This section presents the control loop compensation design procedure for the LM5176 buck-boost controller. The LM5176 operates mainly in buck or boost modes, separated by a transition region, and therefore, the control loop design is done for both buck and boost operating modes. Then, a final selection of compensation is made based on the mode that is more restrictive from a loop stability point of view. Typically, for a converter designed to go deep into both buck and boost operating regions, the boost compensation design is more restrictive due to the presence of a right half plane zero (RHPZ) in the boost mode.

The boost power stage output pole location is given by:

Equation 38.

where R_OUT = 2 ? corresponds to the maximum load of 6 A.

The boost power stage ESR zero location is given by:

Equation 39.

The boost power stage RHP zero location is given by:

Equation 40.

where D_MAX is the maximum duty cycle at the minimum V_IN.

The buck power stage output pole location is given by:

Equation 41.

The buck power stage ESR zero location is the same as the boost power stage ESR zero.

It is clear from Equation 40 that RHP zero is the main factor limiting the achievable bandwidth. For a robust design, the crossover frequency should be less than 1/3 of the RHP zero frequency. Given the position of the RHP zero, a reasonable target bandwidth in boost operation is around 4 kHz:

Equation 42.

For some power stages, the boost RHP zero might not be as restrictive. This happens when the boost maximum duty cycle (D_MAX) is small, or when a really small inductor is used. In those cases, compare the limits posed by the RHP zero (f_RHP / 3) with 1/20 of the switching frequency and use the smaller of the two values as the achievable bandwidth.

The compensation zero can be placed at 1.5 times the boost output pole frequency. Keep in mind that this locates the zero at three times the buck output pole frequency which results in approximately 30 degrees of phase loss before crossover of the buck loop and 15 degrees of phase loss at intermediate frequencies for the boost loop:

Equation 43.

If the crossover frequency is well below the RHP zero and the compensation zero is placed well below the crossover, the compensation gain resistor, R_c1, is calculated using the approximation:

Equation 44.

where D_MAX is the maximum duty cycle at the minimum V_IN in boost mode and A_CS is the current sense amplifier gain. The compensation capacitor C_c1 is then calculated from:

Equation 45.

The standard values of compensation components are selected to be R_c1 = 10 k? and C_c1 = 33 nF.

A high frequency pole (f_pc2) is placed using a capacitor (C_c2) in parallel with R_c1 and C_c1. Set the frequency of this pole at seven to ten times of f_bw to provide attenuation of switching ripple and noise on COMP while avoiding excessive phase loss at the crossover frequency. For a target f_pc2 = 28 kHz, C_c2 is calculated using Equation 46:

Equation 46.

Select a standard value of 560 pF for C_c2. These values provide a good starting point for the compensation design. Each design should be tuned in the lab to achieve the desired balance between stability margin across the operating range and transient response time.