The LED electronic display screen is composed of a plurality of mutually independent pixel points (light emitting elements), and due to the separability of the pixel points, it is determined that the control and driving of the light emission can only be performed in a digital manner. The lighting states of these pixels are synchronously controlled by the controller and driven independently. True-color video display means that the brightness of each pixel is separately controlled and synchronized within a specified scan time. The large screen is made up of tens of thousands of pixels, which makes the complexity of the system much larger than that of the two-valued display screen and puts forward higher requirements for the overall data transmission speed. It is obviously unrealistic to set a regular D/A for each pixel. It is necessary to find a solution that minimizes system complexity and maximizes performance.
It is known from the visual principle that the average brightness perception of a pixel by a person may depend on its on/off duty cycle. That is, as long as the pixel on/off duty ratio is adjusted, brightness control can be achieved. For LED electronic displays, this means that as long as the number representing the luminance of a pixel is converted into the time (D/T conversion) at which the pixel emits light, a D/A conversion of the luminance is achieved.
Let the screen data refresh cycle be: the data that controls the brightness of any pixel is an n-bit binary number.
D=bi2I (where bi=0 or 1)
Ton corresponds to the light-emitting time of D, then the duty cycle of the pixel lighting/extinguishing is:
d=Ton/Ts=D=bi2I
This expression can be implemented by a presettable subtraction counter, but a counter for each pixel will make the display circuit extremely complicated. The above equation is rewritten as: Ton=Ts bi2I, which means that Ton can be divided into several periods, and since the Ton synthesized for several separate periods is the same as the continuous Ton having the same total length, the visual effect is the same. Thus, in general, for n-bit binary data D=bi2I, divide Ts into n segments and select an appropriate time division function f(I) so that the first segment Ti=Tsf(I), where 0Tibi=Tsf ( I)bi, at this time d=Ton/Ts=Tibi/Ts=f(I)bi
That is, the on/off duty cycle of this pixel. Since the function f(I) can be common to all pixels, the above formula shows that as long as f(I) is used to control each pixel uniformly, all pixels of the full screen can be independent and synchronized. T conversion. The above equation can be implemented with the circuit of Figure 1 for a single pixel point. The SFR in the figure is an 8-bit shift register. The figure shows the waveform of the time division function f(I).
Large-screen display driver circuits typically use a "serial shift + latch + drive" architecture to minimize data transfer lines. To achieve the full-screen at the same time, as long as all ST signals can be controlled by f (I). Of course, the premise for doing so is to require the shift register to store the same weight in the control data of each pixel, and this can be done through pre-data processing.