Data driving method of silicon-based liquid crystal chip and silicon-based liquid crystal chip

Abstract

The invention discloses a data driving method of a silicon-based liquid crystal chip and the silicon-based liquid crystal chip, the silicon-based liquid crystal chip comprises a pixel matrix, the pixel matrix is composed of multiple rows and multiple columns of pixels, and the data driving method of the silicon-based liquid crystal chip drives the pixel matrix by taking rows as units through row signals, the liquid crystal on silicon chip comprises a column selection multiplexing circuit, the column selection multiplexing circuit comprises a plurality of selection switches, each selection switch corresponds to each column of the pixel matrix, and the corresponding columns of the pixel matrix are driven by controlling the conduction sequence of the selection switches of the column selection multiplexing circuit. Wherein the column selection multiplexing circuit adopts a rotation mode to control the conduction sequence of the plurality of selection switches.

Description

[Technical Field]

[0001] This invention belongs to the field of semiconductor and integrated circuit technology, and specifically relates to a data driving method for a liquid crystal on silicon (LCoS) chip and a liquid crystal on silicon chip applied to the fields of optoelectronics, optical communication, micro-display and projection technology. [Background Technology]

[0002] Liquid Crystal on Silicon (LCoS) is a microdisplay technology that combines liquid crystal technology with semiconductor processes. Its core structure uses a silicon substrate as its base, and independent voltage control for each pixel is achieved through integrated driving circuitry (commonly based on CMOS complementary metal-oxide-semiconductor technology). Subsequent liquid crystal packaging then forms the actual product. At the circuit level, its basic principles can be found in... Figure 1 As shown, where Figure 1 The example shows a 2×2 pixel circuit structure, where the liquid crystal corresponding to each pixel can be equivalent to a capacitor C. LC One end of it is a common level electrode, and the other end can be used with a switching MOSFET and a storage capacitor circuit to control the voltage, thereby realizing the control of pixel voltage.

[0003] Thanks to the precision of silicon-based processes, pixel sizes can be as low as 5-10 micrometers. Silicon-based liquid crystals (LCDs) can align with the development trends of integrated circuit technology, fully utilizing the advantages of advanced processes. They feature small size, high resolution, high luminous efficiency, and low power consumption, and have gained widespread attention from academia and industry in recent years, with broad application prospects in fields such as displays and optical communications.

[0004] In the display field, driving circuits can load expected data onto the corresponding pixels of an LCoS, thereby controlling the modulation polarization or phase state of the liquid crystal layer. Silicon-based liquid crystal technology, by aligning with the development trends of integrated circuit processes and fully utilizing advanced technological advantages, offers advantages such as small size, high resolution, high light efficiency, and low power consumption. In the field of optical communication, LCoS, as a spatial light modulator (SLM), can control the diffraction angle of light of different wavelengths through wavefront modulation, thereby allocating output ports for different wavelengths. Therefore, in optical communication, LCoS chips mainly serve as the core component of wavelength selective switches (WSS). Through polarization modulation of the liquid crystal, in conjunction with a crystal wedge, the refraction direction of each wavelength is controlled to select the optical signal output port, thus realizing reconfigurable optical add-drop multiplexing (ROADM).

[0005] The most common LCoS driving method is to directly drive the pixel matrix using Gate Driver and Source Driver. This method drives the pixel matrix row by row. (See also...) Figure 1 For example, data-driven operation for the pixels in the second row: the row strobe signal Select provided by the Gate Driver for the second row. <2> The switch is activated, and each SourceDriver is responsible for processing each data column (Data). <1> Data <2> ... Data <n>, n is the last column) to provide the expected analog voltage signal of each column, after the switch tube is turned off by Select<2>, the data driving of the second row is completed, and the corresponding voltage data is stored in C Storage The capacitor, after all rows complete data driving and writing, the LCoS data driving of 1 frame is completed.

[0006] Since the Source Driver occupies a large area and performance overhead resources, a separate dedicated Source Driver is usually not allocated for each column of Data signal line, and the Source Driver is usually multiplexed with other signals, Figure 2 A conventional LCoS driving schematic diagram of a Source Driver performing 4 multiplexing (1x4 as an example) is shown, which is combined with additional signals TP1-4 to realize multiplexing, and its driving process is as follows: the Select signal turns on the switch tube, the Source Driver is responsible for 4 Data lines due to multiplexing, it first prepares the data voltage signal of the first column, the TP1 signal turns on Data<1>, the signal is placed on the Data<1> line, and the data writing of the first column of pixels is completed, then the TP1 signal disconnects Data<1> from the Source Driver, the Source Driver prepares Data<2> data, and completes the driving of the second column in cooperation with the TP2 signal; the third and fourth columns are driven in cooperation with the TP3-4 signals, and the behavior is similar.

[0007] The Ramp driving mode is also a common driving method for LCoS, and the Ramp driving mode does not provide analog voltage signals by the Source Driver, and the column data is provided by the slope voltage signal Ramp, the comparator, and the digital count signal Count. Figure 3 With Figure 4 A schematic diagram of the Ramp driving mode is provided (still taking 1x4 as an example). In the Ramp driving mode, there are a slope voltage signal Ramp and a digital count signal Count, which can be provided by Figure 3 As shown, the digital count signal Count and the slope voltage signal Ramp are periodically repeated and have an alignment relationship, in one period T, the slope voltage signal Ramp completes the change, and the counter completes the counting process. The slope voltage signal Ramp can be generated internally by the chip or provided by an external signal, and a DAC is usually used to provide the slope voltage, which includes the min-max range of the data voltage; the usual counting range of the Count signal is 0-2 n -1, n is the bit number of the digital counter. In the Ramp driving mode, the digital Data signal and the digital counting signal Count are compared by the digital comparator component to generate the Switch signal, which controls the subsequent driving process. When the digital Data signal and the digital counting signal Count are consistent, the Switch signal is ON, otherwise, the Switch signal is OFF. When the Switch signal is ON, the analog voltage corresponding to the digital Data signal on the Ramp can be written into the LCoS pixel. Figure 4 The behavior is intuitively shown. The row selection of the pixel matrix is still controlled by the Select, and the data lines are all associated with the Ramp signal. Each column has a corresponding Switch signal to determine whether the Ramp signal is connected to the corresponding data column. The driving process is exemplified as follows: the Select first selects a corresponding row of pixels. The data to be written into each column of the pixel is represented by the digital Data signal, and the Switch signal is generated by the comparator. After a period T, the Switch<1>~<4> signals all complete the ON behavior only once, so that the analog voltage signal corresponding to the digital Data signal on the Ramp signal is written into the corresponding column pixel. After a period T, the data driving of a row of pixels is completed. When all the rows are driven, the data writing of a frame is completed.

[0008] The LCoS used in the field of optical communication has very high requirements on the accuracy of data driving because it involves modulation of optical signals. The driving based on the Source Driver and the Ramp driving method described above have defects in driving timing and charge transfer, leakage and other phenomena in the driving process, thereby causing driving errors.

[0009] Figure 5 A model for evaluating the errors in the above driving is shown, which shows a basic pixel unit and the capacitances of the corresponding data Data Line and TP / Switch signal and the switching information. S is the storage capacitance of the pixel unit, K is the holding capacitance on the data line; since the switch tube is usually implemented by a MOS, the MOS tube itself has a parasitic capacitance, and the C GD , C GS , C DS in the model around the ideal switch are used to represent these capacitances.

[0010] Figure 6 The timing diagram of part of the driving signals is shown. Save1 / Save2 are the changes of the voltage signals stored by two different pixels, respectively corresponding to Switch1 / TP1 and Switch2 / TP2 signals, and Figure 5 Together, they illustrate the accuracy loss phenomenon occurring during driving. The figures are magnified to make the data changes more obvious for easier explanation. Accuracy loss can be mainly summarized into the following four categories:

[0011] Category 1: When the Select signal is turned on (corresponding to dashed line I), C S With C K The circuit is connected via a switching transistor, but the Switch / TP signal is not yet turned on, the data line is in a floating state, there is no additional charging / discharging path, and the capacitors maintain charge conservation. Therefore, charge transfer occurs between the capacitors to achieve charge conservation, and the main charge transfer path occurs at capacitor C. S With C K Between, C S The potential will move towards C K The approach caused a change in the voltage stored in the Save signal.

[0012] Category 2: At the moment the Switch / TP signal is turned off (corresponding to dashed line III), although the data was rewritten during the Switch / TP signal conduction phase (dashed lines II to III), the Switch / TP signal's turn-off command C... GD1 C GS1 When the voltage at one end of the plate changes, charge transfer occurs between the capacitors to maintain charge conservation, which causes fluctuations in the Save voltage that deviate from the original data.

[0013] Category 3: Select signal turn-off time (corresponding to dashed line IV), similar to the behavior in category 2 C. GD2 C GS2 The voltage at one end of the electrode changes, and the charge transfer causes the Save signal to fluctuate again;

[0014] Category 4: Waiting for the next driving phase (corresponding to the V part). At this point, the pixel has finished its previous driving cycle and is waiting for the next driving cycle. This phase is significantly longer than the driving phase, due to C... S Due to capacitor leakage, the voltage stored in Save is gradually lost.

[0015] Furthermore, the aforementioned precision loss is difficult to compensate for in Source Driver reuse and Ramp Driver due to the different loss values ​​for different pixels; see also [reference needed]. Figure 6 After the first type of precision loss (corresponding to time I, indicated by the dashed line) occurs, the data loss continues until the data is rewritten. If the Switch / TP signal has sequential conduction processes, the duration of this effect varies for different pixels. The second type of precision loss has the same problem. For example, in... Figure 6 Since Switch1 / TP1 always turns on before Switch2 / TP2, the time of impact of Type 1 and Type 2 precision loss on Save1 (corresponding to red double arrow 1) is always different from that on Save2 (see red double arrow 2). This difference in precision loss makes data compensation difficult. [Summary of the Invention]

[0016] The purpose of this invention is to provide a data driving method for a silicon-based liquid crystal chip and a silicon-based liquid crystal chip, so as to solve the problem in the prior art that the precision loss between individual pixels cannot be balanced, resulting in the data loss reflected by different pixels not being consistent over a long period of time.

[0017] To achieve the above objectives, a data driving method for a silicon-based liquid crystal chip according to the present invention is provided. The silicon-based liquid crystal chip includes a pixel matrix, which is composed of multiple rows and columns of pixels. The data driving method drives the pixel matrix on a row-by-row basis using row signals. The silicon-based liquid crystal chip includes a column selection multiplexing circuit, which includes multiple selection switches. Each selection switch corresponds to each column of the pixel matrix. The column corresponding to the pixel matrix is ​​driven by controlling the conduction sequence of the multiple selection switches of the column selection multiplexing circuit. The column selection multiplexing circuit controls the conduction sequence of the multiple selection switches in a round-robin manner.

[0018] Based on the above main features, when the data driving method of the silicon-based liquid crystal chip selects the corresponding row through the row signal, it first turns on multiple selection switches of the column selection multiplexing circuit and pre-charges and discharges the data lines of the selection switches to a preset potential.

[0019] Based on the aforementioned key features, the data driving method for the silicon-based liquid crystal chip further includes modeling the accuracy loss of the output voltage of the pixels in the pixel matrix and providing a data compensation formula, specifically:

[0020] V IN =KV OUT +L,

[0021] Where V IN V is the pixel write voltage. OUT The pixel output voltage is K, and L are compensation coefficients, which are determined by coefficient calibration.

[0022] Based on the above main features, the data driving method of the silicon-based liquid crystal chip is a source driver multiplexing driving mode, and the conduction cycles of each selection switch have the following relationship:

[0023]

[0024] Where n is the number of cycles, TPx is the column select signal, Seq(n,TPx) is the conduction cycle number when the TPx signal is turned on for the nth time in the conduction cycle, N is the number of times the Source Driver is multiplexed, and x takes values ​​from 1 to N.

[0025] Based on the aforementioned key features, the data driving method of the silicon-based liquid crystal chip is a ramp driving mode. The switch control signal for the selection switch is provided by a digital Data signal, a ramp voltage signal Ramp, a digital count signal Count, and a digital comparator. Both the digital count signal Count and the ramp voltage signal Ramp are periodically repetitive and aligned. In one cycle, the ramp voltage signal Ramp completes one full-amplitude change, and the counter completes one counting process. The digital Data signal and the digital count signal Count, via the digital comparator, generate a switch control signal to control the selection switch. This switch control signal controls the subsequent driving process. When the digital Data signal and the digital count signal Count are consistent, the selection switch is turned on; otherwise, the selection switch is turned off. When the selection switch is on, the analog voltage corresponding to the digital Data signal is written into the corresponding pixel. The column selection multiplexing circuit uses a rotation method, achieved by rotating the polarity of two adjacent driven ramp voltage signals Ramp and the digital count signal Count.

[0026] To achieve the above objectives, the present invention provides a silicon-based liquid crystal chip, the silicon-based liquid crystal chip including a pixel matrix, the pixel matrix being composed of multiple rows and columns of pixels, the pixel matrix being driven by row signals on a row-by-row basis, and the pixel matrix being connected to a column selection multiplexing circuit, the column selection multiplexing circuit including multiple selection switches, each selection switch corresponding to each column of the pixel matrix, the column corresponding to the pixel matrix being driven by controlling the conduction sequence of the multiple selection switches of the column selection multiplexing circuit, wherein the column selection multiplexing circuit uses a round-robin method to control the conduction sequence of the multiple selection switches.

[0027] To achieve the above objectives, the present invention provides a data driving method for a pixel matrix, wherein the pixel matrix is ​​composed of multiple rows and columns of pixels. The data driving method drives the pixel matrix on a row-by-row basis using row signals. The pixel matrix is ​​coupled with a column selection multiplexing circuit, which includes multiple selection switches, each corresponding to a column of the pixel matrix. The column of the pixel matrix is ​​driven by controlling the conduction sequence of the multiple selection switches in the column selection multiplexing circuit. The column selection multiplexing circuit controls the conduction sequence of the multiple selection switches in a rotating manner.

[0028] Compared with the prior art, the present invention has the following superior technical effects: by having the column selection multiplexing circuit control the conduction sequence of the multiple selection switches in a rotating manner, the present invention can balance the accuracy loss between each pixel, keep the data loss reflected by different pixels consistent over a long period of time, and reduce the difficulty of data compensation. [Attached Image Description]

[0029] Figure 1 This is a schematic diagram of the pixel driving circuit of an existing silicon-based liquid crystal.

[0030] Figure 2 A schematic diagram of a silicon-based liquid crystal driving circuit that performs 4-channel multiplexing for an existing source driver.

[0031] Figure 3 Existing Ramp driver principle diagram Figure 1 .

[0032] Figure 4 Existing Ramp driver principle diagram Figure 2 .

[0033] Figure 5 This is a model for evaluating the error of the Source Driver circuit.

[0034] Figure 6 This is a timing diagram of some signals in the existing driver circuit.

[0035] Figure 7 This is a schematic diagram illustrating the application of the present invention to the Source Driver reuse method.

[0036] Figure 8 This is a schematic diagram illustrating the application of the present invention to a Ramp driver.

[0037] Figure 9 This is a schematic diagram illustrating the data compensation principle in this invention.

[0038] Figure 10 The simulation results without wheel-driven operation are reused for the Source Driver.

[0039] Figure 11 The simulation results of the round-robin drive are reused for the Source Driver.

[0040] Figure 12 The simulation results are for the Ramp without wheel drive.

[0041] Figure 13 Simulation results for Ramp using wheel-driven operation.

Detailed Implementation Methods

[0042] Please see Figure 7 As shown, this is the first embodiment of the present invention, which is applied to the Source Driver multiplexing method. The present invention uses a "round-robin" method to balance the problem that the impact time of the first and second types of precision loss on different pixels is different. Compared with the prior art, the present invention has the following two main improvements.

[0043] The first improvement is that the TP signal (column select signal) is turned on before each row drive Select signal is activated. This is intended to pre-charge and discharge the data line to a certain potential, which is manually controlled, to enhance the controllability of the aforementioned type 1 accuracy loss.

[0044] The second improvement: For the Source Driver multiplexing method, this invention uses the conduction sequence of TP signals in the round-robin adjacent drive to balance the accuracy loss time. The adjacent drive behavior under the round-robin drive can be described as follows, and the conduction rounds of each TP signal have the following relationship:

[0045]

[0046] Where n is the number of cycles, Seq(n,TPx) is the conduction cycle number of the TPx signal during the nth conduction cycle, N is the number of multiplexing cycles of the Source Driver, and x takes values ​​from 1 to N. For example, if the Source Driver has N=3 multiplexing cycles and has 3 TP signals (TP1 to TP3), the values ​​of n and Seq(n,TPx) can have different combinations. When n=2, Seq(1,TPx)=x, Seq(2,TPx)=N+1-x=4-x, that is, the conduction cycle of TP1 is 1,3; TP2 is 2,2; and TP3 is 3,1. When TP1 is first conducted, it is the first of TP1 to TP3. When it is second conducted, it is the third of TP1 to TP3. The rotation of TP2 to TP3 is interpreted in the same way. Please note that, as mentioned above, the value of n and the setting of Seq(n,TPx) can be freely set according to the actual situation. For example, when n=3, the TP1 conduction cycle can be set to 1, 2, 3; TP2 to 2, 3, 1; and TP3 to 3, 1, 2, still satisfying the above formula. When using the Source Driver multiplexing method to drive LCoS, under the round-robin drive, in the long run, the impact time of the first type of precision loss and the second type of precision loss on different column pixels is balanced. See [link to relevant documentation]. Figure 7 As shown by the red double arrows in the image.

[0047] Please see Figure 8 As shown, this is the second embodiment of the present invention, which is a wheel-based implementation method driven by Ramp, and also has two improvements.

[0048] The first improvement is the addition of a precharge phase to the Ramp signal. All Switch signals are turned on once during this precharge phase before the Select signal is selected, pre-charging and discharging the data line to a certain potential. This potential voltage is manually controlled by the Ramp signal to enhance the controllability of Type 1 accuracy loss.

[0049] The second improvement: For the ramp drive method, the wheel drive scheme uses a wheel ramp signal and

[0050] The polarity of the Count signal is used to balance the time lost in precision. Excluding the Precharge phase, the rotary drive scheme reverses the polarity of the Ramp and Count signals in two adjacent drives; that is, the Ramp ramp is reversed from "small to large" to "large to small," and the Count signal is reversed from "forward sequence" to "reverse sequence." Through this technique, in rotary drive mode, the time lost in precision due to the writing of pixels with different data sizes is balanced. (See [reference needed]). Figure 8 As shown by the double arrows in the middle.

[0051] To address the accuracy loss in the first and second embodiments, this invention also proposes a data compensation method based on accuracy loss assessment, the specific method of which is described below.

[0052] When the pixel write voltage V IN At that time, due to the aforementioned loss of accuracy, the V-shape displayed externally... OUT V OUT ≠V IN In order to make V OUT The performance is as expected; data compensation is required before writing to V. IN Data compensation is based on the following formula:

[0053] V IN =KV OUT +L

[0054] The compensation coefficients K and L are determined through calibration (details to follow).

[0055] Please combine Figure 9 As shown, the derivation process of the data compensation method is explained below, where... Figure 9 The "change" in accuracy loss was magnified when plotting to make the relevant data in the graph clearer. The K and L symbols in the following derivation represent the compensation coefficients that can be calibrated or calculated.

[0056] Over a long period of time, the data value V that a pixel exhibits externally. OUT Its average storage voltage is:

[0057]

[0058] During stages t2 to t3, pixel data is flushed and written, remaining as the written data V. IN ,Right now:

[0059] V5 = V IN

[0060]

[0061] At time t3, the second type of accuracy loss described above occurs. At this time, the Switch / TP signal turns off the switching transistor. To maintain charge conservation, charge transfer occurs, resulting in:

[0062]

[0063] Where ΔV 0i and ΔV 1i It is the capacitance C before and after charge transfer. i The voltage difference between the two ends, as shown in the above equation, indicates that charge is conserved before and after charge transfer.

[0064] Due to the voltage V that turns the switching transistor on and off ON and V OFF And since each capacitor C can be considered a constant, the result of V4 can be linearly expressed as:

[0065] V4=K′2V IN +L′2

[0066] but:

[0067]

[0068] Similarly, V3 can be represented as:

[0069] V3=K′3V IN +L′3

[0070] The most significant electrical behavior between t4 and t5 is capacitive leakage, which can be expressed by the following two equations:

[0071]

[0072] Where V0 is the initial leakage voltage, and τ is a time constant determined by the leakage capacitance and the resistance along the leakage path. In LCoS applications, to ensure display accuracy, leakage behavior is significantly suppressed. The leakage duration t between t4 and t5 is usually t << τ, and τ is usually extremely large. From the derivative expression, it can be seen that the leakage slope changes little and is related to V0. A linear approximation is made. Since the timing setting allows us to consider the leakage duration between t4 and t5 as a constant, the voltage V2 at times t1 and t5 and the integral result between t4 and t5 are:

[0073] V2=V3-K′4V3=K″4V IN +L″4

[0074]

[0075] Since the invention's driving method pre-charges and discharges the data line to a fixed potential, the integral result between t1 and t2 is similar to the related derivation above and can be expressed as:

[0076]

[0077] The duration between t1 and t5 can be considered a constant due to the timing settings. Combining formulas (2) to (6), we can obtain:

[0078] V OUT =K A V IN +L A

[0079] Right now:

[0080] V IN =KV OUT +L

[0081] The above is the formula for data compensation. The compensation coefficients K and L can be determined through coefficient calibration, which is one of the most common methods in engineering. Analysis shows that, theoretically, the compensation coefficients K and L are fixed values. Coefficient calibration involves deriving these coefficients from known "standard values" or "reference conditions" to accurately reflect the actual situation. For example, in LCoS products, for debugging purposes, interfaces are usually reserved or other methods are used to read the output values ​​of some pixels in the pixel matrix. This reading is feasible and common; the read output value corresponds to V. OUT The value written by the driver is clearly controlled manually, corresponding to V. IN .

[0082] For ease of understanding, the following describes a calibration method:

[0083] Step 1: Manually control the writing of V IN1 Read the corresponding V OUT1 ;

[0084] Step 2: Manually control the writing of V IN2 Read the corresponding V OUT2 ;

[0085] Step 3: Obtain the system of equations:

[0086] V IN1 =KV OUT1 +L

[0087] V IN2 =KV OUT2 +L;

[0088] Step 4: Solve the system of two linear equations in two variables that is common in Step 3, and you will get the compensation coefficients K and L.

[0089] The calibration above recorded two sets of data (as in steps one and two), which is the minimum amount of data required to calibrate K and L (K≈1.1 and L≈0.019, as determined in simulation one of the following examples, were determined using the above method). More data can actually be recorded for calibration. Since the data conforms to a linear law according to the formula, these data points can be fitted with a linear function to obtain the compensation coefficients K and L. This calibration method can also determine the values ​​of both, and its accuracy is higher due to the larger amount of data.

[0090] To better illustrate the technical effects achieved by the wheel drive described in this invention, simulations of two embodiments will be used below. The timing settings in the simulations below are only used to illustrate the technical effects of this invention, and the timing settings amplify the impact of accuracy loss to more clearly illustrate the technical effects of the invention. They may differ from the timing settings of actual LCoS products.

[0091] Example Simulation 1

[0092] The following description is based on a driver simulation of three pixels (corresponding to storage results Save1 to Save3) for a Source Driver tri-multiplexing, where the written data is...

[0093] Save1 = Save2 = Save3 = 2V. Assume Type 1 accuracy loss ( Figure 6 Before the dashed line I) occurs, the voltage on the Dataline is set to 0V. Under this condition, the charge transfer result is considered.

[0094] Simulations were performed using timing signals for both standard drive and the invention-based rotary drive, and the impact of accuracy loss on the duration of the Save storage voltage signal was observed. The Save results for both standard and invention-based rotary drives are shown below. Figure 10 and Figure 11 As shown. In Figure 10 In the context of the Save signal, the duration of accuracy loss depends on the pixel position, with Save1 through Save3 experiencing different durations. Figure 11 In the process, the duration of the accuracy loss of Save1 and Save3 is "rotated" in adjacent driving stages, and the difference in the duration of the accuracy loss of the two is balanced. In the long run, the duration of the accuracy loss of Save1 and Save3 is closer to Save2, and the duration of the impact of Save1 to Save3 is close to the same, achieving the technical effect described above.

[0095] To more intuitively illustrate the technical effect, a simulation drive of 20 drive cycles was performed, and the average data voltage on Save1 to Save3 was statistically analyzed. This value reflects the level of color gradation displayed in the actual drive. The results for Save1 to Save3 and related data are shown in the table below:

[0096] Simulation-driven data long-range average statistics

[0097]

[0098] It is evident that the Save1 to Save3 results, when writing the same data, all show a long-term voltage of approximately 1.8V when using a rotary drive, while the differences are significant when the rotary drive is not used. The rotary drive, when writing the same data voltage, balances the impact of accuracy loss on the duration of the write, resulting in smaller differences in color gradation between pixels writing the same data and thus stabilizing the color gradation display effect.

[0099] The technical effect of the invented data compensation scheme is explained below. Simulations of writing 2V and 1.5V data under rotary drive were performed, and the long-term average values ​​of the two stored values ​​were calculated to be 1.801V and 1.346V, respectively. Using these results, the data compensation expression was calibrated, yielding K≈1.1 and L≈0.019. With the target long-term values ​​of 2V, 1.5V, and 1V for Save1 to Save3, the effect was verified again using simulation-driven statistical long-term information. The relevant results are shown in the table below, demonstrating a good compensation effect.

[0100] Simulation-driven data long-range average statistics (compensation)

[0101]

[0102] Example Simulation 2

[0103] Simulation 2 of Example 2 is similar to Simulation 2 of Example 2, but uses a ramp drive and an invented rotating ramp drive. Simulation was still performed on three pixels (corresponding to storage results Save1 to Save3). The ramp signal was provided by an 8-bit DAC (00000000~11111111, corresponding to 0~5V), and the counter signal was also 8 bits. The written data was incremented gradually: Save1 = 01000100 (1.333V), Save2 = 01100110 (2.000V), Save3 = 10001000 (2.667V). Assuming a first-type precision loss (… Figure 6 Before the dashed line I) occurs, the voltage on the Dataline is set to 0V. Under this condition, the charge transfer result is considered.

[0104] Simulations were performed using both a standard ramp drive and an inventive round-robin ramp drive, respectively, to observe the impact of accuracy loss on the duration of the save storage voltage signal. The save results for the standard and inventive round-robin drives are shown below. Figure 12 and Figure 13 .exist Figure 12 In the process, the duration of the precision loss effect of the Save signal is related to the size of the written data; the duration of the effect varies for Save1 through Save3, and the effect is continuous. Figure 13 In the process, the duration of accuracy loss affecting Save1 to Save3 changed significantly in adjacent driving stages. The difference in the duration of accuracy loss affecting the three was averaged over a long period of time, thus achieving a near-consistency and the technical effect described above.

[0105] To more intuitively illustrate the technical effect, a simulation drive of 20 drive cycles was performed, and the average data voltage on Save1 to Save3 was statistically analyzed. This value reflects the level of color gradation displayed in the actual drive. The results for Save1 to Save3 and related data are shown in the table below:

[0106] Simulation-driven data long-range average statistics

[0107]

[0108] Without a rotation drive, the difference between the stored value and the written data is affected not only by the size of the written data but also by the added effect of precision loss, making it difficult to find a pattern among the differences. After using a rotation drive, the added effect of precision loss in the differences among the three is balanced. The average values ​​of Save1 and Save3 data decrease from 1.125V and 1.621V without rotation to 0.9765V and 1.524V, respectively. This is because Save1 and Save2, which write at 1.333V and 2.000V, are always affected by the precision loss that causes the data to shift downward from the original write value for a shorter time than Save3, which writes at 2.667V. The rotation drive balances the time affected by precision loss for each data, which is equivalent to Save1 and Save2 having a longer time affected by the precision loss that causes the data to shift downward from the original write value; Save3 is the opposite, which is equivalent to a shorter time affected by the precision loss that causes the data to shift downward from the original write value, manifested as the data increasing from 2.045V to 2.072V. Since the differences depend more on the written data, the patterns between the differences are easier to discern. For example, the table above shows that when using a rotary drive, the difference between the stored value and the written data changes by about 120mV for every 0.667V increase in the written data.

[0109] The following describes the technical effect of the invention's data compensation scheme under Ramp rotation drive. Using the two sets of results written (long-range average) of 1.333V (0.9765V) and 2.000V (1.524V) in the table above, the data compensation expression is calibrated, yielding K≈1.218 and L≈0.144. With the long-range performance of Save1 to Save3 as Save1=01000100 (1.333V), Save2=01100110 (2.000V), and Save3=10001000 (2.667V) as the target, the actual written digital quantity closest to the compensation result is verified using simulation-driven statistical long-range information. The relevant results are shown in the table below, demonstrating a good compensation effect.

[0110] Simulation-driven data long-range average statistics (compensation)

[0111]

[0112] In addition, the present invention also proposes a silicon-based liquid crystal chip, the silicon-based liquid crystal chip including a pixel matrix, the pixel matrix being composed of multiple rows and columns of pixels, the pixel matrix being driven by row signals on a row-by-row basis, and the pixel matrix cooperating with a column selection multiplexing circuit, the column selection multiplexing circuit including multiple selection switches, each selection switch corresponding to each column of the pixel matrix, the column corresponding to the pixel matrix being driven by controlling the conduction sequence of the multiple selection switches of the column selection multiplexing circuit, wherein the column selection multiplexing circuit uses a round-robin method to control the conduction sequence of the multiple selection switches.

[0113] Furthermore, the present invention also provides a data driving method for a pixel matrix, wherein the pixel matrix is ​​composed of multiple rows and columns of pixels, and the pixel matrix is ​​driven by row signals on a row-by-row basis. The pixel matrix is ​​also used in conjunction with a column selection multiplexing circuit, wherein the column selection multiplexing circuit includes multiple selection switches, each selection switch corresponding to each column of the pixel matrix. The column corresponding to the pixel matrix is ​​driven by controlling the conduction sequence of the multiple selection switches of the column selection multiplexing circuit, wherein the column selection multiplexing circuit controls the conduction sequence of the multiple selection switches in a round-robin manner.

[0114] Compared with the prior art, the present invention has the following superior technical effects:

[0115] 1. This invention balances the precision loss between pixels by rotating different timing signals. When writing the same data voltage, the rotating drive displays smaller color gradation differences and stabilizes the color gradation display effect.

[0116] 2. This invention balances the precision loss between pixels by rotating different timing signals, making the data loss reflected by different pixels more consistent, thereby reducing the difficulty of data compensation.

[0117] 3. This invention utilizes a data compensation method to make the external voltage of pixels over long distances more controllable, thereby achieving refined drive control.

[0118] The method disclosed in this invention was originally intended to address the high-precision driving requirements of LCoS in the field of optical communication. However, those skilled in the art will know that the method disclosed in this invention is not limited to the field of optical communication, but can also be applied to LCoS driving in other fields such as display, and is also beneficial for balancing driving accuracy errors and performing error compensation.

[0119] It is understood that those skilled in the art can make equivalent substitutions or modifications to the technical solution and inventive concept of the present invention, and all such substitutions or modifications should fall within the protection scope of the appended claims.< / n>

Claims

1. A data driving method for a silicon-based liquid crystal chip, the silicon-based liquid crystal chip comprising a pixel matrix, the pixel matrix being composed of multiple rows and columns of pixels, the data driving method driving the pixel matrix by row signals on a row-by-row basis, and the silicon-based liquid crystal chip comprising a column selection multiplexing circuit, the column selection multiplexing circuit comprising multiple selection switches, each selection switch corresponding to each column of the pixel matrix, the column corresponding to the pixel matrix being driven by controlling the conduction sequence of the multiple selection switches of the column selection multiplexing circuit, characterized in that: The column selection multiplexing circuit uses a round-robin method to control the conduction sequence of the multiple selection switches.

2. The data driving method for a silicon-based liquid crystal chip as described in claim 1, characterized in that: The data driving method for the silicon-based liquid crystal chip first turns on multiple selection switches of the column selection multiplexing circuit when the corresponding row is selected by the row signal, and pre-charges and discharges the data lines of the selection switches to a preset potential.

3. The data driving method for a silicon-based liquid crystal chip as described in claim 2, characterized in that: The data driving method for the silicon-based liquid crystal chip further includes modeling the accuracy loss of the output voltage of the pixels in the pixel matrix and providing a data compensation formula, specifically: Let: V IN =KV OUT +L, Where V IN V is the pixel write voltage. OUT K represents the pixel output voltage, and L represents the compensation system coefficients, which are determined through coefficient calibration.

4. The data driving method for a silicon-based liquid crystal chip as described in claim 1, characterized in that: The data driving method for the silicon-based liquid crystal chip is a source driver multiplexing driving mode, and the conduction cycles of each selection switch have the following relationship: Where n is the number of cycles, TPx is the column select signal, Seq(n,TPx) is the conduction cycle number when the TPx signal is turned on for the nth time in the conduction cycle, N is the number of times the Source Driver is multiplexed, and x takes values ​​from 1 to N.

5. The data driving method for a silicon-based liquid crystal chip as described in claim 1, characterized in that: The data driving method of the silicon-based liquid crystal chip is the Ramp driving mode. The switching control signal of the selection switch is provided by the digital Data signal, the ramp voltage signal Ramp, the digital count signal Count, and the digital comparator. The digital count signal Count and the ramp voltage signal Ramp are periodically repeated and have an alignment relationship. In one cycle, the ramp voltage signal Ramp completes one full-amplitude change, and the counter completes one counting process. The digital Data signal and the digital count signal Count generate a switching control signal to control the selection switch through the digital comparator. The switching control signal controls the subsequent driving process. When the digital Data signal and the digital count signal Count are consistent, the selection switch is turned on; otherwise, the selection switch is turned off. When the selection switch is turned on, the analog voltage corresponding to the digital Data signal is written into the corresponding pixel. The column selection multiplexing circuit adopts a rotation mode by rotating the polarity of the ramp voltage signal Ramp and the digital count signal Count twice.

6. A silicon-based liquid crystal chip, the silicon-based liquid crystal chip comprising a pixel matrix, the pixel matrix being composed of multiple rows and columns of pixels, the pixel matrix being driven by row signals on a row-by-row basis, and the pixel matrix cooperating with a column selection multiplexing circuit, the column selection multiplexing circuit comprising multiple selection switches, each selection switch corresponding to each column of the pixel matrix, the column corresponding to the pixel matrix being driven by controlling the conduction sequence of the multiple selection switches of the column selection multiplexing circuit, characterized in that: The column selection multiplexing circuit uses a round-robin method to control the conduction sequence of the multiple selection switches.

7. A data driving method for a pixel matrix, wherein the pixel matrix comprises multiple rows and columns of pixels, the data driving method drives the pixel matrix row by row using row signals, and the pixel matrix cooperates with a column selection multiplexing circuit, the column selection multiplexing circuit including multiple selection switches, each selection switch corresponding to each column of the pixel matrix, and the column corresponding to the pixel matrix is ​​driven by controlling the conduction sequence of the multiple selection switches of the column selection multiplexing circuit, characterized in that: The column selection multiplexing circuit uses a round-robin method to control the conduction sequence of the multiple selection switches.

Images