7.3.1 Pulse Level Modulation (PLM) Control

A proprietary Pulse-Level-Modulation (PLM) control method is used in the TPS92511.  It can regulate the average LED current by sensing only the inductor current at the on-period (Figure 13).  The integrated MOSFET and the sensing and control circuits in the TPS92511 implement the whole PLM control internally so the control does not suffer from tolerance and noise issues that may be coming from external components.  As compared with the conventional method which regulates average LED current by sensing the current over the entire switching cycle, the power dissipation on the sensing circuit in PLM is much lower.  For example, consider a duty cycle of 0.5, the power dissipation on current sensing in PLM can be reduced by half. PLM requires no external loop compensation circuit. Besides, the accuracy of the regulated LED current is high (typically ±3.5% in the TPS92511).

Figure 13. Waveforms of a Floating Buck LED Driver with PLM

7.3.2 Pulse Level Modulation (PLM) Operaion Principles

The Pulse-Level-Modulation is a patented method to ensure an accurate average output current regulation without the need of direct output current sensing. Figure 13 shows the current waveforms of a typical buck converter under steady state, where, I_L1 is the inductor current and I_LX is the current flowing into the LX pin. For a buck converter operating in steady state, the mid-point of the RAMP portion of I_L1 equals to the average value of I_L1 and hence the average LED current I_LED(avg). In short, by regulating the mid-point with respect to a precise reference level, PLM achieves LED current regulation by sensing the main MOSFET current solely, instead of the entire cycle of I_L1.

7.3.3 PLM Control enable Common-Anode Low-Side Sensing (CALS)Technique to Save Wiring

For multi-channel systems with separated driver boards and LED array boards, the Pulse-Level-Modulation (PLM) control scheme enable Common-Anode Low-Side Current Sensing to save inter-board wirings. Figure 14 shows a conventional configuration with a Low-side switching and High-Side Current Sensing. For an n channel system with separated driver and an LED array boards, 2n inter-board wirings are required. For example, an 128-channel system needs 256 inter-board wirings, which implies a high material and manufacturing cost. Figure 15 shows the PLM configuration with Low-side switching and Low-Side Current Sensing. A Common-Anode configuration is used for the LED array board. As shown in the figure, an n channel system with separated driver and LED array boards requires only n+1 inter-board wirings. For an 128-channel system, only 129 inter-board wirings are required.  The wiring cost is cut by half, and the cost of the end product can be reduced.

Figure 14. Conventional Configuration with Low-Side Switching and High-Side Current Sensing Requires 2×n Inter-Board Wirings

Figure 15. PLM Configuration with Common-Anode Low-Side Switching Requires n+1 Inter-Board Wirings

7.3.4 Internal Regulator

The TPS92511 integrates an internal voltage regulator for powering internal circuitry. For stability, an external capacitor C_VCC of at least 1 μF should be connected between the VCC and PGND pins The output of the internal regulator V_CC is 5.4V when V_IN is larger than 6V. If V_IN is lower than 6V, V_CC decreases. The TPS92511 will trigger the VCC under-voltage lock-out if V_CC falls below typically 3.5V. V_CC can be used to bias external circuits subject to a loading of maximum 2 mA, while it has a short circuit current limit at typically 16 mA.

7.3.5 Setting The Switching Frequency

The switching frequency f_SW of the TPS92511 is programmable in the range of 50 kHz to 500 kHz by a single resistor R_FS connecting the FS pin and ground. The following equation shows the relationship between f_SW and R_FS:

Equation 1.

Figure 16 plots f_sw against R_FS. Table 1 shows values of R_FS for commonly used switching frequencies.

Figure 16. Switching Frequency vs R_FS

7.3.6 Setting The LED Current

The LED current I_LED of the TPS92511 is programmable by a single resistor R_IADJ connecting the IADJ pin and ground. The IADJ pin is internally biased to 1.25 V. Equation 2 shows the relationship between I_LED and R_IADJ:

Equation 2.

To ensure stability, R_IADJ must be less that 30 kΩ, implying a minimum I_LED of 50 mA can be programmed.  The tolerance of I_LED of 150 mA is shown in the ELECTRICAL CHARACTERISTICS.  Larger tolerance should be expected for lower I_LED. Figure 17 plots I_LED against R_IADJ. Table 2 shows values of R_IADJ for commonly used I_LED.

Figure 17. LED Current vs _RIADJ

7.3.7 Integrated MOSFET

The TPS92511 integrates a N-channel power MOSFET, the drain of which is connected to the LX pin. When the integrated MOSFET is turned on, the resistance across the LX and GND pins is typically 1.4Ω. The integrated MOSFET has a fixed current limit of 1.2A to protect the application circuit during critical operation conditions like short circuit of the LED string. Once the limit is hit, the integrated MOSFET turns off immediately for 34 µs to let the inductor discharge.

The minimum on-time of the integrated MOSFET is 400 ns. It may be hit at a high switching frequency and a high V_IN/V_LED ratio. Once hit, the I_LED regulation may be affected. In the worst case, I_LED may be boost up to a level higher than the programmed value, and the LED string and/or the inductor may be damaged as a result. Hence, it is recommened that the ratio between V_IN and V_LED should be designed under the following constraint:

Equation 3.

7.3.8 Inductor Selection

Operating in the continuous conduction mode (CCM) is required in the TPS92511 application circuit. In the CCM, considering the on-period, the peak-to-peak inductor current ripple (2ΔI_L1) is shown in Equation 4.

Equation 4.

Because

Equation 5.

L₁ can be a function of V_IN, V_LED, f_SW and ΔI_L1 as shown in Equation 6 .

Equation 6.

The value of L₁ is selected by designers with the consideration of all above parameters. The minimum L₁ calculated by the following equation is a good starting point for designing L₁:

Equation 7.

The following table shows some typical examples of using R_FS and R_IADJ to estimate the minimum L₁:

To maintain the CCM, ΔI_L1 must be smaller than the average LED current I_LED(avg). Hence, the minimum inductance used is:

Equation 8.

In the absence of output capacitors, the TPS92511 can maintain a continuous I_LED throughout the entire switching cycle because in such case the inductor current is the same as I_LED (floating buck topology operating in the CCM). However, the LED peak current must not exceed the rated current of the LED. The peak LED current can be found by the following equation:

Equation 9.

7.3.9 Integrated MOSFET Current Limit

The current limit of the integrated MOSFET is internally fixed at 1.2A to protect the LED string, the inductor and the integrated MOSFET from overdriven.  Once triggered, the integrated MOSFET turns off immediately for 34 µs to let the inductor to discharge.  The triggering of the current limit cycles repetitively until all overdriven conditions disappear.

7.3.10 PWM Dimming Control

The TPS92511 implements PWM dimming by applying a PWM dimming signal to the DIM pin. A low input applying to the DIM pin disables the switching of the integrated MOSFET, and as a result discharges the inductor and then turns off the LED string. To turn on the LED string, the DIM pin should be connected to high or left open (since it is internally pulled high by a current of typically 40 µA and 90 µA when the DIM pin is low and high respectively). The PWM dimming frequency is recommended to be lower than 0.1f_SW to ensure normal operation.

7.3.11 Analog Dimming

Analog dimming can be implemented by injecting a current to R_IADJ (Figure 18) and as a result reduces the current of the IADJ pin, I_ADJ, which is controlled internally by the TPS92511 to bias the voltage on the IADJ pin to be 1.25V.  If the CCM can be maintained, the minimum I_ADJ can achieve 15 µA, which refers to an I_LED of 18 mA.  If I_ADJ is further decreased, I_LED may not follow due to the presence of the minimum on-time of the integrated MOSFET.  If the CCM cannot be maintained, I_LED can still decrease monotonically with I_ADJ.  However, if good line and load regulations are required, the CCM should be maintained by using a large inductance.

Figure 18. Circuit Configuration for Analog Dimming

7.3.12 High Voltage Buck Configuration

The TPS92511 can handle applications with an input voltage higher than 65V, which is the maximum V_IN of the recommended operating condition of the TPS92511, by adding an external high voltage N-channel MOSFET to the application circuit as shown in Figure 19.  PWM dimming can be implemented in this circuit without additional efforts, and analog dimming is also feasible by referencing to additional circuits shown in Figure 18.

Figure 19. Circuit Configuration for Very High Voltage Buck

7.3.13 Thermal Foldback

Thermal foldback is useful to prevent over-temperature of LEDs during operation by sensing the temperature of LEDs and, if the sensed temperature is high, reducing I_LED to decrease the power and as well as the temperature of LEDs.  Thanks to the feature of analog dimming, thermal foldback can be implemented by embedding a negative temperature coefficient (NTC) resistor, R_NTC, into a circuit as shown in Figure 20.  When the sensed temperature increases, R_NTC decreases and thus the emitter current of Q_T1 increases to reduce I_LED by means of analog dimming.  The resistor R_TF can adjust the loop gain of the thermal foldback control loop, which should be high enough to avoid oscillation and maintain stability.

Figure 20. Circuit Configuration for Thermal Foldback

7.3.14 EMI Consideration

Conductive and radiative EMI can be major concerns for lighting applications.  The TPS92511 application circuit can be designed for the EN 55022 class B standard by adding a few external components, as shown in Figure 21.  The input filter which consists of an inductor L₂ and two capacitors C_IN2 and C_IN3 takes care of the conductive EMI, while the output capacitor C_LED and the ferrite bead FB₁ which inserts between the LX pin and D₁ take care of the radiative EMI.

Figure 21. Circuit Configuration with EMI Design Consideration

7.4.1 Operation with V_IN < 4.5 V (minimum V_IN)

For the typical application circuit, when the input voltage drops so that the VCC voltage regulator is under drop-out mode, and the VCC voltage drops below the “VCC UVLO Lower Threshold” (typically 3.48V), the switching of the main MOSFET is stopped, and the LED current will become zero. At the same time, the voltages of both the FS and IADJ pins will become zero .

When the input voltage increases from zero and the VCC voltage is increased to cross over the “VCC UVLO Upper Threshold” (typically 3.75V), the voltages on the FS and IADJ pins will rise to their regulation voltage (typically 1.25V), the switching of the main MOSFET is started upon the DIM pin voltage is HIGH, and the LED current will ramp up to its preset value set by R_IADJ.

7.4.2 Operation with DIM control

For the typical application circuit, when the VCC voltage is not under UVLO condition, the switching of the main MOSFET is enabled and the LED current is conducted if the DIM pin voltage is higher than the “DIM Pin Upper Threshold” (typically 1V).

Alternaltively, the switching is disabled and the LED current is cut off if the voltage of the DIM pin is lower than the “DIM Pin Lower Threshold” (typically 0.675V).

7.4.3 Linear Mode

When the VCC voltage is not under UVLO condition and the voltage on the FS pin is forced to be higher than 4.2V but lower than 5V, the switching of the main MOSFET is disabled, and the TPS92511 is working in the Linear Mode. In the Linear Mode, if the voltage on the DIM pin is higher than the “DIM Pin Upper Threshold” (typically 1V), the TPS92511 will regulate the LX pin in-going current according to the preset value set by R_IADJ. Alternatively, if the voltage on the DIM pin is lower than the “DIM Pin Lower Threshold” (typically 0.675V), the LX pin will open and its in-going current will become zero.

Below is the simple configuration to have the TPS92511 working as a linear current shunt regulator.

Figure 22. Circuit Configuration for Working as a Linear Current Shunt Regulator