Method of realizing grayscale of LED display screen by pulse width modulation

Pulse width modulation is to achieve grayscale by controlling the light-emitting time of LED . The photoelectric response speed of LED is relatively fast, which can reach tens of megahertz. Therefore, we can use the control circuit to light up and extinguish the LED at high speed to obtain fine grayscale. In this way, the current driving the LED becomes a series of pulses, and the width of the pulse is modulated by the image data, so it is called pulse width modulation. This method has low control cost, high control accuracy and good linearity, and has been widely used.

1. Weight Lighting Method

Taking the corrected 12-bit image data D11...D0 as an example to illustrate this method, first convert the parallel data into serial data, and then output the LED lights in turn . When the highest bit D11 is output, the LED lights need to be turned on (or off) for 2 048 clock cycles (the LED lights are turned on when D11 is high, and the LED lights are turned off when D11 is low); when D10 is output, 1 024 clock cycles are required; and so on, only one clock cycle is required when D0 is output. In this way, 4 096 clock cycles are required to complete the grayscale formation of a pixel. Obviously, this method is too time-consuming, and it is generally not used in practice.

2. Weight Extinction Method

Still taking the corrected 12-bit image data D11...D0 as an example to illustrate this method, firstly, the parallel data is still converted into serial data, and then the LED lights are output in turn. However, this method only needs 12 clock cycles to complete the grayscale formation of one pixel. This method defines that when outputting D11, the LED light needs to be turned on (or off) for 1 clock cycle; then, DIO is 1/2 clock cycle; D9 is 1,4 clock cycle; and so on, D0 is 1/2048 clock cycle. It should be noted that the clock mentioned here refers to the lock clock (LAT) of the LED driver output to the LED display screen , which is generally around 1 000-3 000Hz (the internal clock of the main control chip is generally above 100 MHz). When each bit of image data is output and locked on the screen, the main control chip outputs a signal called "blanking pulse" (OE), which is generally defined as turning off the LED when OE is high and turning on the LED when OE is low. The width of the OE pulse must be precisely controlled. When outputting D11, OE outputs 1 clock cycle; when outputting DO, OE outputs 1/2 048 clock cycles. Although this method has a small time cost, it greatly sacrifices the brightness of the display screen . If the image data is fully bright (that is, D11... D0 are all 1), the brightness of the display screen is only 17% (approximate value) of the weighted lighting method. The formula is as follows:

Brightness percentage = 1/12 (1+1/2+1/4+1/8+1/16+1/32+1/64+1/128+1/256+1/512+1/1 024+1/2 048) = 17%

Obviously, such brightness is unacceptable.

3. Hybrid method

In order to increase the brightness of the screen without wasting too much time, the two methods can be mixed. For example, we can make D11 repeat 8 times, D10 repeat 4 times, D9 repeat 2 times, D8 output once, and from D7 on, the pulse width of OE is halved, and when D0 is output, the pulse width of OE is 1/256. This method requires 23 clock cycles to complete the grayscale formation of a pixel. It is much less than the 4,096 clock cycles of the weighted lighting method, and the brightness of the screen can reach 70% (approximate value) of the weighted lighting method. The formula is as follows:

Brightness percentage = 1/23 (8+4+2+1+1/2+1/4+l/8+1/16+1/32+1/64+1/128+l/256) = 70%

Generally speaking, this brightness should be acceptable. If you want to increase the brightness, you can repeat D11 16 times or more; conversely, if you want to reduce the time cost, you can reduce the repetition of D11 a little bit. You can't have both, and you need to consider comprehensively.