Actually, the root cause analysis of upside ghosting is similar to what has been discussed in the downside ghosting session. The main cause of upside ghosting is the buildup charge in the parasitic capacitance on SW_n when transitioning between scan lines, the parasitic capacitance is charged through the LED causing this to light up when this must not be. Figure 3-9, Figure 3-10 and Figure 3-11 describe the root cause analysis of upside ghosting.

So, in P1 operation, when only wants to light up LED₀₂, the driver turns ON SW₀. And at this moment, the yellow parasitic capacitor on Scan Line 0 is discharged to 0, through these current paths in red dotted line.

During P1 operation, SW₀ is ON, LED₀₂ lights up.

Figure 3-9 Upside ghosting - P1 Operation

Next, in P2 operation, when only wants to light up LED₁₁, the driver turns on SW₁ and the yellow parasitic capacitor on Scan Line 1 is discharged to 0, through these current paths in red dotted line. However, at this moment, the yellow parasitic capacitor on Scan Line 0 is also charged, through the current path in blue dotted line. So, LED₀₁ is slightly light up even SCAN line 0 is closed. Because the parasitic capacitor on SCAN line 0 has been discharged previously, and LED₀₁ is light up through the current path in blue dotted line. So, LED₀₁ is calling upside ghosting.

During P2 operation, SW₁ is ON, LED₁₁ lights up, LED₀₁ slightly lights up.

Figure 3-10 Upside ghosting - P2 Operation

Similarly, in P3 operation, when only wanting to light up LED₂₀, we can see that LED₁₀ is slightly lit up and LED₀₀ is even so slightly lit up, through the current paths in blue and orange dotted line, respectively. The reason is the same as previously due to the scan line parasitic capacitors.

The reason that LED₁₀ and LED₀₀ have different brightness is because the corresponding scan line parasitic capacitors accumulate different charges before the LED being faulty lighted up (different quantities of charge flow through the LED to charge the corresponding scan line parasitic capacitor).

During P3 operation, SW₂ is ON, LED₂₀ lights up, LED₁₀ slightly lights up, LED₀₀ ever so slightly lights up.

Figure 3-11 Upside ghosting - P3 Operation

So, how to solve this issue? The upside ghosting is because of discharges on SCAN line parasitic capacitors. So, if we can stop or weaken the discharge process, then we can solve this issue.

As shown in Figure 3-12, TLC698x uses scan line clamp circuits to pull up the line voltage to a high value (Vclamp) during the line switch time to maintain no current flow to the previous scan line, thus eliminating upside ghosting.

Figure 3-12 Upside Ghost Effect Elimination by SW Line Clamp Circuit

Figure 3-13 shows the comparison demo with or without upside ghosting. We can see that the clamp circuit works very well. With this being enabled, we do not see ghosting anymore.

Figure 3-13 Comparison Demo With or Without Upside Ghosting