This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

UCC28070: Improve the input harmonics by change the CSA/CSB pin sense resistor

Part Number: UCC28070

Hi Team,

As title, Our PSU vendor using modify the CSA/CSB sensing resistors values to improve the input harmonics. To let the power supply can pass the EN61000-3-2 class A at 10%~100% load.

I'd like to learn from TI's point of view, if change the sensing resistors that means change the voltage level of CSA/CSB. Is this design change impact any PFC function? For example OCP percentage? Function? or other input characteristic? 

I mean what testing should be verified after the design change? Please TI recommend. 

Thank you.

Muhsiu

  • Hello Mushui, 

    Thank you for your interest in the UCC28070 Interleaved PFC controller. 

    Am I correct to assume that "...modify the CSA/CSB sensing resistors values..." means increasing their resistance values at light load (or maybe the opposite (decreasing value at heavy load)?  In either case, changing resistance does change the voltage values at CSA/CSB.   
    A higher resistance at light load will generate a larger signal at CSA/B which can help improve the internal synthesized downslope and overall average current waveshape, so iTHD may be slightly improved by this method.  However, by making the CSx signals appear larger, the current amplifiers CAOA and CAOB will reduce their output voltage and duty-cycle will drop, so VAO will increase to compensate.  This leaves less headroom to the 5-V max limit on VAO to respond with higher current in case of a load step.   
    The CSx resistors must change back to high-load values to support the load step. 

    Most of the light-load THD comes from operating in discontinuous mode.  Distortion is low when operating in CCM, but DCM introduces some reverse currents in the current-sense transformers (CT) which result in varying distortion along the input sine curve of the current.  This effect can be reduced by choosing Lm of the CT to be as high as possible, but that usually increases physical size and adds cost. 
    An alternative idea for reducing iTHD is to modify the Rsynth resistance at light load.  This can be done empirically (by trial and error) to see if increasing or decreasing Rsynth has a beneficial impact on the THD.    

     You can use these methods independently or together.  But be sure to restore the normal values for high-load operation, and incorporate sufficient hysteresis around the high/low  load-change point so that there is no chatter back and forth at the thresholds. 

    Regards,
    Ulrich

  • Hi Ulrich,

    Thank you very much for your detailed feedback. 

    Please help to check if my understanding is correct or not. 

    It seems to me, the PSU vendor are trying to increasing the CSA/CSB resistance to let power stage into CCM at light load to improve the iTHD.

    But "CSx signals appear larger, the current amplifiers CAOA and CAOB will reduce their output voltage and duty-cycle will drop, so VAO will increase to compensate." , This is because controller have to keep the desired output voltage, so VAO increase the voltage level trying to increasing the duty cycle?

    And the current will increasing fast during the load step, so CSx signal are larger => turns out the VAO need to compensate more.  But there was internal 5V Zener diode to limit the VAO amplitude. That might be a trade off between load step response and iTHD performance. I am not sure they can restore the normal  values for high-load operation. This request the design have to switch the resistors between light load and heavy load.

    Thanks to share alternative method to resolve iTHD issue, I will pass this idea to them.  

    Thank you.

    Muhsiu

  • Hello Muhsui,

    Yes, your understanding is correct.   

    Your customer needs to understand this:  Changing the CSx resistance does not change the input current for any given load and line condition. 
     All that does is change the signal level that the controller sees, and the controller acts on that signal to adjust PWM to keep Vout regulated while shaping the average input current to follow the input voltage wave-shape.  

    Changing signal level may have a small second-order effect on the overall current shape and iTHD, but the main cycle-by-cycle current peaks are set by the line and the load, regardless of whether the CSx signals are higher or lower.  
    Therefore, increasing Rsense to increase CSx cannot and will not force light load to operate in CCM.  The load level for beginning of DCM can be pushed down to lighter load only by increasing the boost inductance.    

    Regards,
    Ulrich

  • Thank you very much, Ulrich.