What Causes LED Display Ghosting, Dead Pixel Crosses, and Low-Gray Color Shift?

The image quality of an LED display is directly tied to the performance. Among the key components, the constant-current driver chip is crucial for column driving. Its output accuracy and response speed are vital for final picture quality. Common display flaws like ghosting, cross-shaped dead pixels, and low-gray color shift are often linked to mismatches in these driving parameters.

In contrast, the row driving module initially had a simpler role in early LED displays. It handled basic row switching and was considered low-tech. However, with the rise of fine-pitch displays , demands have changed. Higher pixel density requires faster scanning and more current precision. Consequently, the row driver’s function has evolved. It’s no longer just a simple switch. Now, it integrates features like blanking, over-current protection, and ESD protection.

This evolution makes row driver design much more complex. Engineers now face six key challenges:

• Upper Ghosting
• LED Reverse Voltage
• Short-Circuit “Caterpillar” Effect
• Open-Circuit “Cross” Effect
• High LED Forward Voltage (Vf)
• High-Contrast Coupling

Let’s break down each issue.

Upper Ghosting

In scanning LED displays, the timing synchronization of row switching and device characteristics directly affect the display effect. Upper ghosting mainly comes from two factors: the switching delay of PMOS transistors and the slow discharge of residual charges in the row-line parasitic capacitor Cr.

Scanning screens work by lighting rows one by one. When Row(n) turns on, the row drive circuit charges the parasitic capacitor Cr through the PMOS transistor. Eventually, the voltage across Cr stabilizes at VCC .

When the system sends a row switching command, the PMOS transistor for Row(n) should turn off immediately in an ideal scenario. The residual charges on Cr also need to discharge quickly.

But in practice, PMOS transistors have inherent switching delays. Meanwhile, the parasitic capacitor Cr retains charges. This causes a significant lag in charge discharge.

When the VLED and OUT channel of the next row turn on simultaneously, the unfully discharged residual charges from the previous row form a conduction loop through the LEDs. This makes the LEDs in Row(n) glow faintly. This phenomenon is what the industry calls upper ghosting.

This problem is particularly obvious in small-pitch displays with high scan frequencies. It seriously reduces image contrast.

The core technical solution to upper ghosting is to discharge the residual charges in Cr early and quickly during row switching. This breaks the condition for forming a conduction loop between residual charges and the next row’s drive signal.

From an engineering perspective, the mainstream industry solution uses dedicated row drive tubes with integrated blanking functions. These devices have built-in independent pull-down discharge circuits.

The circuits activate accurately during row switching. They accelerate Cr charge discharge through a low-impedance path.

In this solution, the value of the blanking voltage VH directly determines the charge discharge efficiency and upper ghosting elimination effect. The lower the VH setting, the larger the potential difference between the pull-down circuit and Cr.

This creates stronger driving force for charge discharge and faster discharge speed. The upper ghosting elimination effect becomes better.

Engineering practice data verifies that when VH < VCC－1V, upper ghosting can be completely eliminated within the conventional scan frequency range (120Hz~360Hz). It also does not significantly affect other display performances.

Note that VH cannot be adjusted too low. Excessively low VH will cause other derivative problems, which will be detailed in subsequent chapters.

LED Reverse Voltage

LEDs are the core light-emitting components of LED displays. Their reliability directly determines the display’s service life. Reverse impact voltage is one of the key stress factors affecting the long-term reliability of LEDs.

The PN junction inside LEDs has weak tolerance to reverse voltage. When the reverse voltage exceeds its threshold, the PN junction characteristics degrade. In severe cases, the PN junction breaks down directly. This ultimately appears as dead pixels on the display.

This problem is more prominent in small-pitch LED displays. On one hand, small-pitch LEDs have smaller packaging sizes and more compact internal structures. Their reverse voltage tolerance is relatively lower.

On the other hand, small-pitch displays have higher drive density and faster drive signal switching frequencies. This makes reverse voltage surges more likely. Controlling LED reverse voltage has thus become a long-term core pain point in the industry.

From a technical principle perspective, LED reverse voltage generation relates closely to parasitic parameters in the drive circuit. When the constant current output channel of the LED display turns off, the parasitic inductance in the drive loop produces a freewheeling effect due to sudden current changes.

This freewheeling current continues to charge the parasitic capacitor at the channel. It forms a high-voltage glitch signal across the capacitor.

This high-voltage glitch signal superimposes with the output voltage of the row drive tube. It ultimately forms the reverse voltage applied across the LEDs.

Thus, the blanking voltage VH of the row drive tube is a key variable affecting the LED reverse voltage. Under the premise of a fixed voltage level after the constant current output channel turns off, the higher the VH value, the smaller the superposition value with the high-voltage glitch signal.

The final reverse voltage applied to the LEDs becomes lower. Conversely, the lower the VH value, the higher the superimposed reverse voltage. The risk of LED damage also increases.

According to industry general standards, the nominal reverse voltage tolerance of conventional LEDs is 5V. LED manufacturers derive this parameter based on standard test conditions.

But a lot of LED display manufacturer test data shows that long-term operation in a reverse voltage environment close to or exceeding 1.4V accelerates PN junction degradation. The probability of dead pixels rises sharply.

When the reverse voltage is strictly controlled below 1.4V, the LED service life can approach the theoretical design value. The dead pixel probability can drop below 0.1%.

Based on this test conclusion, the blanking voltage VH of the row drive tube cannot be too low to control LED reverse voltage. A reasonable VH value must provide reverse voltage protection for LEDs.

Engineering practice verification shows that VH generally needs to meet VH≥VCC-2V. This ensures the superimposed reverse voltage does not exceed the 1.4V safety threshold. It effectively guarantees long-term LED reliability.

Short-Circuit Caterpillar

LED short circuits are common failures in practical LED display applications. When an LED short-circuits, the corresponding column often stays on continuously. The industry vividly calls this phenomenon “short-circuit caterpillar.”

Its formation principle can be broken down as follows: When an LED in the middle of a row short-circuits, its anode and cathode conduct directly. This destroys the original normal drive loop.

When the scan signal switches to the faulty row, VLED forms an abnormal conduction path through the short-circuited LED. If the voltage difference between VLED and point A of the faulty column exceeds the forward lighting threshold of other normal LEDs in the same column, the entire column of LEDs will stay on continuously. This forms a long bright line like a “caterpillar.”

This failure seriously affects the integrity of the display image. Especially in black background or low-brightness display scenarios, the abnormal bright line is more obvious. It greatly damages the display effect.

To help engineers diagnose faults quickly, we need to clarify the core differences between short-circuit caterpillars and another common fault .

In terms of manifestation, short-circuit caterpillars are persistent faults. As long as the display is in scanning mode, the abnormal bright line appears continuously. It has nothing to do with image signals .

Open-circuit crosses are conditional faults. They only appear when the pixel with the open-circuited LED receives a lighting signal . Potential interference in the drive loop causes the cross-shaped faint glow. If the pixel is not lit, the fault does not appear.

In terms of root cause, short-circuit caterpillars come from LED short-circuit faults. Open-circuit crosses come from LED open-circuit faults. Their fault natures are completely different.

The core technical logic to solve short-circuit caterpillars is to cut off the abnormal conduction loop of the faulty column. The key is to make the voltage difference between VLED and blanking voltage VH smaller than the forward lighting voltage VF of the LED. That is, strictly meet VLED－VH < VF.

This is because when the voltage difference is smaller than the LED forward lighting threshold, even if an LED short-circuits, VLED cannot provide enough conduction voltage for other normal LEDs in the same column. This avoids the formation of the entire column bright line.

Note that the forward voltage VF varies significantly among different color LEDs. The VF range of red LEDs is 1.62.4V. For green and blue LEDs, it is 2.43.4V.

Industry tests find that red LEDs have a lower actual lighting threshold. They can glow faintly at 1.4V.

Taking the widely used red LED as an example for parameter adaptation: When VH > VCC－1.4V, the voltage difference between VLED and VH can be controlled below 1.4V. This completely cuts off the abnormal conduction loop and solves the short-circuit caterpillar problem.

When VH is in the range of VCC-2VVCC-1.4V, the voltage difference is between 1.4V2V. At this time, only one red LED below the short-circuit point will glow faintly (since the voltage is close to the lighting threshold). The overall fault impact is relatively controllable.

Open-Circuit Cross

In scanning LED display faults, open-circuit crosses are typical column-level faults alongside short-circuit caterpillars. Their formation relates closely to LED open-circuit faults and the parasitic capacitance coupling effect in the drive loop.

The specific formation process is as follows: During normal operation of the scanning screen, when an LED open-circuits and the system sends a lighting signal to the pixel with the open-circuited LED, the corresponding constant current drive channel OUT1 tries to output drive current.

But due to the LED open-circuit, the current cannot form a loop. This pulls the voltage of the OUT1 channel down to below 0.5V quickly (close to 0V in extreme cases).

This low-voltage signal produces a coupling effect through the parasitic capacitors C1 (between the channel and column line) and C2 (between the column line and row line) in the drive loop. It then affects the row-line parasitic capacitor Cr on the column side. This abnormally pulls down the potential across Cr.

The drop in Cr potential destroys the normal cut-off state of other LEDs in the same column. It causes other LEDs in the same row as the open-circuited LED to glow abnormally faintly. This ultimately forms an abnormal cross-shaped bright line, known as an open-circuit cross.

For example, when the blanking voltage VH of the scanning row is set to 3.5V, this voltage forms a significant potential difference with the pulled-down Cr potential. It further strengthens the conduction path. This makes the open-circuit cross phenomenon more obvious.

From a technical solution perspective, reducing the blanking voltage VH of the blanking row tube effectively eliminates open-circuit crosses. Specifically, VH needs to meet VH < 1.4V.

When VH is below 1.4V, even if the Cr potential is pulled down by the low-voltage signal of the open-circuited LED, it cannot form a potential difference with VH sufficient to conduct other LEDs. This cuts off the formation loop of the abnormal bright line.

Currently, some high-end row drive tubes in the industry adopt adjustable blanking voltage designs. They allow engineers to adjust the VH value flexibly according to actual fault conditions. Reducing VH below 1.4V can eliminate open-circuit crosses quickly.

But this solution has obvious performance trade-off issues. On one hand, reducing VH increases the reverse voltage across the LEDs. This accelerates the aging of the PN junction inside the LEDs and shortens their service life.

On the other hand, when VH is below 1.4V, it breaks the suppression condition for short-circuit caterpillars. This may cause the short-circuit caterpillar fault to recur.

Therefore, in practical engineering applications, this solution is usually used as an auxiliary measure. It needs to be combined with other technical measures for comprehensive use.

Large LED VF Value

In the mass production and practical application of LED displays, the dispersion of LED forward voltage VF values is unavoidable. When some LEDs have VF values exceeding the nominal range (i.e., large VF values), it may cause abnormal continuous lighting of columns.

This problem directly damages the integrity of the display image. It is a common application difficulty that troubles engineering technicians.

Taking the widely used green LED as an example, its nominal forward voltage VF range is 2.4~3.4V. But a lot of test data shows that the actual lighting threshold of some green LEDs is lower than the nominal lower limit.

When the voltage difference between the anode and cathode of the LED reaches 1.8V, it can conduct and glow faintly.

In this case, if the blanking voltage VH of the row drive tube is set too high, it will form a voltage difference exceeding 1.8V across the unscanned LEDs in the faulty column. This causes the column to stay on continuously. The abnormal phenomenon is more obvious in low-gray display scenarios.

To analyze the relationship between large LED VF values and VH settings more clearly, we conduct quantitative derivation using an LED with a nominal forward voltage VF1=3.4V as an example.

When the scan signal switches to the row where the LED is located, the output voltage VOUT of the constant current drive and the LED power supply voltage VLED1 turn on simultaneously.

According to the conduction characteristics of LEDs, the voltage at the channel end must meet VOUT=VLED1－VF1 to ensure the LED gets a stable drive current.

For other unscanned LEDs in the same column, the voltage difference V△ across them is determined by the difference between the blanking voltage VH and VOUT. That is, V△=VH－VOUT.

Substituting VOUT=VLED1－VF1 gives V△=VH－VLED1＋VF1.

Combined with the previous test conclusion, when V△>1.8V, even if the LED is not scanned and driven, it will glow faintly because the voltage difference exceeds the lighting threshold. This ultimately causes the column to stay on continuously.

In practical engineering applications, the voltage drop of the row drive tube is relatively small (usually below 0.2V). So, we can approximate VLED≈VCC.

Thus, we can further derive: When VH>VCC－1.6V, V△ will exceed the dangerous threshold of 1.8V. This is not conducive to solving the column continuous lighting problem caused by large LED VF values. This derivation conclusion provides a key quantitative basis for optimizing VH settings.

High-Contrast Coupling

High-contrast coupling is a typical performance bottleneck for small-pitch LED displays in high-image-quality scenarios. Its specific performance is as follows: When the display shows a low-brightness background image ,such as black or dark gray)with a superimposed high-brightness area ,such as white text or bright patterns, the part of the low-brightness background area in the same row as the high-brightness image appears obvious color shift or dimming. Red shift and dimming are the most common.

From a technical root cause perspective, this phenomenon comes from mutual interference between the column drive channel and the row drive tube. When displaying high-brightness images, the column drive channel outputs a large drive current.

This current produces dynamic voltage fluctuations in the drive loop. Through coupling paths such as parasitic capacitance and mutual inductance between row and column lines, it interferes with the drive signals of adjacent rows.

In low-brightness backgrounds, the drive current is small. The signal anti-interference ability is weak. This coupling interference is further amplified. It ultimately causes the display parameters of the corresponding area to deviate from the normal range, resulting in color shift or dimming.

From a technical optimization perspective, we can address high-contrast coupling from multiple dimensions. Optimization at the row drive level is an important direction.

Designing a clamping circuit to set a fixed clamping voltage keeps the row drive tube at this fixed level after discharge. This can reduce the blanking voltage VH and improve high-contrast coupling to a certain extent.

This is because reducing VH lowers the potential difference between row and column lines. It weakens the intensity of coupling interference and alleviates color shift or dimming.

But this optimization solution has obvious derivative problems. First, it causes dimming of short-circuited columns, affecting display integrity in fault states.

Second, it causes red shift in low-gray display, reducing low-gray image quality. Third, it further deteriorates the adaptability of LED VF values, increasing the risk of column continuous lighting faults.

If we only optimize the row drive module itself (without involving clamping circuit design) and reduce the blanking voltage VH, it can also alleviate high-contrast coupling. But this method repeats the above derivative problems.

It also increases the LED reverse voltage, accelerates LED aging, and may cause the short-circuit caterpillar problem to recur.

Therefore, optimizing high-contrast coupling at the row drive level requires multi-dimensional performance trade-offs. It usually needs to combine measures such as adjusting column drive chip parameters and designing anti-interference PCB wiring to form a comprehensive solution.

Optimal Selection of Row Drive Tube Blanking Voltage

Comprehensive technical analysis of the six core challenges above shows that selecting the blanking voltage VH of the row drive tube is a multi-objective optimization process. The VH value cannot be too high or too low. We need to find the best balance between various performance indicators.

Specifically, reducing VH can effectively solve problems such as upper ghosting and high-contrast coupling. But it worsens LED reverse voltage control, triggers short-circuit caterpillar faults, and increases the risk of column continuous lighting caused by large LED VF values.

Increasing VH can optimize LED reverse voltage protection and suppress short-circuit caterpillars. But it intensifies problems such as upper ghosting, open-circuit crosses, and high-contrast coupling.

It should be emphasized that among the six core challenges, reducing VH is not the preferred way to solve open-circuit crosses. The mainstream industry solution uses the open-circuit detection function of constant current drive chips to achieve accurate fault isolation.

This function can detect the open-circuit state of LEDs in real time. When an open-circuit fault is found, it immediately cuts off the drive signal of the corresponding channel to avoid potential coupling interference.

Relying solely on reducing VH to solve open-circuit crosses significantly reduces the long-term reliability of LEDs. It does not meet the long-life design requirements of displays.

To help engineers quickly determine the VH value in practical design and selection, the following table summarizes the adaptation range, core constraints, and precautions of blanking voltage in various application scenarios.

Based on the above multi-dimensional performance trade-off analysis and the mainstream VCC=5V power supply scheme in the industry, setting the blanking voltage VH between 3V~3.4V is a reasonable choice that balances various performance requirements.

Quantitative verification shows that this range meets the constraint VH≥VCC-2V (3V≥5V-2V). It ensures the LED reverse voltage is controlled within the 1.4V safety threshold.

At the same time, it is close to the critical value of VH>VCC－1.4V (3V>5V-1.4V=3.6V is not true; 3.4V is close to 3.6V). It can effectively suppress short-circuit caterpillar faults. Only in extreme cases may a single LED glow faintly, with controllable impact.

In addition, this range can alleviate upper ghosting and high-contrast coupling to a certain extent. It is also relatively friendly to the adaptation of large LED VF values.

Practical application verification shows that this VH range can adapt to the design requirements of most P1.0~P2.5 scanning modules. It can effectively alleviate the above various application problems.

While ensuring the display quality of LED displays , it also balances the long-term reliability of LEDs. It has high engineering practical value.