UCC25660: ucc256601 VCR Synthesizer operation and Loop stability

Part Number: UCC25660

Tool/software:

Hello Everyone,

 

We are currently working on a design that uses the UCC256601 and as we push towards layout a few questions have been asked about loop stability.

 

1) Can the Loop Stability tested using Phae Gain analyzer (Bode)

 

2) Does the VCR Synthesizer maintain loop stability.

I looked around the TI website and even an internet search for “VCR Synthesizer” and totally struck out.

 

3) Is the loop compensation handled on the input of the opto-coupler in the feedback path.

 The UCC256601 datasheet doesn’t cover this at all.

However, the TI Eval Board UCC25660EVM-064 has a complex compensator on the input of the opto.

 

4) What does “VCR” stand for?

  • Hi Gerald,

    Please see my responses below:

    1) Can the Loop Stability tested using Phae Gain analyzer (Bode)

    Yes, you can use bode 100 to measure the loop stability.

    The HVP064 EVM user guide does show load transient and bode 100 loop response in the user guide.

    2) Does the VCR Synthesizer maintain loop stability.

    I looked around the TI website and even an internet search for “VCR Synthesizer” and totally struck out.

    In charge mode control of the LLC, the resonant cap voltage is sensed directly using a capacitor divider. This voltage is used to compare against the feedback thresholds in order to determine the switching frequency of the converter for a given input and load conditions.

    However, In the UCC25660, We have removed the capacitor divider and reconstructing this resonant cap voltage information inside IC by integrating the resonant current. This is the primary role of the VCR Synthesizer. And this scaled version of resonant capacitor voltage (VCR) information is used to compare against the Feedback thresholds in order to determine the switching frequency of the converter.

    This is shown in following screen shots:

    Since IPPC control is similar to charge mode control, the control to output voltage transfer function is still a single pole response which is easier to stabilize using Type 2 compensator. 

    3) Is the loop compensation handled on the input of the opto-coupler in the feedback path.

     The UCC256601 datasheet doesn’t cover this at all.

    However, the TI Eval Board UCC25660EVM-064 has a complex compensator on the input of the opto.

    Yes, The loop compensation handled on the input of the opto coupler. Here we have used Type 3 compensator to improve the cross over frequency. This kind of compensation technique is very common in isolated topologies.

    Regards

    Manikanta P

  • Thank you Manikanta for the information.

    Does TI have a tool for selecting a starting point for the compensator values?

  • Hi Gerald,

    Usually this is what I follow as a starting point for the compensator design:

    I usually start with type 2 compensator

    Rup and Rlow are set as per the output voltage requirement and shunt regulator reference. 

    Cf and Rv together sets the high frequency pole that would attenuate the high frequency noise.

    low frequency zero (fL) should be placed below 100Hz to eliminate the system pole created by output cap and output load.

    So, for fL=88Hz, and by substituting Rv and Rup obtained previous steps, you can find the Cv value from above equation.

    Rbias can be obtained using the following equation:

    Vopto is forward voltage of optocoupler and Ibias is minimum bias current for the shunt regulator.

    For Rled: You need to follow below equation:

    Regards

    Manikanta P

  • Hello Manikanta,

    I know this maybe an log shot, but is there a transfer function for the plant, as in Vout(s)/(input to the UCC25660_FB(s))?

  • Hi Gerald,

    I will reach out to systems engineer regarding this.  Will get back to you next week if he has any transfer function derivation with him.

    Regards

    Manikanta P