Reduced noise touch screen

IS 2 173 136 T3

DESCRIPTION

Low noise touch screen device.

The present invention is directed to a touch screen user input device for computers and, in particular, to a touch screen that has reduced noise while providing high resolution.

Background of the invention

Computer touch screens allow the user to record or extract information on a computer screen or to select from various regions of a computer generated presentation, typically by means of the user's fingers or by means of a free or linked plotter. The cost of touchscreens depends in part on the desired resolution. When the application requires only a rough resolution (such as determining, to within about two inches or more, the position of the finger or tracer on the screen) there is typically only moderate cost involved. However, when a fine resolution is required (that is, a resolution of less than about an inch (1 inch = 2.54 cm), preferably less than 1/2 inch, more preferably less than 1/4 inch, and more preferably less than 1/8 inch), the cost of the touch screens offered according to previous solutions has been relatively high. One reason for the cost is that additional hardware or methods were needed to distinguish from noise (ie, signals that are not indicative of finger or tracer position) small signals that represent high-resolution positions or movements. Many prior devices have also been relatively expensive to make linear (ie, to obtain an indication of the location of the finger or tracer based on the output signal of the device). Typically, the position is a non-linear function of the output signal. The linearization process, in some earlier devices, is believed to involve manually modifying the size or shape of the screen electrodes. As this must be done individually for each screen coating that is manufactured, the cost can be high.

Another factor contributing to the high cost of high-resolution touchscreens is the need, in some processes, to provide electrodes or other layers deposited on both front and back surfaces, and / or to provide two or more independent process steps in the process. empty. Also, when the final product is made by attaching or attaching a separate coating to a computer screen (typically a cathode ray tube or CRT), this additional step adds more cost to the final product.

Consequently, it will be useful to provide a touch screen that offers fine resolution while reducing or eliminating noise at low or low cost.

From US-A-4,853,498 a position measuring apparatus for capacitive touch panel systems is known. The touch screen has a transparent conductive coating with four electrodes located along the respective edges of the screen. The position in which said screen is touched is detected by measuring the current drawn through the electrodes when the tracer or finger makes contact with the conductive layer, thus forming a directional signal indicative of the position. US-A-4,853,498 discloses the preamble to claim 1.

From EP-A-0199243, a method for automatically calibrating a touch screen is known. This method is executed before each interactive session. By means of áel, three points are presented, which define perpendicular axes, and the operator is instructed to touch the touch screen at each of these points successively. The coordinates of these three points are checked and if the checks are fulfilled, a translation matrix is ​​generated to convert the coordinates of the touch screen into coordinates of the presentation.

Summary of the invention

The present invention provides an apparatus as defined in claim 1. In order to offer a reduction or elimination of noise, which is particularly annoying in fine resolution devices, a relatively high sampling rate is provided in conjunction with a filter. distinguishing noise (typically low frequency). Furthermore, one embodiment of the invention can distinguish the desired signal from spurious signals, such as a ground or static discharge, using automatic gain control.

Signal linearization includes a measurement of the touch signals at different known locations on the screen and a conversion of the output signal at the location obtained from the finger or tracer using the measured data, thus eliminating the need for manual or manual switching. individually or by “tuning” the electrodes or other features of the screen.

It is also possible to place the electrodes directly on the CRT or another computer screen. "Directly" means that the electrodes are in contact with the CRT, rather than on a coating that is later bonded to the screen. One embodiment allows the electrodes to be positioned in such a way as to eliminate the electrodes from the back surface or the layers, thus eliminating one or more vacuum process steps, preferably requiring only one vacuum process step to manufacture a computer screen such as touch screen.

Touch screens according to the present invention can be used for various purposes. One of the purposes involves the use of an electronic gaming machine, such as a slot machine, an electronic lottery machine, and the like. Other uses include common computing, such as a personal computer, a laptop computer, a pocket computer, a computer without a keyboard, a personal communication device, a telephone, interactive television and the like, the execution of software. such as word processing, spreadsheet, communications, databases, programming, networking, and other popular software.

ES 2

The invention may be used in connection with user software or with computer operating systems designed for writing computing, such as Penpoint® from GO Corporation, Windows for Pen Computing® from Microsoft Corporation, or with operating systems or other software intended for use with a pointing device, such as a mouse, rolling ball, joystick, and the like.

Brief description of the drawings

Figure 1 is a schematic front view of a touch screen in which an apparatus according to an embodiment of the present invention may be used;

Figure 2 is a flow chart of a prior production method of a touch screen;

Figure 3 is a flow chart of a method for providing a touch screen in which an apparatus according to an embodiment of the present invention may be used;

Figure 4 is a schematic diagram of a touch screen signal processing apparatus in accordance with one embodiment of the present invention;

Figure 5 is a schematic diagram of an analog signal processing apparatus according to an embodiment of the present invention;

Figure 6 is a block diagram of a touch screen and its associated treatment system, comprising an apparatus according to an embodiment of the present invention; ^Y

Figure 7 is a flow chart of a calibration and linearization procedure.

Detailed description of the preferred embodiment

As illustrated in FIG. 1, a computer display device, such as a CRT, is provided configured so that electrodes 102a-102d overlap at perimeter regions of the screen. Various configurations of the electrodes 102a, 102b, 102c, 102d are possible. In the embodiment shown in Figure 1, the electrodes 102a, 102b, 102c, 102d are in the form of conductive distribution bars located along most of each of the four edges of the screen 104, preferably without extend into the corner regions of the screen. Lead wires 106a, 106b, 106c, 106d are in electrical contact with each of the electrodes 102a, 102b, 102c, 102d to provide a communication channel with the circuitry described later.

Aspects of the present invention are best understood in the context of prior methods of providing a touch screen. Figure 2 shows a method used in connection with the manufacture of a touch screen according to previous devices. As illustrated in Figure 2, previous devices provided touch screens by placing certain items on a glass coating that was then attached to the front surface of a computer screen, such as a CRT,

136 T3 4 for example, using a transparent adhesive. In earlier processes, the glass coating first had a conductive coating applied to its back surface, 212 (ie, the surface that had to contact or be bonded to the surface of the CRT). This coating was provided, topically, by means of an evaporation process, usually in a multi-furnace. Next, a similar conductive coating was placed on the front of the glass coating also by means of a process, 214, of evaporation. After these steps, an additional layer was placed on the back surface of the glass coating. In some cases, a conductive layer was provided on the back surface. In most cases, it was necessary to place an electrode on the posterior surface, often using a screen printing procedure. The back electrode is believed to have been offered in earlier devices as part of a noise reduction technique, that is, a technique to distinguish the desired position signal from unwanted signals or noise. The electrodes were then placed on the front surface of the glass, topically by means of a process, 218, of screen printing. At this time, it is believed that prior techniques have included a step of modifying or changing the shape of the electrodes in order to help make the device linear, 220. This is believed to have been done on an individual basis in order to "tune" each screen individually so as to provide the coating with a known or linear electric field. This tuning is believed to have involved measuring the electrical characteristics of the screen and scraping or abrading parts of the electrodes in order to modify the measured characteristics to fit within predetermined parameters.

Finally, after "manual tuning", 220, a protective layer such as SiO₂, at least on the front surface of the screen, in a process carried out in a vacuum oven, 222. As can be seen, the process of Figure 2 includes several stages that are relatively expensive including evaporation processes that may require the treatment in a vacuum oven, 212, 214, for the placement of at least one deposited layer, such as an electrode on the posterior surface, 216, and manual or individual tuning of the electrodes, 220.

Figure 3 depicts a process for providing a touch screen device. In the embodiment of Figure 3, electrodes are placed directly on the glass front surface of a computer screen such as a CRT. Although the process in Figure 3 could also be used to fabricate a glass or other transparent coating for a CRT by placing the electrodes directly on the CRT, it is no longer necessary to include a step for bonding or placing a coating on a CRT. . Topically, when providing a coating, a joint is used rather than a mere placement on the CRT, as the joint helps provide strength and resistance.

ES 2 173 136 T3 necessary strength and also provides the desirable optical qualities. However, by directly placing the electrodes on the CRT, as illustrated in Figure 3, the cost of the bonding (positioning) step is eliminated.

According to the process of Figure 3, the electrodes (such as the electrodes shown in Figure 1) are screen printed directly on the front of the computer screen such as a CRT, 312. After the screen printing step, a conductive coating on the electrodes and on the front surface of the screen. In addition, a protective coating is made, such as SiO₂, on the front of the screen. Preferably, both the conductive coating and the protective coating are applied in a single vacuum stage, 314, that is, without the need to remove the CRT from the vacuum oven between the conductive and protective coatings.

Various materials can be used in connection with the process depicted in Figure 3. Electrodes can be made from various conductive materials including silver and low-melting glass and silver epoxies. The conductive coating is preferably made using materials that can be provided in substantially transparent form, such as indium tin oxide or antimony tin oxide. In one embodiment, indium tin oxide is used since it can be applied in a single stage process.

Before describing the details of an embodiment of the present invention, a circuit for dealing with noise was described, generally as shown in Figure 4. Although the circuit of Figure 4 can be used in connection with various touch screen devices, it is particularly useful for touch screen manufactured according to the method of Figure 3, since this method does not involve an electrode on the back surface of the screen. the screen, which was a feature believed to be used by earlier devices in relation to noise reduction.

When the touchscreen is arranged with an electric potential at the electrodes 102a, 102b, 102c, 102d, under normal circumstances (without touching the screen), the potential at the electrodes will remain constant over time. When a human being touches a part of the screen, for example with a finger, a small amount of current will flow, such as about 5-10 μ amperes per volt of excitation potential, through the human body towards the ground. The apparatus of Figure 4 is oriented to provide a signal from which the value of the current from one of the electrodes through the human body to the ground can be measured in a relatively noise-free manner. The general method used to distinguish the desired signal from noise involves relatively high frequency sampling that allows filtering out the lower frequency modulation or "envelope" associated with noise. In order to deduce the amount of current flowing from a given electrode (and thus allow calculation of the distance from the electrode, as described later), it is useful to make some assumptions about the electrical characteristics of the human body at through which the current flows. To this end, it is assumed that the human body has the electrical characteristics corresponding to a body model 412 that has, in series, a first capacitor, a resistor, and a second capacitor. These elements are not circuit elements but are actually a model of certain characteristics of the human body. In current practice, element 412 will be the path through the human body from the point where the body touches the touch screen to ground 416. One of the Lines 106a of the touchscreen electrode 102a was connected to the positive input of an amplifier such as op amp 418. Preferably, the op amp has a relatively high gain, such as around 2,000 gain, and at a embodiment, it has a common mode rejection of 127 db. A resistor such as a 100 kilo ohm resistor 420 connects output node 422 to negative input terminal 424. An RC filter 426 is provided at the output of the operational amplifier 418 in order to reduce dV / dt. This eliminates filter transients and is useful for reducing emissions, for example in order to comply with regulations such as Federal Communications Commission (FCC) rules.

In one embodiment, the signal provided to electrode 102a (through resistors 430, 432) as well as that supplied to the negative input of operational amplifier 418 (through resistors 434, 436) has a generally sinusoidal shape provided at a frequency such as 10 kilohertz 438. Similar circuits are used to provide signals (as well as sample signals) to the other electrodes 102b, 102c, 102d, although the phases of the four signals are preferably 90 shifted.^or.

Noise rejection is provided by a band pass filter 440 that samples the amplified signal from the electrode at a relatively high frequency, for example about 100 kilohertz, preferably about 200 kilohertz, and preferably about 250 kilohertz or higher. . In one embodiment, the band pass filter 440 is a fourth order Butterworth filter with a gain of about 20 and a Q factor of about 40. A Butterworth filter offers the advantageous characteristic of fast attenuation from the peak and because of this shows a high performance-to-cost ratio. However, in some situations the Butterworth filter can be unstable and can create, for example, transient oscillations. Consequently the present invention can also be used with other types of filter. For example, a Bessel filter can be used which tends to be more stable than a Butterworth filter under transient conditions, but does not fade as rapidly. Another type of filter that can be used is an eloptic filter.

In the embodiment shown, both filter 440 and oscillator 438 are controlled by a clock signal such as a 20 MHz signal 442. In one embodiment, the filter

ES 2 provides a sampling rate (defining the time span over which filtering is applied) of at least three times the speed of clock signal 442, in one embodiment, a sampling rate of at least 40 kilohertz. Other devices can be used for high sampling rates and filtering, such as a fast-acting rectifier or an integrator.

The magnitude of the drive signal 438 is set in relation to a gain controller 446. One of the functions of the gain controller 446 is to handle a situation in which there is a rapid change in the environment, such as may result when the user touches a grounded metal object, thereby changing the effective electrical characteristics of the user's body 412. Gain controller 446 receives an indication of the magnitude of the signal received from the touch screen, preferably from information provided by CPU 450, described below. If any of the four signals becomes greater than a predetermined amount, for example greater than 4.5 volts, the system gain is reduced, decreasing the drive signal, ie, the output signal 452 of oscillator 438. In the embodiment of FIG. 5, drive signal 452 is attenuated by attenuation signals 516 that control multiplexer 518.

Figures 5 and 6 represent a way of carrying out the general configuration of Figure 4 in a system having four electrodes on a touch screen, as shown in Figure 1. Figure 5 shows an analog section 512 of circuits in which each of the leads 106a, 106b, 106c, 106d was connected to the positive input of an operational amplifier 418a, 418b, 418c, 418d. The output of the op amp, optionally through an RC filter, is connected to a band pass filter 440a, 440b, 440c, 440d. Under the control of the control signals mux 0, mux 1 512, a multiplexer 514 selects the outputs of the filters 440a, 440b, 440c, 440d to feed them, one by one, to an analog-to-digital converter 612 (figure 6) . As noted above, filters 440a, 440b, 440c, 440d are capable of very fast sampling and preferably all four channels are sampled and fed to an A / D converter 612 in less than 800 microseconds, preferably less than 400 microseconds and preferably less than 300 microseconds. In one embodiment, the system provides a digital sample of the filtered analog signal at a rate of about 1,000 samples per second (for each of the four electrodes), preferably about 2,500 samples per second, more preferably about 3,000. samples per second or more. This rapid sampling is used to reduce or eliminate the effects of noise and variation between samples. The analog-to-digital converter is, according to one embodiment, a 12-bit tracking analog-to-digital converter.

As shown in figure 6, the central processing unit (CPU) is controlled by a crystal oscillator 442 which also, by means of dividers 614a, 614b, 614c, 614d, provides

136 T3 8 clock signals 441 and 552 at 200 kHz and 10 kHz, respectively. A power supply 616 provides power to the CPU, preferably through power management circuit 614, preferably configured to reset the CPU 450 if V_DC drops below 4.75 volts. Coupled to the CPU is a communications circuit 620 which may include, for example, optical couplers to convert the TTL level to 20 µΑ drive. EPROM memory 622 is used to store utility programs and startup programs, for example, to download program memories. Flash memory 624 is used to store the main program and is preferably non-volatile memory, so that the content remains intact when the power goes out. Flash memory can be reloaded or altered through the communications device. In this way, the flash memory program can be downloaded from a remote location. The communications front end contains optical couplers for serial data input and output, and a global input for remote reset of the master unit. A CMOS memory is used for annotation and calculation purposes.

Several CPU 450 can be used in this connection, although preferably a CPU of the type used in a personal computer environment is preferable.

In operation, the user touches a portion of the screen, and the current flowing from each of the four conductors 102a, 102b, 102c, 102d through the body is a function of the distance from the touch to each of the conductors. , respectively. Preferably, the screen has a substantially equipotential surface. Therefore, if the user touched the screen precisely in the center, equidistant from the four electrodes 102a, 102b, 102c, 102d, the total current flow through the user's body would be approximately 8/4 microamperes, that is, 2 microamps. . For touches that are not in the center, that is, that are closer to some electrodes than others, the current from the closest electrodes will be greater than those from the more distant electrodes. Knowing the relationship between the amount of current and the position of the screen, the amount of current that flows through each of the electrodes and through the user can be used to deduce the position of the touch, as fully described later. Returning to the example of a touch in the center of the screen, as represented in figure 4, the current of the touch is measured through a resistance of 11.3 ohms, so that, converted to voltage, the current becomes at 22.6 microvolts per volt of excitation voltage. Amplifier 418 multiplies the voltage by 2,000 and band-pass filter 440 adds a gain of 20. As shown in figure 5, the signal selected by multiplexer 514 is fed to an intermediate amplifier BA 516 which, in the embodiment shown, has a gain of 2. Therefore, as a whole, the system has a profit of 80,000. The signal is fed to the analog to digital converter. The order for

ES 2 173 136 T3 converting the analog signal into digital form is synchronized with the output of the drive oscillator. The timing is configured such that, when the digital data is obtained, it represents the peak voltage of the touch signal 5 for that channel. Several samples (eg, four samples per channel) are taken to average and eliminate any DC values ​​on which the sine wave may be mounted. 10

The digitized voltages of the four electrodes 102a, 102b, 102c, 102d represent the voltages of the left, top, right, and bottom electrodes, respectively, and are designated V_L, V_T, V_R, and VB. The digitally represented values ​​of these four voltages are combined to give two values ​​of tension Y_V and X_V, which represent values ​​in a vertical direction (Y_V) and in a horizontal direction (X_V) as follows: 20

Yv =

Vt - Vb

Vt + Vb (1)

Xv =

Vl - V_R

Vr + Vl (2)

Y_V and X_V They are related to the vertical and horizontal position of the touch on the screen, but the relationship, in general, is not linear. In order to obtain the actual position based on the values ​​of Y_V and X_V, a conversion or linearization procedure is used.

In general, the linearization procedure 35 is a procedure for converting from one coordinate system to another coordinate system. However, the conversion is general, not linear, and may be different for each particular touch screen, as well as possibly changing over time due to aging, changes in your environment, and so on. Various measurements are taken of the values ​​of Y_V and X_V in various known places on the screen. These measurements are then used to determine the 45 parameters of a conversion method to convert the values ​​of Y_V and X_V in values ​​that indicate the situation of the touch on the screen. In general, although it is desirable to provide a method for converting from the Y coordinate system_V, 50

X_V In a location coordinate system of the location on the screen, the method begins by converting in the opposite direction, that is, defining certain locations in the location coordinate system on the screen and measuring the values ​​of Y_V and X_V, which result from touches in predetermined places. These values ​​are then used to calculate the parameters that can be used to convert the coordinate systems in the opposite direction, that is, 60 from the Y coordinate system._V, X_V in the place of the location coordinates on the screen.

Various methods can be used for conversion between coordinate systems. In general, conversion systems that have high accuracy require a high level of computational resources and / or time. However, it has been found that the conversion method described below has only moderate computational requirements, but results in a relatively high precision or fine resolution system in which the touch location calculated by the system was within a distance default of the actual touch situation. In one embodiment, the calculated location is within about 1/8 inch of the actual location.

According to one embodiment, a pseudo-linear conversion method is used, using the following conversion equations:

Xs = ao + a₄Xv + a₂Yv + a₃Xv Yv (3)

Y_s = bo + biYv + b₂Xv + b₃Xv Yv (4) where

X_S is the horizontal distance calculated from the left edge of the screen

Y_S is the vertical distance calculated from the top edge of the screen.

Other conversion methods can be used including higher order conversion . However, the method of equations 3 and 4 has been found to provide acceptably high resolution, at least for parts of the screen that do not include the corners of the screen.

In the method depicted in Figure 7, once the device has been connected, 912, the system determines if this is the first connection made in this system, 914 (for example, by checking a flag). If so, the system performs a calibration procedure, 916, described below. Otherwise, various setup functions are performed, 918 such as clearing records, checking memories, enabling interrupts, and establishing communications. The system then determines if there is already a valid calibration stored in the system, 920. If not, the system then starts the calibration procedure, 916. Otherwise, the system continued to set the signal gain to a high level. , 922, for example using the gain controller, 446. In the next step, the system performs a check to determine if the screen was being touched. In the embodiment represented in figure 7, this is determined by checking if any of the four channels 106a, 106b, 106c, 106d indicates a voltage that is greater than twice the ambient voltage, that is, the voltage that is in the four electrodes 102a, 102b, 102c, 102d in the absence of touches. If not, the system continued to check for a touch, 924, until a touch is detected. In one embodiment, the determination of the occurrence of a touch is done by detecting whether there is a peak of a signal sampled by the filter that is above a threshold level. In the case where the sampled signal is represented by the following equation:

S = A (wt + φ) + T sin (wt + φ) (5) where

ES 2

S represents the sampled signal;

A is the amplitude of the sampled signal;

ω is the phase angle;

φ is the phase of the sampled signal;

T is the amplitude of the current flowing through the tap device the peak (P) of the sampled equation is a function of an equation that has the form:

P = (A² + B²)^1/2 (6)

Upon detecting a touch, the system checks to ensure that none of the channels 106a, 106b, 106c, 106d indicate a voltage that is greater than a predetermined amount, such as 4.5 volts, 926. As described above , this check is done in order to detect sudden changes in the environment, such as a user touching a grounded metal object or the like. If such a large voltage is detected, the drive voltage is reduced, 928, for example using the gain controller, 446, and the system returns to a test state to see if there is a touch on the screen.

If the voltage is within predetermined parameters, the system takes a predetermined number of samples, for example, four samples on each of the four channels, 930, using the high-frequency sampling bandpass filter, the A / D and other circuits represented in figures 5 and 6. The ambient voltages are subtracted, 932, and the Y values_v and X_v are calculated according to equations (1) and (2), 934. The Y values_v and X_v are then converted into X values_S and Y_S, 936, using equations (3) and (4). Y position values_S and X_S they are then sent to the central computer over communication circuit 620. Normally, the system would be configured to monitor the touches continuously and thus the system would be configured to obtain another location, 940, returning to the state in which the system determines if it is doing a touch on the screen, 924. Otherwise, the routine ends, 942.

When performing a calibration, 916, the system first determines whether the monitor has warmed up sufficiently, 950, and cycles successive ten-second timeouts, 952, until warm-up is complete. Next, the CPU 450 sends a request, through the communication device 620, to the central computer, requiring the central computer to present the instructions to the user of the touch screen, 954. During the time that the user is not touching the touchscreen, the computer measures the voltages in the four channels 106a, 106b, 106c, 106d and stores these values, for example, for use in steps 924 and 932. The CPU 450 then requires the host computer to present a touch point (a point or other indicia on the touch screen at a predetermined location). The computer also presents

136 T3 12 instructions that tell the user to touch the screen precisely in the place of the point presented, 958. The system determines if there is a touch, 960, and if the voltage is within the defined limits 962, 964, as described above for steps 924, 926 and 928. Ambient voltages are subtracted, 966, and the measured values ​​are stored, 968. The CPU 450 then requires the host computer to place the next point on the screen, 970, in the next predetermined location, unless all points have been displayed, 972. In one embodiment, a total of nine points are used. touch or calibration points, preferably defining four substantially rectangular and identical quadrants or regions.

Once all the touch points have been presented and the stress values ​​of Y_v and X_v have been stored, corresponding to each of the touch points, the system calculates the parameters at₀, to₁, to₂, to₃, b₀, b₁, b₂, b₃, which will be used in equations (3) and (4), 974. This can be done by solving or adjusting equations (3) and (4) for each of the variables to₀, to₁, to₂, to₃, b₀, b₁, b₂, b₃. Since nine measurements have been made, this would result in a system of 18 equations and eight unknowns and would therefore be overdetermined. Consequently, a “best fit” method is used to determine the values ​​of the unknowns a0, a₄, to₂, to₃, b₀, b₁, b₂, b₃ which provides the best fit with the nine measured values ​​of X_v and Y_v. These values ​​of the unknowns are stored later for use in the calculation of X_s and Y_sfrom step 936. The system then returns to step 918, described above.

In light of the above description, it will be apparent to those skilled in the art that the present invention provides several advantages. The present invention is capable of determining the place of a touch on a touch screen with a fine resolution and at a relatively low cost and high speed. The present invention reduces costs by eliminating steps, such as the steps of providing a back electrode and / or a conductive back coating, and reduces the requirement for the number of steps that must be performed in independent vacuum chamber processes. The system provides data linearization to determine the location of a touch with a high degree of precision without the need for individual manual linearization of touchscreen devices.

Various variations and modifications can be used. It is possible to use some aspects of the invention without using the others. For example, it is possible to use the simplified production technique of the display of the present invention without using the linearization of Figure 9 and / or the circuits of Figures 5 and 6. In addition to being used in connection with gaming machines, PCs and other devices as described above, the present invention can also be used in connection with other interactive applications, such as location or navigation devices, automated teller machines (ATMs), in connection with control devices for vehicles and / or machines and the like.

IS 2 173 136 T3

Although the application has been described in the form of a preferred embodiment and certain variations and modifications, other variations and modifications may also be used, the invention being defined by the following claims.