Posted on 01 November 2019

Large Form Factor Video Displays

Free Bodo's Power Magazines!




Dot correction reduces system cost

The LED lighting market is growing rapidly with applications ranging from alphanumeric informational displays to high quality, large form-factor video displays. These markets suffer from two opposing market forces: pressure for higher quality displays, and pressure for lower cost solutions.

By Michael Day, Power Management Application Supervisor for Portable Power, Texas Instruments


High-quality displays require precise control of LED brightness and color. Unfortunately, the inherent variations in LED characteristics force video display manufacturers to implement processes and procedures to compensate for these variations. This adds cost to the system. High-end LED driver ICs implement a simple, software-programmable technique called “dot correction” that helps manufacturers to achieve high-quality products while still keeping system level production costs in check. This paper shows how LEDto- LED brightness variations affect video display manufacturers’ costs, how dot correction compensates for LED brightness variations while helping to decrease production costs, and provides a simple dot correction example.

The System

Large form factor LED video displays integrate hundreds of individual display panels and hundreds of thousands of LEDs to create a seamless large screen video image. Figure 1 shows a typical display. Each panel contains a 16 x 16 pixel matrix. Each individual pixel contains multiple LEDs, usually two red, one green, and one blue LED.

Progression from individual pixels to large form factor LED video display.

Figure 1 caption

LED-to-LED Brightness Variations

The key to a high-quality display is precise control of LED brightness. To faithfully reconstruct the original video image, a video display must accurately control LED brightness to a fraction of a percent. Unfortunately, LED brightness varies from LED to LED, even within a single manufacturer’s production lot. This “LED production variation” places limitations on a display’s picture quality. The root cause of LED brightness variations stems from the fact that all semiconductor- manufacturing processes contain inherent and uncontrolled variations. These manufacturing variations result in uncontrolled LED brightness variations. Typical offthe- shelf LED products may have a +/- 60 percent variation in lumen output for the same forward current. This excessive variation is not acceptable for higher end video displays or in many other applications. LED manufacturers use a method called “binning” to help reduce the lumen variation. Binning is a process that separates LEDs into separate bins or lots of similar brightness. The customer can then buy specific brightness bins to get LEDs that are similar in brightness. For example, the datasheet for a specific LED part number from company X shows a lumen variation between 71 and 280 lumens (+/- 60 percent) for two LEDs with the same forward current. The manufacturer bins this LED into six smaller groups to significantly reduce LED-to- LED variations within each group. Even within these smaller groups, LEDs have significant variation. The highest brightness bin still varies +/- 11 percent (224 to 280 lumens). Although binning provides the end user with higher quality LEDs, it comes at the expense of higher cost. Since the LED manufacturer cannot control the percentage of LEDs that end up in specific bins, they incur an effective reduction in yield when trying to fill specific bins to support customer orders. Lower yield drives up manufacturing costs. Test costs also increase since the manufacturer must individually test each LED. Manufacturers always pass these costs on to the customer.

LED Aging

LEDs age with time and their light output decreases. Typical aging specifications define an LED’s end-of-life when the lumen output at a specified current drops to half of its initial value. A typical LED’s lifetime ranges from 50,000 to 100,000 hours, depending on the LED and its operating conditions. While LED-to-LED brightness variations affect panel makers at production, LED aging affects them when the time comes to service existing panels. A typical video application may have hundreds or thousands of individual panels. If one of these panels becomes defective or damaged, it must be replaced. The new panel will be brighter than the rest of the display because the newer LEDs have not aged. The manufacturer must calibrate the new panel to match the brightness of the aged display. This calibration adds cost to the system, especially when the calibration involves changes in hardware. Fortunately, higher-end LED driver ICs provide the panel maker with a simple to use method for managing differences in LED brightness and aging.

Dot Correction

Dot correction helps manage variations in LED brightness by adjusting the current supplied through each individual LED in the array. When properly implemented, dot correction provides a uniform brightness across a panel, even with unmatched LEDs. Its ability to match LED brightness provides significant improvements in display quality. The concept of dot correction is simple. Manufacturers measure the brightness of individual LEDs at full-rated current. The dimmest LED in the system is designated as the “base” LED to which every other pixel is matched. They then program a dot correction value for each LED that appropriately scales the forward current to set all LEDs to the same brightness. Dot correction can be implemented at the hardware or software level. Devices like the Texas Instruments TLC5940 implement dot correction at the hardware level, but with software control. This method provides significant benefits over other purely hardware or purely software implementations. The TLC5940’s dot correction values are calculated and then programmed into the IC using a software interface. The TLC5940 controls the dot corrected LED current at the hardware level. The hardware-software method is superior to a purely hardware method because it eliminates the need for expensive and time-consuming hardware changes such as component value changes in the production flow. It is also superior to a purely software method, since the TLC5940 controls the dot corrected LED currents during a panel’s operation. As a result, the system or board level microprocessor is free to perform other tasks. The TLC5940 also eliminates the need for expensive look-up tables and complex software multiplication routines for each LED in every video refresh cycle.

Implementing Dot Correction in Production

The mechanics to implement dot correction are relatively easy and are best explained with a simple example. If a panel’s specification requires its green LEDs to have a maximum luminosity of 80 mcd (millicandela), the manufacturer must set the LED driver’s maximum current to produce at least 80 mcd in all LEDs. This current setting must take LED brightness variations into account. After each individual panel is built, the microprocessor programs all LEDs to their maximum brightness. The manufacturer measures each LED’s brightness, calculates the appropriate dot correction values, and programs the TLC5940 with the new dot correction values. This process can be repeated until the command to turn all LEDs on at maximum brightness results in uniform brightness across the panel. If an Osram LP E675 LED is used, 43mA of forward current guarantees the dimmest LED produces at least 80 mcd. During a panel’s production test and calibration, the brightness of all LEDs is measured at the driver’s full-programmed current of 43mA, which is set by an external resistor. A typical pre-dot corrected LED luminous intensity histogram for sixteen LEDs might resemble that in Figure 2. The data in the foreground is the LED current in mA, while the data in the background is the LED brightness in mcd. LED brightness variations are evident without dot correction, which may be unacceptable in higher-end displays. The manufacturer can now use the pre-dot correction information to calibrate the LED brightness. When programmed to full brightness, the IC must dot correct the luminous intensity of LED1 from 83 mcd to 80 mcd. The TLC5940 has six-bit dot correction (63 steps), which corresponds to a full-scale resolution of 1.59 percent per step.

LED brightness and toward current histogram before dot correction

The following formula calculates the correct dot correction level for each LED:

DCProduction = LDesired / LInitial * 63

Where DCproduction is the calculated dot correction value at production, Lbaseline is the desired brightness level, and Linitial is the initial measured brightness at the reference current. Since dot correction values are whole numbers, the equation’s result must be rounded up or down.

After each LED’s DOT correction value is calculated and stored, the TLC5940 automatically generates a uniform brightness in all LEDs. Since dot correction is controlled by the hardware system, the processor is not concerned with variations in LED brightness. When the processor commands full brightness, the TLC5940 accepts the brightness command and automatically scales the current in each LED to provide full brightness. Figure 3 shows the LED currents and resulting brightness after dot correction is applied.

LED brightness and forward current histogram after dot correction

Implementing Dot Correction in the Field

Dot correction also simplifies calibration of replacement panels for existing video displays. A replacement panel for an aged video display is already factory dot-corrected to ensure uniform brightness. However, the replacement panel’s brightness varies from the existing panels’ brightness. The manufacturer can calibrate the replacement panel’s brightness much the same way they calibrated individual pixels on the panel.

After installing the replacement panel, the manufacturer measures both the new panel’s brightness and the older panel’s brightness. They then calculate a “replacement panel adjustment factor” using the equation below.

KRPAF = LumensOriginal_Panel / LumensReplacement_Panel

Where KRPAF is the replacement panel adjustment factor, LumensOriginal_Panel is the brightness of the original panels in the video display, and LumensReplacement_Panel is the brightness of the replacement panel. KRPAF scales each LEDs dot correction value to provide a uniform dimming for the replacement panel. The equation below calculates the new dot correction value for each individual LED.

DCReplacent_panel_LED = DCProduction * KRPAF

Where DCReplacent_panel_LED is the new dot correction value for an individual LED in the replacement panel, DCProduction is the original dot correction value that was programmed in production at the factory, and KRPAF is the replacement panel adjustment factor. This replacement panel calibration method results in an overall video display with a uniform brightness. The calibration process is software-based, so it can be easily changed. This saves the additional time and expense required for hardware-only based calibration routines.


Dot correction easily compensates for inherent variations in LED brightness, allowing display manufacturers to produce high-quality video displays while still keeping costs down. The software-hardware based TLC5940 dot correction provides distinct advantages over other methods. The hardware portion allows cheaper, higher performance systems by performing dot correction at the hardware level.



VN:F [1.9.17_1161]
Rating: 0.0/6 (0 votes cast)

This post was written by:

- who has written 791 posts on PowerGuru - Power Electronics Information Portal.

Contact the author

Leave a Response

You must be logged in to post a comment.