33 formed at intersections of the source and gate lines. Each picture element 32, 33 includes a TFT 29, and a liquid crystal cell 30, which comprises a picture element electrode, and opposite electrode 31 and a layer of liquid crystal material between the electrodes. It is possible for other circuits, such as a counter or a de-coder having the function of selecting the source lines and the gate lines in order, to be substituted for the shift registers 13 and 20. A clock signal CLX, a start signal DX and video signals V1, V2 and V3 are supplied in use to input terminals 34, 35 and 36 respectively of the source line drive circuit 12. A clock signal CLY and a start signal DY are supplied to input terminals 37 and 18 respectively of the gate line drive circuit 21.

0013The shift registers 13 and 20 shown in Figure 1 can be static or dynamic circuits composed of complementary TFTS, consisting of P type and N type TFTS, or they can be dynamic or static circuits composed of monoconductive TFTs. Given the performance of a TFT as a device, a static type of circuit composed of complementary TFTS is the optimum one. Since a TFT for an active matrix panel is generally formed on an insulating substrate with poly-crystalline or amorphous silicon, a smaller ON current and a larger OFF current are produced than is the case with a metal oxide semi-conductor field effect transistor (hereinafter referred to as a MOSFET). The reason for this is that the trap density existing in a silicon thin film is very high in comparison with that in a mono-crystalline silicon whereby the carrier mobility is low and re-combination of carriers at a P-N junction which is reverse biased occurs frequently. In view of these characteristics of the TFT as a device, a static shift register composed of complementary TFTS is preferred generally and such a shift register is also preferred for the following reasons: <ul id="ul0004" list-style="none" compact="compact"><li>(1) As a TFT has a large OFF current, the operating voltage range, operating frequency range, and operating temperature range of a dynamic circuit consisting of TFTS is very small.</li><li>(2) A drive circuit needs to have a CMOS structure with a low consumption of current for making the best use of the advantage of the active matrix liquid crystal panel, which is its low current consumption.</li><li>(3) The ON current value can be smaller than that in a mono-conductive MOS dynamic shift register.</li></ul>

0014Figure 2(a) shows a suitable circuit structure for each of the shift registers 13 and 20 in Figure 1. Inverters 41 and 42 in Figure 2(a) consist of a P type TFT 47 and an N type TFT 48 as shown in Figure 2(b). Further, clocked inverters 43 and 46 in Figure 2(a) consist of P type TFTS 49, 50 and of N type TFTS 51, 52, respectively, as shown in Figure 2(c). A clock signal CL is input to the gate of the N type TFT 52, and an inverted clock signal <maths id="math0001"><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">CL</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0001.tif" /></maths> is input to the gate of the P type TFT 49. Likewise, clocked inverters 44 and 45 in Figure 2(a) consist of P type TFTS 53, 54 and N type TFTs 55, 56, respectively, as shown in Figure 2(d). The inverted clock signal <maths id="math0002"><math display="inline"><mrow><mover accent="true"><mrow><mtext mathvariant="italic">CL</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0002.tif" /></maths> is input to the gate of the N type TFT 56 and the clock signal CL is input to the gate of the P type TFT 53. An alternative possibility is for a circuit made up of an analog switch consisting of an inverter 57, an N type TFT 58 and a P type TFT 59, as shown in Figure 2(e), to be substituted for the clocked inverters 43 and 46, and for a circuit made up of an analog switch consisting of an inverter 60, an N type TFT 61 and a P type TFT 62, as shown in Figure 2(f), to be substituted for the clocked inverters 44 and 45.

0015As described above, it is very beneficial to construct a drive circuit for an active matrix panel from TFTs having a CMOS structure. However, the mere application of a complementary TFT integrated circuit in the prior art has disadvantages as described below: <ul id="ul0005" list-style="none" compact="compact"><li>(1) A method of production for integrating both a P type TFT and an N type TFT on the same substrate is complicated and renders the overall circuit expensive.,</li><li>(2) It is difficult to form a P type TFT and an N type TFT having balanced characteristics, which is essential for forming a complementary TFT integrated circuit.</li><li>(3) A P type TFT and an N type TFT do not have enough driving ability for realising a drive circuit.</li></ul>

0016Such disadvantages are solved in the present instance by employing a method of manufacture, and a structure having dimensions and utilising materials as described below.

0017Figure 3(a) is a section through a pair of complementary TFTS for forming either the source line drive circuit 12 or the gate line drive circuit 21 shown in Figure 1, and Figure 3(b) is a section through the respective picture element 32 of Figure 1 which includes a liquid crystal cell.

0018In Figure 3(a), a P type TFT 99 and an N type TFT 100 are formed on an insulating substrate 71 made of glass, quartz crystal or the like. Silicon thin film regions 73, 76 are channel regions of the TFTS 99, 100 and silicon thin film regions 72, 74, 75, 77 are source and drain regions. The silicon thin film regions 72 and 74 are impurity doped to be P type, and the silicon thin film regions 75 and 77 are impurity doped to be N type. Gate insulating film regions 78 and 79 are made of silicon oxide (for example Si0₂), silicon nitride or the like, and gate electrodes 80 and 81 are made of polycrystalline silicon, metal, metal silicide or the like. The TFTs further comprise an insulating film layer 82 made of silicon oxide (for example Si0₂), silicon nitride or the like, a line layer 83 made of metal or the like, an insulating film layer 84 made of silicon oxide (for example Si0₂), silicon nitride or the like, and a passivation film layer 85.

0019In Figure 3(b), the reference numeral 86 indicates the same insulating substrate 71 as in Figure 3(a), a picture element electrode 94 consisting of transparent conductive film made of ITO (indium tin oxide) or the like, and a picture element TFT 101 being formed thereon., The reference numerals 87, 88 and 89 indicate silicon thin film regions which are formed from the same layers as the regions 72, 73, 74 and 75, 76, 77, the silicon thin film region 88 forming a channel region and the silicon thin film regions 87 and 89 forming source and drain regions, respectively, of the TFT 101. The regions 87 and 89 are impurity doped to be P type or N type, the impurities in such regions being the same as those included in the regions 72 and 74 or the regions 75 and 77. The reference numeral 90 indicates a gate insulating film region, which is formed from the same layer as the gate insulating film regions 78 and 79, and the reference numeral 91 indicates a gate electrode, which is formed from the same layer as the gate electrodes 80 and 81. The reference numeral 92 indicates an insulating film region which is formed from the same layer as the insulating film region 82, the reference numeral 93 indicates a line layer, which is formed from the same layer as the line layer 83, and the reference numeral 95 indicates an insulating film layer, which is formed from the same layer as the insulating film layer 84. A layer 96 of liquid crystal is provided between the insulating film layer 95 and a counter electrode 97, which includes a transparent conductive film layer and which is provided on a transparent substrate 98.

0020Accordingly, the source/drain region, the channel region, the gate insulating film region and the gate electrode of each of the TFTs 99 and 100 forming the drive circuit are provided by the same thin film layers as the corresponding regions of the picture element TFT 101. Further, the TFTs in the source line drive circuit and the gate line drive circuit are connected through the line layer 83, which is designed to have a low sheet resistance (being formed, for example, from a metal such as aluminium), and the line layer 93 consisting of the same layer as the line layer 83 provides a source line in the picture element matrix., A picture element electrode 94 is made of a transparent conductive film layer such as ITO. In the case that the line layer 93 is made of aluminium or alumi-silicide and the transparent conductive film layer is made of ITO and no insulating film layer is arranged therebetween, through holes 102 and 103 opened simultaneously can be used for connecting the line layer 93 and the electrode 94 and the silicon thin film regions 87, 89 whereby the manufacturing process can be simplified. In this case, the aluminium and the ITO are processed in respective etching solutions and the ITO is formed for patterning in the stage prior to the aluminium, taking advantage of the fact that the ITO is not soaked in the etching solution for aluminium.

0021In Figure 3(b), the insulating film layer 95 provides a capacitor for preventing application of a DC voltage to the liquid crystal layer 96 and the capacitance value thereof should be large enough in comparison with that of the picture element for this purpose. Thus, the thickness of the layer 95 should be a predetermined value (for example, approximately 3000 Angstrom) or less. On the other hand, the TFTs 99, 100 of the drive circuit shown in Figure 3(a) need to be covered by the passivation film layer 85, which is designed to have a thickness larger than a pre-determined value (for example, approximately 1 micron) for ensuring wet-proofing. For forming the passivation film layer 85, it is most effective to form a film all over the active matrix substrate and then to remove the film except in the region of the drive circuits. Hence, the passivation film layer 85 is made from a material such as polyimide or the like processed with an etching solution, in which the insulating film layer 84 is not soaked.

0022The method of manufacture for such a structure and the characteristics of the complementary TFTs obtained therefrom are described below. In a method of manufacturing a CMOS integrated circuit utilising conventional mono-crystalline silicon, at least four photo processes are needed (a process for forming a low concentration P well, a process for forming a P type stopper layer, a process for forming the source/drain regions of each P type MOSFET, and a process for forming the source/drain regions of each N type MOSFTET) in comparison with a method of manufacturing an N type MOSFET integrated circuit. Contrary to this, a complementary TFT integrated circuit according tot he invention can be realised by adding no more than one photo process as compared with a method of manufacture for a mono-conductive TFT integrated circuit.

0023Figures 4(a) to 4(d) illustrate an example of the main stages in a method of manufacturing the active matrix panel. First, as shown in Figure 4(a), a desired pattern is formed after accumulating a layer of silicon thin film over a transparent insulating substrate 110 and thereby forming a channel region 111 of a P type TFT and channel regions 112 and 113 of N type TFTs. Next, gate insulating film regions 114, 115 and 116 are formed by means of thermal oxidation or chemical vapour deposition, and then gate electrodes 117, 118 and 119 are formed.

0024Next, as shown in Figure 4(b), acceptor impurities 120, for example boron, are implanted all over the surface utilising ion implantation techniques. Implanted acceptor impurities in regions 121, 122 are activated by subsequent heat treatment to form a P type semi-conductor,. Acceptor impurities are also added to regions 123, 124, 125 and 126, which are to be the source/drain regions of N type TFTs.

0025After this, as shown in Figure 4(c), the P type TFTs covered by a mask material such as photo resist 128, and donor impurities 127, such as phosphorus or arsenic, are implanted elsewhere with a higher concentration than that of the acceptor impurities 120. Implanted donor impurities in the regions 123 to 126 are activated by subsequent heat treatment. If the implant dosage of the acceptor ions is 1 x 10¹⁵ cm ^-2 and the implant dosage of the donor ions is 3 x 10¹⁵ cm^-2, then the regions 123 to 126 contain the equivalent of an implant dosage of the donor ions having a concentration of 2 x 10^-15 cm^-2. In this way, source/drain regions 123 to 126 of the N type TFTS are formed.

0026Subsequently, as shown in Figure 4(d), an insulating film layer 129 is accumulated after removing mask material 128, through holes are opened, a picture element electrode 131 made of transparent conductive film is formed and a line 130 made of metal or the like is formed.

0027As described, a P type TFT 132 and an N type TFT 133 for forming the drive circuit and an N type TFT 134 for forming a picture element TFT of the matrix panel are thus produced. It is a matter of course that a P type TFT for the matrix panel can also be produced. The TFTs thus obtained include a P type TFT having acceptor impurities in its source/drain region, and N type TFTs having both acceptor impurities and donor impurities in their source/drain regions but with the latter having a higher concentration than the former.

0028In the above process of manufacture, if donor impurities 127 are substituted for acceptor impurities 120 in Figure 4 (b) and acceptor impurities 120 are substituted for the donor impurities 127 in Figure 4 (c), an N type TFT 132 and P type TFTs 133, 134 as shown in Figure 4 (d) are obtained instead. The TFTs thus obtained include an N type TFT having donor impurities in its source/drain region and P type TFTs having both donor impurities and acceptor impurities in their source/drain regions but the latter having a higher concentration that the former.

0029A complementary TFT integrated circuit can thus be formed by adding the single photo process required to form the mask pattern 128 in Figure 4 (c), as compared with the method of manufacture for a mono-conductive TFT integrated circuit. As a result, an active matrix panel with a built-in drive circuit can be realised. From an economical point of view, the above method of manufacture is an improvement, and it is not inconvenient to form a mask pattern independently for the ion implantation of acceptor impurities and donor impurities. Further, in the complementary TFT integrated circuit manufactured by the above method, each TFT is separated in the form of an island on the insulating substrate 110 so that a specific process for element separation is not required. Further, being different from an integrated circuit made of monocrystalline silicon, no parasitic MOSFET is produced so that a channel stopper does not have to be formed.

0030Next, a means for realising a P type TFT and an N type TFT having balanced characteristics as required for a complementary integrated circuit will be described. TFTs making use of the groups II to VI semi-conductor compounds have been known in the art for a long time. However, a complementary TFT making use of a compound semi-conductor has not been realised for the following two reasons:- <ul id="ul0006" list-style="none" compact="compact"><li>(1) It is actually impossible to control and realise both P type and N type semi-conductors in the semi-conductor compound.</li><li>(2) It is very difficult to control the interface between the semi-conductor compound and the insulating film for a MOS construction.</li></ul>

0031Accordingly, source/drain regions and a channel region are made of silicon thin films. The carrier mobilities of amorphous silicon thin film and of poly-crystalline silicon thin film are shown in Table 1. As shown by the Table, the poly-crystalline thin film is the better for realising a complementary TFT integrated circuit because characteristics of P type and of N type semi-conductors can easily be made well balanced and the current supply capacity of the TFT can be made larger. <tables id="tabl0001" num="0001"><table frame="all"><title>TABLE 1</title><tgroup cols="3" colsep="1" rowsep="1"><colspec colnum="1" colname="col1" colwidth="52.50mm" /><colspec colnum="2" colname="col2" colwidth="52.50mm" /><colspec colnum="3" colname="col3" colwidth="52.50mm" /><thead valign="top"><row><entry namest="col1" nameend="col1" rowsep="0" /><entry namest="col2" nameend="col3" align="center">carrier mobility (cm²/V sec)</entry></row><row><entry namest="col1" nameend="col1" /><entry namest="col2" nameend="col2" align="center">N type</entry><entry namest="col3" nameend="col3" align="center">P type</entry></row></thead><tbody valign="top"><row rowsep="0"><entry namest="col1" nameend="col1" align="left">amorphous silicon</entry><entry namest="col2" nameend="col2" align="right">0.1 - 1</entry><entry namest="col3" nameend="col3" align="right">10^-4 - 10^-3</entry></row><row rowsep="1"><entry namest="col1" nameend="col1" align="left">poly-crystalline silicon</entry><entry namest="col2" nameend="col2" align="right">5 - 50</entry><entry namest="col3" nameend="col3" align="right">5 - 50</entry></row></tbody></tgroup></table></tables>

0032Next will be described a means for increasing the current supply capacity of a TFT, especially P type and N type TFTs for forming a drive circuit as in the present invention. As described above, the trap density is high in a TFT made of silicon thin film which is not mono-crystalline silicon so that the ON current is small and the OFF current is large in comparison with a mono-crystalline silicon MOSFET. Figure 5 is a graph showing a characteristic 140 of a mono-crystalline silicon MOSFET and a characteristic 141 of a TFT made of silicon thin film in the case in which the gate length, gate width and source/drain voltage VDS are the same. The abscissa represents the voltage VGS based on the source as the reference, and the ordinate represents the current IDS between the source and the drain. As shown by the graph, since the ON/OFF ratio of the TFT is small, the TFT 29 for the picture element matrix and the TFTs forming the drive circuit 12 and 21 in Figure 1 respectively should be formed with the optimum dimensions. For instance, when an image is to be displayed making use of an NTSC video signal, the TFT 29 for the picture element matrix should satisfy the following equations within the entire temperature range wherein the active matrix panel is used.<maths id="math0003"><math display="block"><mrow><mtable><mtr><mtd><mrow><mtable><mtr><mtd><mrow><mtable><mlabeledtr><mtext>(1)</mtext><mtd><mrow><msub><mrow><mtext>0.1 x C</mtext></mrow><mrow><mtext>1</mtext></mrow></msub><mtext> x R OFF1 ≳ 1/60 (sec)</mtext></mrow></mtd></mlabeledtr></mtable></mrow></mtd></mtr><mtr><mtd><mrow><mtable><mlabeledtr><mtext>(2)</mtext><mtd><mrow><msub><mrow><mtext>5 x C</mtext></mrow><mrow><mtext>1</mtext></mrow></msub><mtext> x R ON1 ≲ 10 (µ sec) </mtext></mrow></mtd></mlabeledtr></mtable></mrow></mtd></mtr></mtable></mrow></mtd></mtr></mtable></mrow></math><img file="EP0806701A1_D0003.tif" /></maths> where C₁ represents the total capacity of a picture element, and R ON1 and R OFF1 represent ON resistance and OFF resistance of the TFT, respectively. Equation (1) is a condition to be satisfied by all of the picture elements in the matrix panel during a holding operation (holding condition), and 90% and more of the electric charge which is written into any picture element can be held over for one field if it is satisfied. Equation (2) is a writing condition to be satisfied by all of the picture elements in the matrix panel during a writing operation (writing condition), and 99% and more of a video signal can be written into a picture element if it is satisfied. On the other hand, a TFT for forming a drive circuit should satisfy the following equation within the entire temperature range wherein the active matrix panel is used.<maths id="math0004" num="(3)"><math display="block"><mrow><mtext>K x (C2RON2 + C3RON3) ≲ 1/2 f</mtext></mrow></math><img file="EP0806701A1_D0004.tif" /></maths> where C2 and C3 represent the capacities of junctions 142 and 143 in Figure 2(a), RON2 and RON3 are the output resistances of the clocked inverter 43 and the inverter 41 of Figure 2(a) respectively, <u>f</u> is the clock frequency of the shift register 13 or 20, and K is a constant (K is approximately from 1.0 to 2.0, by experience). According to the actual survey or simulation performed. RON2 and RON3 of the TFT for forming the drive circuit should be a tenth or less of RON1 of a picture element TFT in order to realise a shift register with approximately 2 MHZ of the clock frequency <u>f</u>.

0033For realising such low output resistance, the gate length of the TFT for the drive circuit can be formed as short as possible within the limits of the breakdown voltage permitted. Further TFTs for the sample and hold circuit 17, 18 and 19 shown in Figure 1 permit a lower breakdown voltage than the TFTs for the shift register 13 so that the gate lengths of the former can be shorter than those of the latter. Figure 6 illustrates the definition of gate length L and Table 2 shows an example of the gate lengths for each of the TFTs. In Figure 6, a gate electrode 142 and a silicon thin film 143 making up a channel region together define a gate length 144 and a gate width 145. <tables id="tabl0002" num="0002"><table frame="all"><title>TABLE 2</title><tgroup cols="3" colsep="1" rowsep="1"><colspec colnum="1" colname="col1" colwidth="52.50mm" /><colspec colnum="2" colname="col2" colwidth="52.50mm" /><colspec colnum="3" colname="col3" colwidth="52.50mm" /><thead valign="top"><row><entry namest="col1" nameend="col1" rowsep="0" /><entry namest="col2" nameend="col3" align="center">Gate length L (µm)</entry></row><row><entry namest="col1" nameend="col1" /><entry namest="col2" nameend="col2" align="center">P type TFT</entry><entry namest="col3" nameend="col3" align="center">N type TFT</entry></row></thead><tbody valign="top"><row><entry namest="col1" nameend="col1" align="left">TFT for picture element matrix</entry><entry namest="col2" nameend="col2" /><entry namest="col3" nameend="col3" align="char" char=".">20.0</entry></row><row><entry namest="col1" nameend="col1" align="left">TFT for shift register</entry><entry namest="col2" nameend="col2" align="char" char=".">4.0</entry><entry namest="col3" nameend="col3" align="char" char=".">5.5</entry></row><row rowsep="1"><entry namest="col1" nameend="col1" align="left">TFT for sample and hold circuit</entry><entry namest="col2" nameend="col2" /><entry namest="col3" nameend="col3" align="char" char=".">4.5</entry></row></tbody></tgroup></table></tables>

0034It is also possible to raise the current supply capacity of a P type TFT and an N type TFT by constructing the TFT such that the thickness of the silicon thin film forming the channel region is smaller than the maximum width of the depletion layer extending over the surface of the silicon thin film.

0035The maximum width X_PMAX of the depletion layer of a P type TFT made of silicon thin film and the maximum width X_NMAX of the depletion layer of an N type TFT are represented by the following equations, respectively.<maths id="math0005"><math display="block"><mrow><mtable><mtr><mtd><mrow><mtable><mtr><mtd><mrow><mtable><mlabeledtr><mtext>(4)</mtext><mtd><mrow><msub><mrow><mtext>X</mtext></mrow><mrow><mtext>PMAX</mtext></mrow></msub><msup><mrow><mtext> = (2ε x 2 φ FP)</mtext></mrow><mrow><mtext>1/2</mtext></mrow></msup><msup><mrow><mtext> x (q x ND)</mtext></mrow><mrow><mtext>-1/2</mtext></mrow></msup></mrow></mtd></mlabeledtr></mtable></mrow></mtd></mtr><mtr><mtd><mrow><mtable><mlabeledtr><mtext>(5)</mtext><mtd><mrow><msub><mrow><mtext>X</mtext></mrow><mrow><mtext>NMAX</mtext></mrow></msub><msup><mrow><mtext> = (2ε x 2 φ FN)</mtext></mrow><mrow><mtext>1/2</mtext></mrow></msup><msup><mrow><mtext> x (q x NA)</mtext></mrow><mrow><mtext>-1/2</mtext></mrow></msup><mtext></mtext></mrow></mtd></mlabeledtr></mtable></mrow></mtd></mtr></mtable></mrow></mtd></mtr></mtable></mrow></math><img file="EP0806701A1_D0005.tif" /></maths> where <u>q</u> represents a unit electric charge, ε is a dielectric constant of silicon thin film, φ FP represents a fermi function for a P type TFT, φ FN represents a fermi function for an N type TFT, and ND and NA represent equivalent donor and acceptor densities for the channel regions, respectively. Further, equivalent donor and acceptor densities are determined by the density of the donor impurities existing in the region, the density of the acceptor impurities and the trap density acting as donor and acceptor. According to the present invention, the thickness of the silicon thin film of the channel region in a P type TFT and an N type TFT is formed to be smaller than either X_PMAX or X_NMAX. Figure 7 is a cross sectional view of a TFT where a depletion layer is formed. An insulating substrate 146 carries a silicon thin film region 147 forming a channel region, and silicon thin film regions 148, 149 forming source/drain regions. A gate insulating film layer 150 is formed on the regions 147, 148 and 149, and a gate electrode 151 is provided on the layer 150. The thickness of the silicon thin film is indicated by <u>tsi</u>, and the width of the depletion layer formed on the silicon thin film surface is indicated by <u>x</u>.

0036The basic criteria for providing the preferred drive circuit for an active matrix panel are thus established as follows:- <ul id="ul0007" list-style="none" compact="compact"><li>(1) Each drive circuit is given a static form composed of complementary TFTs.</li><li>(2) A method of manufacture and a structure for a complementary TFT integrated circuit are provided.</li><li>(3) The characteristics of the P type and of the N type semi-conductors are made to be well balanced.</li><li>(4) The TFT design is improved so as to enhance the capacity for driving the load.</li></ul>

0037Various further improvements will now be described.

0038A first modification in the form of a layout for an active matrix panel is shown in Figure 8, which is a plan view of an active matrix panel 160 showing the layout of each of the circuit blocks. A source line drive circuit 161 (and/or 162) is formed at the periphery of the active matrix panel 160 at the top (and/or bottom) as illustrated, in order to form a normal image, and a shift register 163, a buffer 164, a video signal bus 165 and a sample hold circuit 166 are arranged one inside the other relative to the periphery within the source line drive circuit 161. A gate line drive circuit 167 (and/or 170) is formed at the left (and/or the right) hand periphery of the active matrix panel 160 and a shift register 168 and a buffer 169 are arranged in order inwardly from the periphery within the gate line drive circuit. A picture element matrix 171 is formed at the centre or the active matrix panel 160 for connection to the source line drive circuit 161 (162) and the gate line drive circuit 167 (170), and Input terminals 172, 173, 174 and 175 are arranged at the corners of the panel. Signals are transmitted within the panel in use in the directions indicated by arrows 176 to 180. Limited space can be put to the most effective practical use by the layout as described above.

0039Further, a layout as shown in Figure 9 is utilised within the source line drive circuit and/or the gate line drive circuit for forming unit cells within a limited dimension equivalent to the spacing between picture elements (or twice such spacing). In Figure 9 the cells 196 to 198 have a width D corresponding to the spacing between picture elements (or twice such spacing). Arranging the drive circuit in repeating cells having a cell width D, while also employing the layout in Figure 8 makes effective use of the overall space available. In Figure 9, the cells 196 to 198 are shown surrounded by broken lines. A positive power source line 184 and a negative power source line 185 are connected to regions 186 to 191 of silicon thin film forming the source/drain and channel regions of a P type TFT and to regions 192 to 195 of silicon thin film forming the source/drain and channel regions of an N type TFT. The elements of each TFT can be separated by etching the silicon thin film to form isolated areas regardless of homo-polarity and hetero-polarity. Therefore, for instance, a distance <u>a</u> between the region 192 of silicon thin film of the N type TFT and region 187 of silicon thin film of the P type TFT and a distance <u>b</u> between the two regions 187 and 188 of silicon thin film of the P type TFT can be made substantially equal to each other. The number of elements in the direction in which a unit cell is repeated can be increased by arranging an isolated area of silicon thin film of the P type TFT and an isolated area of silicon thin film of the N type TFT alternately, utilising the above characteristics positively.

0040In order to increase the number of elements further, the following means may also be employed. Figures 10(a) and 10(b) illustrate an inverter composed of complementary TFTs and formed between a positive power source line 199 and a negative power source line 200. In Figures 10(a) and 10(b), through holes 201 and 202 are provided for forming contact between a source region and a gate electrode 203. At first, a P type region 204 and an N type region 205 are provided in an isolated area of silicon thin film with a boundary 208. Next, a contact with the drain region is formed by a through hole 206 and an output line 207 for the inverter is added.

0041A second modification concerns the reduction of clock noise in the source line drive circuit. As shown in Figure 1, the source line drive circuit 12 is provided with video signal buses 14 to 16 and a line (connected to the terminal 34) for transmitting at least a pair of relatively inverted clock signals CL and <maths id="math0006"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0006.tif" /></maths> for driving the shift register 13. The line shown connected to the terminal 34 is in fact a pair of lines and if there is a difference between the stray capacitance formed between any video signal bus and the CL line and that formed between the same video signal bus and the <maths id="math0007"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0007.tif" /></maths> line, noise in the form of a spike synchronised with the clock signal is added to the video signal. As a result, unevenness occurs in the display in the form of a line in the picture produced by the active matrix panels.

0042As shown in Figure 11(a), the above described clock noise may be reduced by a twisted arrangement for the CL and <maths id="math0008"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0008.tif" /></maths> lines. Figure 11(a) illustrates a source line drive circuit comprising a shift register having unit cells 210 to 213, a sample hold circuit comprising TFTs 214, 215, a picture element matrix 216, and a video signal bus 217. CL and <maths id="math0009"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0009.tif" /></maths> lines 218 and 219 cross each other (i.e. are twisted) at their centres approximately. Hence, the average distance between the CL line 218 and the video signal bus 217 and the average distance between the <maths id="math0010"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0010.tif" /></maths> line 219 and the video signal bus 217 are made to be approximately equal to each other. As a result, the value for the stray capacitance (CS1 + CS3) which is formed between the CL line 218 and the video signal bus 217 is made to be equal to the value for the <u>stray</u> capacitance (CS2 + CS4) which is formed between the CL line 219 and the video signal bus 217. Further, as shown in Figure 11(b), the rising edge of the clock signal CL corresponds to the trailing edge of the clock signal <maths id="math0011"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0011.tif" /></maths>, and vice versa. Consequently, clock noise added to the video signal can be reduced sharply to realise a clear display in the picture. Further, the CL and <maths id="math0012"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0012.tif" /></maths> lines may be twisted (cross over) a plural number of times.

0043A third modification seeks to make the resistance of the connections to the sample hold circuit uniform. Figure 12 illustrates the parts of a source line drive circuit such as the one shown in Figure 1, including a shift register 230, video signal buses 231 to 233, a sample hold circuit comprising TFTs 234 to 236, and a picture element matrix 240. Picture element signals corresponding, for example, to the primary colours red (R), green (G) and blue (B), are transmitted to the three video signal buses 231 to 233 and then the combination is changed by a single horizontal scan. Since these three video signal buses require a low resistance, a metal such as aluminium is preferably used for each line. On the other hand, when the structure of Figures 3(a) or 3(b) which is considered to be the most effective from an economical point of view is adopted, the same material as that employed for the gate electrodes 80 and 81, such as polycrystalline thin film, is used for lines 237 to 239 from the video signal buses to the sample hold circuit. In this case, the sheet resistance of poly-crystalline silicon thin film is much higher than the resistance of the video signal buses, and the lengths of the lines 237, 238 and 239 cannot be equal to each other if these lines are straight, whereby the resistance of the lines 237 to 239 also will not be equal to each other. As a result, the picture displayed may be uneven and include lines. However the lines 237, 238 and 239 are formed such that their resistances are made equal to each other. For example, the line width W may be kept constant and the line length L of each line 237 to 239 may be made equal, or the line length L of each line may differ and the line width W may vary respectively between the lines 237 to 239.

0044A fourth modification is directed towards increasing the slow speed of a drive circuit composed of TFTs. As shown in Figure 5, since the performance of a TFT is inferior to the performance of mono-crystalline silicon MOSFET, the speed of a shift register made from TFTs is not fast enough to drive an active matrix panel. In order to make up for this slow speed, the circuit structure as shown in Figure 13(a) and operation as shown in Figure 13(b) may be utilised.

0045In Figure 13(a), a start signal DX and clock signals CL x 1 and <maths id="math0013"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL x 1</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0013.tif" /></maths> are applied to a first shift register 250 included in a source line drive circuit for generating sampling pulses 252, 254 and so on. Further, a start signal DX and clock signals CL x 2 and <maths id="math0014"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL x 2</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0014.tif" /></maths> are applied to a second shift register 251 included in the source line drive circuit for generating sampling pulses 253, 255 and so on. A video signal bus 265 is connected by way of TFTs 256 to 259 of a sample hold circuit and source lines 261 to 264 to a picture element matrix 260. The signals V, DX, CL x 1, <maths id="math0015"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL x 1</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0015.tif" /></maths>, CL x 2, <maths id="math0016"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL x 2</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0016.tif" /></maths>, and the pulses 252 to 255 output from the shift registers 250 and 251 are shown in Figure 13(b).

0046The source line drive circuit shown in Figure 13(a) is provided with the two shift registers 250 and 251, which are driven by clock signals CL x 1 (<maths id="math0017"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL x 1</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0017.tif" /></maths>) and CL x 2 (<maths id="math0018"><math display="inline"><mrow><mover accent="true"><mrow><mtext>CL x 2</mtext></mrow><mo>¯</mo></mover></mrow></math><img file="EP0806701A1_D0018.tif" /></maths>) which are approximately 90° out of phase. When the source line drive circuit is provided with a number N of shift registers, each shift register should be driven by a respective clock signal and the inverted clock signal wherein the phases lag by approximately 180°/N. If the frequency of CL x 1 and CL x 2 is taken as <u>f</u>, the sampling pulses 252 to 255 are generated sequentially at intervals of 1/4f hours. The video signal V is sampled at the trailing edges 266 to 269 of the sampling pulses and is held at the source lines 261 to 264. As a result, sampling with a frequency 4f is realised by means of shift registers which are driven by a clock signal of frequency <u>f</u>, which results in an effective compensation for the slow speed of a shift register made up of TFTs. When the above described source line drive circuit is provided with a number N of shift registers, sampling with a frequency 2Nf can be realised by means of shift registers which are driven by a clock signal of frequency <u>f</u>.

0047A fifth modification provides means for testing each output from the source line and gate line drive circuit. Figure 14 illustrates this modification, which includes a shift register 280 as part of the source line drive circuit, a video signal bus terminal 281, a sample and hold circuit comprising a TFT 282, a source line drive test circuit 283, a control terminal for the test circuit 283, a test signal output terminal 285 and a source line 286. A respective test circuit 283 is added to each source line 286. Likewise, a shift register 287 forming part of a gate line drive circuit is connected to a gate line test circuit 288 having a test signal input terminal 289 and a test signal output terminal 290. A gate line 291 leads from the gate line drive circuit to a picture element matrix 292. A respective test circuit 288 is added to each gate line 291.

0048The test circuits operate as follows:

0049The test circuit 283 is turned ON by way of the terminal 284 for testing operation of the source line drive circuit. In this condition, the shift register 280 is scanned after a predetermined test signal is supplied to the video signal bus terminal 281. At this time, if a signal which meets standard requirements is serially output at the test output terminal 285, the result is considered as "Good" and, if not, it is considered as "Poor". For testing the gate line drive circuit, the shift register 287 is likewise scanned when a predetermined test signal is input. At this time, if a signal which meets standard requirements is serially output at the test output terminal 290, the gate line drive circuit is considered as "Good" and, if not, it is considered as "Poor". In this way, electrical testing of an active matrix panel, can be done automatically, whereas in the prior art testing has been by visual observation of a displayed test pattern.

0050A sixth modification involves providing a storage capacitor in a picture element without an additional process during manufacture. Figure 15(a) illustrates the circuit of such a picture element, and Figure 15(b) is a cross section through the picture element. As shown in Figure 15(a), the picture element comprises a TFT 302, which operates as a switch and which is connected to a source line 300 and a gate line 301, a liquid crystal cell 303, a counter electrode terminal 304, a metal oxide semi-conductor capacitor 305 (hereinafter referred to as a MOS capacitor) having a gate electrode 306. Turning to Figure 15(b), transparent insulating substrates 310 and 324 have sandwiched therebetween silicon thin film regions 311 to 315, gate insulating film regions 316 and 317, gate electrodes 318 and 319, an insulating film layer 320, a line layer 321 which forms a source line, a transparent conductive film layer 322 which forms a picture element electrode, a counter electrode 323 including a transparent conductive film layer, and a layer of liquid crystal material 325.

0051The TFT 302 formed by the portion indicated by the reference numeral 326, the regions 311 and 313 providing the source/drain regions thereof and the region 312 forming the channel region. The MOS capacitor 305 is formed by the portion indicated by the reference numeral 327, with the regions 313 and 315 providing the source/drain regions thereof and the region 314 forming the channel region thereof.

0052As shown clearly in Figure 15 (b), since the MOS capacitor 305 has entirely the same cross sectional structure as the TFT 302, It is not necessary to add any special extra manufacturing process in order to form the MOS capacitor 305. However, If the MOS capacitor 305 is to be used as a storage capacitor, it is necessary to maintain a condition wherein a channel, that is an inversion layer, is formed at the region 314. A predetermined voltage potential is applied to the gate electrode 306 (319) of the MOS capacitor 305 so as to turn the latter on for maintaining this condition. For example, a positive potential in the case of an N type MOS capacitor or a negative potential in the case of a P type MOS capacitor is appropriate for this. Since a gate insulating film is normally formed to be extremely thin, a storage capacitor produced as described above can be from five to ten times as large as a storage capacitor making use of a conventional insulating film layer. This is very effective for economising on the area employed by the storage capacitor. Therefore, it becomes possible to increase the aperture ratio of the active matrix panel.

0053A last modification concerns the mounting of an active matrix panel having a built-in drive circuit. Figure 16(a) is a cross section through an arrangement for the mounting, which features a transparent substrate 330 wherein a picture element matrix and a drive circuit composed of TFTs are formed, a transparent substrate 331 wherein a counter electrode is formed, a sealing member 334, a filling of liquid crystal material 333, a mounting substrate 335 having an aperture 340, a wire 338 made of a metal such as gold or aluminium, and a protective member 339. A recess 336 in the mounting substrate 335 adjacent the aperture 340 is effective for ensuring a strong connection by the wire 338. A shading member 337 is advantageously provided over all or a part of the mounting substrate 335, and a further shading member 332 in the form of a strip surrounding a portion of the picture element matrix is advantageously provided on the transparent substrate 330 for improving the external appearance of the active matrix panel when used as a display device. Figure 16(b) is a plan view of the arrangement showing the picture element matrix 341 supported by the mounting substrate 335 over the aperture, which is indicated by the dotted line 342.

0054Since the stress on the wire 338 is uniform with this arrangement, the strength of the connection is improved. Further, when the active matrix panel in accordance with this embodiment is utilised as a transmission type display device and a light source is provided at its rear face, the leakage of unnecessary light from the periphery of the picture element matrix can be prevented and an improved external appearance is obtained.

0055Two applications of the invention will now be described.

0056A first application of the invention is in an electric view finder (hereinafter referred to as an EVF) of a video camera or the like. By integrating a drive circuit made of complementary TFTs with the picture element matrix as described above, a small, inexpensive and reliable active matrix panel having a low power consumption and a high resolution density can be obtained. As a result an EVF having the structure shown in Figure 17 can be realised. This EVF 353 is connected to an image sensing device 350, recording apparatus 352, and a video signal processing circuit 351 arranged to supply a composite video signal on a line 362. The EVF 353 comprises a drive circuit 354, including a chroma circuit, a synchronised timing signal formation circuit, a liquid crystal panel driving signal formation circuit, a power source and a back light driving circuit. The EVF further comprises a light source 356 for providing the back light, a reflector 355, a diffuser 357, polarisers 358 and 360, an active matrix panel 359 according to the invention and a lens 361.

0057Such a structure offers the following benefits, which have not been obtained by an EVF employing a conventional cathode ray tube. <ul id="ul0008" list-style="none" compact="compact"><li>(1) A colour EVF having an extremely high resolution with a picture element spacing of 50 microns and less can be produced if the active matrix panel is provided with a colour filter. Low power consumption is also promoted.</li><li>(2) An extremely small, light weight, space saving design of EVF can be realised.</li><li>(3) The shape of the EVF can be more freely chosen, and even a novel design in which the EVF is flat is possible.</li></ul>

0058Another application is in a colour display projector, in which the active matrix panel is employed as a liquid crystal light valve. Figure 18 is a plan view of such a colour display projector. White light emitted from a light source 370, such as a halogen lamp, is condensed by a parabolic mirror 371 and is filtered to cut out heat rays in the infra-red region by heat ray cutting filter 372 so that only light rays in the visible spectrum enter the dichroic mirror system of the projector. First, blue light (light having a wavelength of about 500 nm) is reflected by a blue light reflection dichroic mirror 373, other light (e.g. yellow light) being transmitted therethrough. The reflected blue light is directed by a reflection mirror 374 into a blue light modulation liquid crystal light valve 378. The light transmitted through the blue light reflection dichroic mirror 373 strikes a green light reflection dichroic mirror 375, and green light (light having a wavelength of about 500 nm to 600 nm) is reflected while the other light, that is red light (light having a wavelength of about 600 nm and longer), is transmitted therethrough. The reflected green light enters a green light modulation liquid crystal light valve 379. The red light transmitted through the green light reflection dichroic mirror 375 is directed by reflection mirrors 376 and 377 into a red light modulation liquid crystal valve 380.

0059The blue light, green light and red light are modulated by the liquid crystal light valves 378, 379 and 380, each of which comprises an active matrix panel according to the invention driven by a primary colour signal, such as blue, green and red, and are then synthesized by a dichroic prism 383. The dichroic prism 383 is so constructed that a blue light reflection surface 381 and a red light reflection surface 382 cross one another at right angles. The thus synthesised colour image is magnified and projected onto a screen by a projection lens 384.

0060Such a structure offers the following benefits, which have not been obtained by a colour display projector employing a conventional cathode ray tube as a projection tube. <ul id="ul0009" list-style="none" compact="compact"><li>(1) A projection lens with a small aperture can be used for the projection lens 384 because the liquid crystal light valve can be made smaller and of higher resolution than is possible with a cathode ray tube. Therefore, a small, light weight and inexpensive projector can be realised.</li><li>(2) Having a high aperture ratio, the active matrix panel of the invention permits a light display even if a projection lens with a small aperture is utilized.</li><li>(3) Good registration of the three colours can be obtained because the optical axes of the red, green and blue light valves are fully aligned relative to each other by the dichroic mirrors and dichroic prisms.</li></ul>

0061The present invention, therefore, provides a number of benefits and advantages.

0062Firstly, the following advantages can be obtained by forming the gate line/source line drive circuit made of complementary TFTs on the same transparent substrate as the picture element matrix in accordance with the invention. <ul id="ul0010" list-style="none" compact="compact"><li>(1) In the prior art, the degree of resolution is limited by the mounting of the drive integrated circuit. This limitation is removed by the built-in drive integrated circuit of the invention. As a result, a liquid crystal panel having a picture element spacing of 500 microns and less can be obtained in accordance with the invention.</li><li>(2) The external dimension of the mounting substrate of the matrix panel can be miniaturised so that a display device employing a liquid crystal panel according to the invention can be made smaller, thinner and lighter.</li><li>(3) Since the step of attaching an external drive integrated circuit is unnecessary, a low cost display device can be obtained.</li><li>(4) Since the connection of an external drive integrated circuit is not required, the reliability of the display device can be improved.</li><li>(5) The low current consumption which is an intrinsic feature of the present active matrix panel obtained by forming the drive circuit with complementary TFTs results in low current consumption in the whole display device. This is an important feature, which makes the application of the invention in an EVF for a video camera and a portable image monitor feasible.</li></ul>

0063Secondly, not only is low power consumption achieved but also the voltage range and operating frequency range can be extended by the use of complementary TFTs and a circuit structure with a static type of shift register. A TFT has a high OFF current characteristic as shown in Figure 5 and the temperature dependency of the OFF current is also large. This defect of the TFT is compensated by a static type of shift register, and so the voltage range and frequency range can be expanded.

0064Thirdly, due to the structure wherein first polar impurities are included in the first polar TFT source/drain region and both first and second polar impurities, with the latter having a higher density than the former, are included in the second polar TFT source/drain region, an inexpensive complementary TFT integrated circuit can be obtained by the addition of a single photo process to the conventional process for manufacturing a known mono-conductive type TFT. Further, P type and N type TFTs with well balanced performances can thus be obtained.

0065Fourthly, the gate which forms the drive circuit is formed to be shorter than that of the TFT which forms the picture element matrix, so that the actuating speed of the drive circuit can be improved and the writing and holding of electric charge in each picture element can be maintained in the optimum condition.

0066Further, the various modifications described above provide the following beneficial effects.

0067Firstly, the pattern layout as shown in Figures 8, 9 10(a) and 10(b) enhances the arrangement within the drive circuit(s) so that unit cells can be formed within a defined pitch corresponding to the spacing of the picture elements.

0068Secondly, clock noise in the video signals can be removed to the point of rendering any corresponding unevenness or unwanted line in the picture displayed below the level of visibility.

0069Thirdly, the resistances of the connections to the sample and hold circuit are made uniform along each source line so that the writing level of the display signal to all the source lines is rendered uniform, which avoids unevenness in the form of a line in the picture display.

0070Fourthly, the source line drive circuit as shown in Figure 13 (a) and driven by the signals as shown in Figure 13 (b) allow sampling of the video signal with a frequency 2Nf by means of a system of N shift registers which are driven by a clock signal having a frequency <u>f</u>. Therefore, an active matrix panel having a built-in drive circuit of high resolution can be obtained even when TFTs, whose ON current is not necessarily large enough, are employed.

0071Fifthly, by providing a test circuit for the output of each drive circuit, checking the active matrix panel, which has previously been carried out by visual examination of a display test pattern, can be carried out electrically and automatically.

0072Sixthly, the structure for a storage capacitor as shown in Figures 15 (a) and 15 (b) may be employed in each picture element to hold the electric charge in each picture element more steadily but without either increasing the cost of production or decreasing the aperture ratio.

0073Seventhly, the mounting structure as shown in Figures 16 (a) and 16 (b) not only improves the connection strength and the reliability of the panel but also prevents the leakage of unnecessary light around the periphery of the picture element matrix in the event that back lighting is used with the active matrix panel of the invention to construct a transparent display device.

0074Finally, the application of the invention to the particular display systems described has the following advantages.

0075Firstly, effects which cannot be obtained by the use of a conventional cathode ray tube in an EVF can be obtained by the application of the invention to the EVF of a video camera. <ul id="ul0011" list-style="none" compact="compact"><li>(1) An extremely high resolution colour EVF with a picture element spacing of 50 microns or less can be realized by the use of an active matrix panel which is provided with a colour filter. Further, low power consumption results.</li><li>(2) An extremely small and light weight EVF can be realised in a space saving manner.</li><li>(3) The shape of the EVF can be selected more freely so that a novel design, for example a flat EVF, can be provided.</li></ul>

0076Secondly, effects which cannot be obtained by the use of a conventional cathode ray tube can be obtained by the application of the invention to a colour display projector. <ul id="ul0012" list-style="none" compact="compact"><li>(1) Since the liquid crystal valve can be made smaller and of higher resolution than in the case of the cathode ray tube, a projection lens with a small aperture can be used. Therefore, a small, light weight and inexpensive colour display projector can be realized.</li><li>(2) Due to the high aperture ratio of the active matrix panel, light display can be obtained even with a projection lens with a small aperture.</li><li>(3) Unlike the projection tube employing the cathode ray tube, the optical axes of the light valves for each colour, namely red, green and blue, can be made to coincide by the dichroic mirrors and dichroic prisms so that the registration of the three colours can be achieved satisfactorily.</li></ul>