DLPS039F December 2015 – April 2019 TPS99000-Q1
As seen in Figure 12, the actual LED current pulse is distorted due to the rising (Tr) and falling (Tf) edges rates not being equal, and/or the turn-on (Tp1) and turn-off (Tp2) propagation delays not being equal. The rising edge turn-on time of the current pulse is primarily a function of the voltage across the inductor and the desired current, plus the inductor current initial condition. This distortion causes both the time attenuation and amplitude attenuation of the pulse to become non-linear functions of the control settings. This can lead to image artifacts.
Blanking time is the period of no light output in between two LED segments. The inductor current during this time is called blanking current. This current is controlled to provide an optimized Tr and Tf.
Blanking current control reduces image artifacts by preventing light overshoot and undershoot.
The blanking current time periods are split into two parts. The first is a dissipation phase where the residual current in the inductor from the previous light pulse is reduced using a dissipative shunt. The second phase is a non-dissipative (low series resistance) shunt phase, where the inductor is charged up to the appropriate current for the next light pulse before current is applied to the LED. This process is illustrated in Figure 13.
During the first phase of the blanking time, shunt 2 (S_EN2) is enabled while the LEDs are disconnected. This places a load with a higher effective resistance in place of the LEDs. The residual energy in the inductor is dissipated into this load and the inductor current decreases rapidly. Without this feature, a high current in one pulse could cause excessive brightness in the next pulse.
During the second phase of the blanking time, the LED driver charges the inductor through a short circuit shunt (S_EN1). Charging continues until the peak current limit is reached. The peak current limit is set by the ILIM DAC. The peak current limit setting is coordinated by DLPC230-Q1 software to match the expected operating current during photo feedback operation. (The expected current level is determined from ADC measurements of LED current during prior frames.) When the blanking current time period is over, the S_EN1 short circuit shunt is turned off, the next LED is enabled, the DRV_EN signal is toggled, and the system reverts to photo feedback, hysteretic operation. Because the inductor is pre-charged to the ideal current and the system capacitance is low, light output rising edge is extremely fast, and the transition to stable hysteretic control is nearly immediate. This results in a more rectangular pulse. An illustration of the current paths is shown in Figure 14.
Precise control of the LED pulse shape results in greater dimming range, more display bit depth, and better color and gray ramp accuracy.