Figure 3-16 shows the root cause Analysis of the brighter coupling. The left sub-figure shows the whole capacitor modeling of matrix LED display. Apart from parasitic capacitors on channels and scan lines, this modeling also considers parasitic capacitors across the LED, which is crucial to analyze the coupling issue.

The middle sub-figure shows the analysis when LED₀₁ is turned ON. LED₀₁ ON means channel OUT₁ and SW₀ are chosen. As shown in the green shadow block, this is called the chosen line. For the rest of the lines shown in the red shadow block, this is called unchosen lines.

The chosen line is connected to ground and the unchosen lines are floating. In addition, since channel OUT₀ is closed, the anode of LED₀₀ is also floating.

For further equivalence, simplify the capacitor modeling of all unchosen lines to 3 capacitors shown in the upper right sub-figure. The 2 capacitor C_{a_N-1} and C_{b_N-1} in blue represent LED’s parasitic capacitors. The capacitor C_{sw_N-1} in yellow represents SCAN lines’ parasitic capacitors.

Since the fact that the parasitic capacitor of SCAN line is much less than that of LED, this can further simplify the modeling as shown in the lower right sub-figure. With this simplified modeling, now the root cause of brighter coupling becomes clearer. When the LED₀₁ is lit up as the current path in red dotted line, the voltage on the anode of LED₀₁ is increased a delta. The voltage across the capacitor can not suddenly change. So, the voltage on the anode LED₀₀ is coupled to increase another delta. This makes the LED₀₀ slightly light up, as shown in the current path in blue dotted line.

Figure 3-16 Root Cause Analysis of Brighter Ghosting

Figure 3-17 shows a brighter coupling simulation example in spice. From the simulation results, we can see that OUT₀ (blue cure) is coupled by OUT₁(green curve) and ramps from 0 to around 2.6V. That’s larger than the LED forward voltage and can make this light up. The peak current through the LED₀ is around 400uA.

Figure 3-17 Brighter Coupling Simulation

So, how to solve the brighter coupling issue? From the previous modeling, the unchosen lines are floating, which means the unchosen lines are easily to be affected by variable voltage. Supposing we can make voltage on the unchosen lines be a certain level, what happens? In fact, this is the design shown in Figure 3-18. This can make the unchosen lines voltage be upper clamped to a fixed voltage to isolate the Switch ON and Switch OFF channels, as the SW line ceiling circuit block shown. Now we can see that the voltage on anode of LED₀₀ is keeping constant and not coupled.

TLC698x can set the line ceiling voltage (upper line clamp voltage) that clamps the maximum voltage level on scan lines to maintain voltage across the unwanted LED is smaller than the LED forward voltage. thus, eliminating the brighter coupling issue.

Figure 3-18 Brighter Coupling Elimination by SW Line Ceiling Circuit

Line ceiling is different from SW line clamp that we have discussed in previous sections for upside ghosting. To distinguish, we also call line ceiling as upper line clamp, and SW line clamp as lower line clamp, shown in Figure 3-19.

Figure 3-19 Line Ceiling and Line Clamp

Figure 3-20 shows the line ceiling simulation example in spice. From the simulation results, we can see that the clamping method works and LED₀ is not coupled to light up.

Figure 3-20 Line Ceiling Simulation

Figure 3-21 and Figure 3-22 show the comparison demo with or without brighter coupling. The bright horizontal grid lines in area C couples the corresponding lines in areas A and B to be brighter than the other lines (such as area D). The dark area A an B are coupled by the high grayscale area C to be brighter. With this being enabled, we don’t see coupling anymore.

Figure 3-21 With Brighter Coupling

Figure 3-22 Without Brighter Coupling