Predicting the loop stability of the UCC289X active clamp forward converter

Author: Richard Yang, Power FAE, Texas Instruments (TI) China

Abstract
In this paper, we use the peak current mode (PCM) improved small signal model of active clamp forward converter under continuous inductor current condition to predict the actual loop stability of UCC289X application. To verify the validity of the calculation results, we further prove it by actual measurement based on UCC2897 EVM prototype and establish UCC2897A simulation model. The results show that the calculation results based on the improved small signal model can also accurately predict the actual loop stability.

1. Introduction
With the continuous development of power supply reliability assessment, loop stability testing with specific loop analysis instruments has become the only indispensable requirement. However, in the actual development process, engineers often spend too much time on debugging the loop stability function. For example, when using active clamp converters, we always encounter loop stability issues because it is difficult to achieve a good balance between large signal load dynamics and small signal loop stability, especially in peak current control mode. MOS voltage stress has a huge impact on large signal load dynamic optimization. In order to obtain better optimization, engineers usually spend a lot of time repeatedly debugging loop stability.

Since its small signal model is not accurate, there is controversy as to whether this calculation method is applicable to actual measurement, but if an accurate small signal model can be obtained, this problem can be solved.

The goal of this article is to establish a calculation platform to verify the loop stability according to the improved CCM small signal model of the active clamp forward converter used in the UCC289X application. Figure 1 shows an example of actual EVM verification, and this article will also provide you with a lot of comparison data. Finally, it has been proven that the loop stability calculation is very useful for actual design and debugging during development when using the UCC289X application.

 

Figure 1 Schematic diagram based on EVM

2. Calculation of power stage transfer function
Please refer to the BOM of the schematic diagram shown in Figure 1. The following table lists the power stage parameters.

Solve the duty cycle as:
 
The output load is calculated as:
 
The equivalent primary sense resistor is:
 
Solving for mc is:
 
Based on the previous parameters, the Bode diagram and phase characteristic diagram of the transfer function can be drawn using MathCAD or simulation software.
 
in:
 
For the UCC289X application, the coefficient KC is determined by the internal resistor divider; it is set to 0.2, and the final power stage transfer function from control to output transfer function is:
 

Figure 2 shows the calculation results:

 
Figure 2. Bode calculation of control-to-output transfer function

3. Transfer function of the feedback loop
In UCC2897X applications, the voltage compensation circuit is mostly used with the circuit shown in Figure 3.
 

Figure 3 Voltage compensation circuit

OPTO modeling is most important to obtain the feedback loop transfer function. Normally, accurate modeling depends on two parameters. The first parameter is the CTR of OPTO, which depends on its stable value and can be easily solved. Many times, the second parameter is a bit difficult to obtain because of its high-frequency characteristics.
 

Figure 4 Switching time comparison of SFH690BT with related load resistance

However, the most important parameters affecting this high-frequency characteristic are RL and Cin. Cin refers to the internal capacitance; we assume that it is added between the output terminals of the current-controlled current source for transient analysis. Cin is calculated according to the following formula:
 
In this application where Ic is 1mA, we can assume that Tr is about 40u, so Cin is:
 
From the above results, we can choose Cin as 10n.
Then the feedback transfer function is:
 
Therefore, the closed total transfer function is:
 
The loop is closed using the following function:
 
The result of drawing using MathCAD is:
 

Figure 5. Calculation results of the total voltage loop stability of the closed loop.

4. Use simulation to verify loop stability
In order to demonstrate the effectiveness of the above transfer function, we created a UCC2897A simulation model based on the typical circuit created by the EVM application solution. The circuit parameters are basically consistent with the EVM BOM.
 

Figure 6 Loop stability verification simulation circuit

Figures 7 to 9 show the comparison between calculation and simulation.
 

Figure 7 Comparison of calculated and measured total voltage loop curves at 38Vdc input and 3.3V/30A output


Figure 8 Comparison of calculated and measured total voltage loop curves for 48Vdc input and 3.3V/30A output operation


Figure 9 Comparison of calculated and measured total voltage loop curves for 72Vdc input and 3.3V/30A output operation

The following table shows the comparison:
 

It shows that the calculated results can match the simulation results very well.

5. Verify the loop stability using actual measurements
To further verify the calculated loop curve, we compare the calculated results based on the UCC2891EVM with the actual measured results at 48-Vdc input and 75Vdc input:
 

Figure 10 Comparison of calculated and measured total voltage loop curves for 48Vdc input and 3.3V/10A output operation

Figure 11 Comparison of calculated and measured total voltage loop curves for 75Vdc input and 3.3V/10A output operation

The following table provides a comparison:
 
It shows that the calculated results can match the measured results very well.

Please note that the calculated gain margin of the measured results is a bit too large. This is because the prediction of the resonant parasitic parameters is highly complex when the frequency is high.

6. Conclusion
The use of the improved small signal model involving the UCC289X active clamp forward converter to predict the actual loop stability is very useful for actual loop debugging. Engineers can achieve more efficient results when debugging loop stability using this method.

7. References
1. UCC289/1/2/3/4 Current Mode Active Clamp PWM Controllers, Data Sheet (SLUS542)
2. UCC2897A Current Mode Active Clamp PWM Controllers, Data Sheet (SLUS829D)
3. UCC3580/-1/-2/-3/-4 Single-Ended Active Clamp Reset PWM, Data Sheet (SLUS292A)
4. Active Clamp Reset 48V to 1.3-V, 30A Forward Converter UCC2891EVM, by Steve Mappus, in the UCC2891EVM User Guide (SLUU178)
5. Understanding and Designing Active Clamp Current Mode Control Converters Using the UCC2897A (SLUA535)
6. Improved CCM Small Signal Model for PCM Control for UCC284X/UCC289X/LM5026