10.2.1.2.1 Inductor Selection

A boost converter requires two main passive components for storing energy during the conversion. A boost inductor and a storage capacitor at the output are required.

The TPS6131x device integrates current-limit protection circuitry. The valley current of the PMOS rectifier is sensed to limit the maximum current flowing through the synchronous rectifier and the inductor. The valley peak current limit (1250 mA or 1750 mA) is user selectable through the I²C interface.

To optimize solution size the TPS6131x device is designed to operate with inductance values from a minimum of 1.3 µH to a maximum of 2.9 µH. TI recommends a 2.2-µH inductance in typical high-current white LED applications.

The highest peak current through the inductor and the power switch depends on the output load, the input and output voltages. The maximum average inductor current and the maximum inductor peak current can be estimated using Equation 2 and Equation 3:

Equation 2.

Equation 3.

where

• f = switching frequency (2 MHz)
• L = inductance value (2.2 µH)
• η = estimated efficiency (85%)

The losses in the inductor caused by magnetic hysteresis losses and copper losses are a major parameter for total circuit efficiency.

10.2.1.2.2 Input Capacitor

For good input-voltage filtering, TI recommends low ESR ceramic capacitors. TI recommends a 10-µF input capacitor to improve transient behavior of the regulator and EMI behavior of the total power supply circuit. The input capacitor must be placed as close as possible to the input pin of the converter.

10.2.1.2.3 Output Capacitor

The major parameter necessary to define the output capacitor is the maximum allowed output-voltage ripple of the converter. This ripple is determined by two parameters of the capacitor, the capacitance and the ESR. It is possible to calculate the minimum capacitance required for the defined ripple, supposing that the ESR is zero, by using Equation 4:

Equation 4.

where

• f is the switching frequency
• ΔV is the maximum allowed ripple

With a chosen ripple voltage of 10 mV, a minimum capacitance of 10 µF is required. The total ripple is larger due to the ESR of the output capacitor. This additional component of the ripple can be calculated using Equation 5:

Equation 5.

The total ripple is the sum of the ripple caused by the capacitance and the ripple caused by the ESR of the capacitor. Additional ripple is caused by load transients. This means that the output capacitor must completely supply the load during the charging phase of the inductor. A reasonable value of the output capacitance depends on the speed of the load transients and the load current during the load change.

For the standard current white LED application, a minimum of 3-µF effective output capacitance is usually required when operating with 2.2-µH (typical) inductors. For solution size reasons, this is usually one or more X5R or X7R ceramic capacitors.

Depending on the material, size and therefore margin to the rated voltage of the used output capacitor, degradation on the effective capacitance can be observed. This loss of capacitance is related to the DC bias voltage applied. Therefore, TI recommends checking that the selected capacitors show enough effective capacitance under real operating conditions.

10.2.1.2.4 NTC Selection

The TPS6131x requires a negative thermistor (NTC) for sensing the LED temperature. Once the temperature monitoring feature is activated, a regulated bias current (approximately 24 µA) is driven out of the TS port to produce a voltage across the thermistor.

If the temperature of the NTC-thermistor rises due to the heat dissipated by the LED, the voltage on the TS input pin decreases. When this voltage goes below the warning threshold, the LEDWARN bit in REGISTER6 is set. This flag is cleared by reading the register.

If the voltage on the TS input decreases further and falls below hot threshold, the LEDHOT bit in REGISTER6 is set and the device goes automatically in shutdown mode to avoid damaging the LED. This status is latched until the LEDHOT flag gets cleared by software.

The selection of the NTC-thermistor value strongly depends on the power dissipated by the LED and all components surrounding the temperature sensor and on the cooling capabilities of each specific application. With a 220-kΩ (at 25°C) thermistor, the valid temperature window is set from 60°C to 90°C. The temperature window can be enlarged by adding external resistors to the TS pin application circuit. To obtain proper triggering of the LEDWARN and LEDHOT flags in noisy environments, the TS signal may require additional filtering capacitance.

Figure 57. Temperature Monitoring Characteristic

10.2.1.2.5 Checking Loop Stability

The first step of circuit and stability evaluation is to examine these signals from a steady-state perspective:

• Switching node (SW)
• Inductor current (I_L)
• Output ripple voltage (V_OUT(AC))

These are the basic signals that must be measured when evaluating a switching converter. If the switching waveform shows large duty-cycle jitter or the output voltage or inductor current shows oscillations, the regulation loop may be unstable. This is often a result of improper board layout or L-C combination.

As a next step in the evaluation of the regulation loop, test the load transient response. VOUT can be monitored for settling time, overshoot or ringing that helps judge the converter's stability. With no ringing, the loop usually has more than 45° of phase margin.

Because the damping factor of the circuitry is directly related to several resistive parameters (MOSFET r_DS(ON)) that are temperature dependant, the loop stability must be analyzed over the input voltage range, output current range, and temperature range.

10.2.1.2.6 LED Flash Current Level Optimization Versus Battery Droop

In cell phone applications, the camera engine is normally specified over an operating temperature down to 0°C or –10°C. To achieve a reliable system operation, the LED flash current must be rated according to the maximum tolerable battery voltage drop, highest battery impedance and lowest ambient temperature.

To dynamically optimize the LED flash current (light output) versus battery state-of-charge and temperature, we could consider the self-adjustment procedure in Figure 58. This algorithm could be embedded into the auto-exposure, auto white-balance, or red-eye reduction pre-flash algorithms.

Figure 58. Image Capture Sequence

• Phase 1: Pre-Flash, Battery Impedance Estimation

The battery voltage usually drops by a few hundreds of millivolts during a high-power flash strobe. For short durations, this voltage droop should not be subject to the battery intrinsic capacitance (relaxation effect) but rather to its cell impedance.

Based on the state of the Tx-MASK input, the battery voltage drop, during pre-flash, and the LED current level, the base-band processor can compute an estimated cell-impedance value (ESR).

Depending on the ambient temperature, the battery state-of-charge (SoC), the flash (capture) duration and the actual status of the various RF interfaces, the base-band processor can determine a safe battery voltage droop, to be tolerated during the forthcoming strobe sequence, as well as a maximum flash current rating. The maximum flash current setting can be estimated by considering nominal LEDs and approximately 85% power efficiency in the driver.

• Phase 2: Battery Loading Monitoring Before Image Capture

For a reliable system operation, the base-band processor must make sure that no 'parasitic' high-current load suddenly impacts the budgeted battery voltage sag. The most critical timing is referenced as t_CRITICAL.

The interrupt subroutine, running on the base-band processor, must be ready to detect any 'parasitic' battery load event that could occur before the image capture (see REGISTER3 (address = 0x03) for SFT bit description). In such a situation, the battery voltage droop budget and the maximum LED current settings would need to be revised.

Figure 59. 900-mAh, Li-Ion Battery Transient Response vs SoC and Temperature