Understand Power Supply Loop Stability and Loop Compensation —Part 3: Loop Design in Three Simple Steps
Understand Power Supply Loop Stability and Loop Compensation —Part 3: Loop Design in Three Simple Steps
著者
Henry Zhang
2025年06月08日
Read other articles in this series.
要約
Part 3 of this article series explains a simple method for designing loop compensation in current-mode controlled switch-mode power supplies. This control architecture is extensively used in power management solutions, including many of ADI’s power products. It enables the use of a simple Type 2 compensation network to design and optimize the supply feedback loop, ensuring rapid transient responses and robust stability margins. This article introduces fundamental loop design concepts, offers clear explanations of the Type 2 compensation network, and examines the role of each compensation component. The loop design process is then streamlined into three straightforward steps. Additionally, the LTpowerCAD® design tool further simplifies the loop design and optimization process.
Introduction—Basic Concepts
Switch-mode power supplies are extensively used in modern electronic systems to achieve high efficiency and power density. For less experienced system engineers, power supply loop compensation design optimization can be a critical yet challenging task. Most of ADI’s switch-mode regulators employ current-mode control architecture to achieve high performance and high reliability. For example, Figure 1 shows the basic feedback loop block diagram of a popular current-mode control step-down buck converter.1 This architecture features an inner current sensing loop and an outer output voltage regulation loop. The inner current sense loop forces the inductor current to follow the compensation network output voltage at the ITH node. In this way, conceptually the inductor becomes a current source controlled by the voltage loop error amplifier output VITH. Consequently, the buck converter power stage including the current loop behaves as a single-pole system at lower frequencies below the current loop bandwidth. Therefore, a simple Type 2 compensation network is adequate to optimize supply loop stabilities and transient performances. The example Type 2 compensation network is shown in Figure 1 as the RTH, CTH, and CTHP network on the error op amp output ITH pin.
Figure 2 shows the conceptual power supply loop gain diagram. KREF is the feedback resistor-divider network gain from supply output VO to the FB pin. A(s) is the voltage loop compensation error amplifier network gain from the FB pin to the ITH pin. GCV(s) is the power stage transfer function from error amplifier output node VITH to power supply output voltage Vo, including the inner current loop. Therefore, the total power supply loop gain T(s) is given by Equation 1:
Switching Power Supply Loop Design and Optimization Goals
An optimized power supply loop should be designed with high loop bandwidth to achieve fast transient response while maintaining sufficient stability margins. Furthermore, for a switching power supply, it is important to attenuate the switching noises in the feedback loop to minimize switching waveform jitters. In summary, here are key supply loop design targets:
- Loop bandwidth (fBW): While high loop bandwidth is generally desirable for fast transient response, it is practically limited by the switching frequency (fSW). Typically, the maximum bandwidth is set up to 1/10 or 1/5 of fSW.
- Phase margin: A phase margin greater than 45° is usually required, with a margin greater than 60° phase margin being recommended.
- Gain margin: Gain margin, defined as the gain attenuation where the loop phase is –180°, should be at least 8 dB to 10 dB.
- Switching noise attenuation margin: For a current-mode control switching supply, it is important to attenuate switching noises in the feedback loop to minimize jittering of the switching node waveform. Practically, greater than 8 dB attenuation at fSW/2 is preferred.
Intuitive Understanding of Type 2 Compensation Network
To design and optimize the compensation network, a power supply designer first needs to understand the effect of each compensation R or C value on the loop gain and load transient response. Figure 3 shows the Type 2 compensation network, including a typical transconductance error amplifier (that is, a voltage-controlled current source) with a gain of gm, the amplifier parasitic output resistance R0, and the compensation network including RTH, CTH, and CTHP. These three key ITH pin R/C components are to adjust the compensation gain A(s) and therefore determine the supply loop gain bandwidth, stability margins, and transient response performances.
The compensation gain A(s) is defined as shown in Equation 2.
Since the internal gm and R0 are fixed for a given controller IC, the impedance of the RTH, CTH, and CTHP network ZITH (s) determines the compensation gain A(s). Detailed equation derivations are given in ADI application note AN149 (see Equation 3): 1
where
For power supply designers, instead of memorizing this A(s) equation, the compensation gain should be understood in an intuitive way by observing how the ZITH impedance changes vs. the frequency shown in the conceptual plot in Figure 4.
Figure 4 conceptual plot can be explained below from left to right, as the frequency increases from lower to higher values through the following ranges:
Range 1: From DC to low frequency pole fP1, all capacitors are treated as open circuits with high impedance. Therefore, the ZITH magnitude is determined by the error amplifier parasitic output resistance RO, which is usually a very large value (~MΩ). In this range, the error amplifier A(s) has a flat high DC gain equal to gm × R0.
Range 2: As the frequency increases, the first CTH capacitor impedance drops (note that CTH>>CTHP, so the CTHP impedance is still very high). At frequency fP1, the CTH impedance magnitude is comparable to RO. After fP1, as the frequency further increases, the CTH impedance determines the total ZITH impedance.
Range 3: As the frequency further increases, the CTH impedance magnitude eventually drops to the value close to the RTH value in the series path. After that, the RTH value dominates the total ZITH value and keeps it flat in this range. The corner frequency where the CTH and RTH impedance values are close to each other is defined as the zero frequency fZ1. The target supply loop bandwidth fBW is usually set within frequency range 3
Range 4: As the frequency increases even more, eventually the smaller parallel CTHP impedance drops to the value comparable to the RTH value. After that, the ZITH magnitude is determined by the CTHP impedance. The corner frequency where the RTH and CTHP impedance are close is defined as the second HF pole fP2. When needed, this high frequency pole should be located below the supply switching frequency fSW to attenuate switching noises.
With this explanation, it shows the ZITH impedance amplitude is determined by each different R or C component at different frequency ranges. This observation helps us to understand the effect of each component on the supply loop gain and transient response.
How Does Each Compensation Component Affect Loop Gain and Transient Response?
(1) Compensation Resistor RTH
Usually, the compensation network is designed so that the power supply loop bandwidth fBW is located between zero fZ1 and pole fP2 in frequency range 3. In this range, the ZITH magnitude is determined by the RTH value. So, the RTH value directly determines the power supply loop bandwidth fBW. Figure 5 illustrates that the increased RTH value can increase the compensation gain A(s) between fZ1 and fP1. Therefore, larger RTH leads to higher loop bandwidth, as shown in Figure 6a. Higher loop bandwidth usually can reduce supply VOUT undershoot and overshoot amplitude during load transient, as shown in Figure 6b, without much change in the VOUT settling time after a transient event.
(2) Compensation Capacitor CTH
In a typical design, the compensation capacitor CTH should only affect loop gain in frequency range 2 between fP1 and fZ1. Figure 7 shows a smaller CTH value (with higher impedance) increases the A(s) gain in frequency range 2. Figure 8 shows that smaller CTH does not affect the supply loop bandwidth and, therefore, does not change much of the VOUT undershoot and overshoot amplitude during a load transient event. However, a smaller CTH helps to reduce transient settle time, thanks to its higher gain in the lower frequency range 2.
(3) Compensation Capacitor CTHP
CTHP is usually a smaller high frequency, low ESR, and low ESL capacitor to attenuate high frequency noises in the feedback loop. Therefore, the power supply has a clean and low jitter switching waveform. It can be helpful for current-mode supplies that are sensitive to noises on their current comparator inputs. CTHP should be selected to be much less than CTH and contribute its effect only in high frequency (range 4). For the supply loop gain, CTHP can help to achieve the desired >8 dB attenuation at fSW/2. Figure 9 shows how increased CTHP helps to reduce A(s) gain at higher frequencies. Figure 10 shows that increased CTHP reduces the supply loop gain at higher frequencies, so high frequency noises can be attenuated. As long as CTHP is kept as a small value (<<CTH), the loop bandwidth and load transient responses are not affected.
Simple 3-Step Loop Compensation Design Process
With a good and intuitive understanding of the ZITH network, the Type 2 compensation design can be done in a simple, 3-step process for a given target loop bandwidth. It can be done with the LTpowerCAD supply design tool or with bench bode plot measurement equipment.
Step 1—Setting the RTH Value for the Target Loop Bandwidth
Conceptually, the RTH value can be directly calculated for a given target loop bandwidth fBW from Equation 1. At the loop bandwidth (cross-over frequency), the loop gain magnitude is 0 dB, which is 1:
Therefore, the RTH value can be derived in the following steps:
Practically, if it is not convenient to estimate the GCV value with a typical controller IC, ADI’s LTpowerCAD program can be used instead. LTpowerCAD is a complete power supply design tool supporting power stage and loop compensation optimizations.2 On the LTpowerCAD loop design page, users can preset a very large CTH value first, then start with a small RTH and increase it gradually until the target loop bandwidth is achieved. In this case, with current-mode control, the phase margin is preferred to be greater than 60°. If not, simply reduce the RTH value back to reduce the loop bandwidth until the desired 60° phase margin is achieved. See Figure 11.
Without using LTpowerCAD, if bode plot measurement equipment is used in the lab, a designer can still start using a large CTH value then increase RTH from very small to large until the desired loop bandwidth and phase margin is achieved in measurement.
Step 2—Setting the CTHP Value to Attenuate Noises
Next, using the LTpowerCAD tool or bode plot measurement, CTHP should be increased from 0 until the loop gain at fSW/2 is below –8 dB for switching noise attenuation. In addition, the supply gain margin (at phase margin = 0) should be 8 dB to 10 dB as well. See Figure 12.
Step 3—Set the CTH Value for Fast Transient Settle Time
In this step, decrease the very high preset CTH value until the supply’s phase margin starts to drop noticeably, which means fZ1 is approaching the loop bandwidth. Smaller CTH helps to reduce load transient settle time. However, if CTH is too small, eventually the supply phase margin will be affected. Maintain the 45° to 60° phase margin. See Figure 13.
(Optional) Step 4—Additional Phase Boost from Resistor Divider Capacitors
In case adjusting CTH, CTHP, and RTH still cannot achieve the desired loop bandwidth and stability margins, the resistor divider network can be further adjusted by adding the feedforward capacitor CFF and filter capacitor CFLT. The goal is to achieve a phase boost around the targeted loop bandwidth, usually with the help of CFF. This can be done on the LTpowerCAD loop compensation page by opening the Feedback tab. Figure 14 shows good and bad examples of designing the R-divider capacitors. The good example maximizes phase boost at the target loop bandwidth frequency.
Further Simplification Using LTpowerCAD One-Click Loop Design
To further simplify the loop design effort, ADI’s LTpowerCAD program provides a one-click automatic loop compensation design function based on the 3-step method explained in this article. As shown in Figure 15, a user can set the target loop bandwidth, which is usually in the range of 1/10 to 1/5 of the switching frequency, then click on the Use Suggested Compensation checkbox. The LTpowerCAD program will provide a set of RTH, CTH, and CTHP values to achieve the targeted loop bandwidth with a good phase margin. If the target bandwidth or phase margin is not achievable, the user should reduce the target loop bandwidth frequency. This one-click loop design function can be disabled by unchecking the Use Suggested Compensation box to allow users to manually design and fine tune loop and load transient performances.
Conclusion
For a switching power supply utilizing a popular current-mode control architecture, loop compensation design and optimization can be easily accomplished using the straightforward and simple 3-step method outlined in this article. The LTpowerCAD design tool provides real-time results, and its one-click automatic loop design function greatly simplifies the design and optimization process.
References
1 Henry Zhang. “AN-149: Modeling and Loop Compensation Design of Switch-Mode Power Supplies.” Analog Devices, Inc., January 2015.
2 Henry Zhang. “AN-158F: Designing Power Supply Parameters in Five Simple Steps with the LTpowerCAD Design Tool.” Analog Devices, Inc., September 2015.
Zhang, Henry. “Understand Power Supply Loop Stability and Loop Compensation—Part 1: Basic Concepts and Tools.” Analog Devices, Inc., January 2022.
Zhang, Henry. “Understanding Power Supply Loop Stability and Compensation—Part 2: Unusual or Problematic Bode Plots.” Analog Devices, Inc., June 2024.