• State Resolved
  • +1 person also asked this people also asked this
  • Locked Locked
  • Replies 7 replies
  • Subscribers 64 subscribers
  • Views 191 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

UCC28056: Output not regulating issue

Samo Smodiš
Prodigy 200 points
Part Number: UCC28056

Tool/software:

Hello,

I am designing a PFC using the UCC28056A IC. I have an issue where the PFC does not regulate to the set voltage. For example, when I set the output voltage to 300V, the output reaches this voltage and works normally. But when I start increasing the output voltage setpoint towards 500V at some point it just stops increasing the voltage and stays at 425V. I have noticed that when I try increasing the output voltage setpoint above this voltage, the VOSNS voltage starts dropping from 2.5V to 2.0V.

Here is a capture of the VOSNS, ZCD and fet drain voltages:

The blue signal is VOSNS and it is steady around 2.0V

I have also measured the ZCD signal closely to see if there is an overcurrent protection OCP1 happening, but  the ZCD signal during on-interval does not exceed 500mV to trip OCP1. It only reaches ~50mV. Here is a capture of ZCD and drain voltage:

After this I also measured the COMP voltage. Unfortunatelly I don't know how to interpret this signal and I ask you for help and advice on what should the signal look like.

To make things worse, when i increase the input AC voltage, the PFC starts pulsating. I deduced that OCP2 is triggering since the interval between separate pulses is 1s, as is the delay when OCP2 is triggered. Here is a capture of this pulsation.

I dont understand why OCP2 is triggering since the ZCD/CS signal during on-time reaches only 60mV which is far from 750mV threshold to trigger OCP2.

I am attaching the Altium project so you can see the schematics and PCB layout. I am also attaching the UCC28056 excel calculator I used to design the circuit.

I thank you for your help.

Samo

240613 UV Bulb Driver Altium - TI e2e ticket.zip

260613 UCC28056X_Design Calculator v02 120Vac TI e2e ticket.xlsm

  • 0
    TI__Mastermind 31130 points

    Hello Samo, 

    The Altium .zip file is "empty", it has no contents.  But that's okay.  I don't need to see a pcb file at this time. 
    Please generate a .PDF file of the schematic from Altium and post that.  It will be easier for me to work with. 

    Meanwhile, the COMP voltage signal looks very strange.  Normally, this signal should be almost a flat DC level (with very small 100Hz AC ripple), with DC ranging between 0V and 5V.  The level corresponds roughly proportionately to the output power. 

    Your signal is varying significantly over a half-line cycle.  This indicates that your compensation network is not correct somehow, with much too fast response time.

    Your Excel design file came in partially damaged and was repaired by deleting some Macro items, so many calculations do not work. 
    However, the user-selected items (your inputs in the Yellow cells) are intact and show that your values for Cp, Cz, and Rz are an order of magnitude different from the usual values that I see at this pin.   I presume that the values in the tool are the values used in your actual board, and that presumption is supported by the COMP signal behavior.

    I'm not sure what the tool recommended for these components (because the Macros that calculate the recommendations are deleted), but a non-optimal very-rough guess is to increase both Cp and Cz by 10X, and decrease Rz by 1/10.    

    At this time, I don't know what is affecting the ZCD signal.  I'll investigate more once I see your schematic. 
    Note:  the schematic diagram represents the design that you intend to achieve. 
    The pcb and parts on it should match the schematic completely.  Please verify that all components on the board are what is shown on the schematic, in correct polarities, and no shorts or opens.  Conversely, if you made any modifications to the board, make sure that the schematic has been updated to show those modifications.  

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Prodigy 200 points

    Ulrich thank you for your quick response, I really appreciate it.

    I have re-attached the Altium project and also a PDF of the schematic. I also re-attached the Excel calculator, hopefully with macros working.

    The values I have for the compensator component are:

    C_z = 200n

    R_z = 275k

    C_p = 20n

    Should I try to increase C_z and C_p by 10x and reduce R_z by 1/10 ?

    5481.240613 UV Bulb Driver Altium - TI e2e ticket.zip

    240614 SCH UV Bulb Driver v1_1.pdf

    5633.260613 UCC28056X_Design Calculator v02 120Vac TI e2e ticket.xlsm

  • 0 in reply to Samo Smodiš
    TI__Mastermind 31130 points

    Hello Samo, 

    Yes, while I spend some time looking through the information that you provided, please try increasing C_z and C_p by 10x and reduce R_z by 1/10 and see if performance improves significantly.

    Those may not be the final optimal values, but I think they will be a lot closer to normal than what you have now, and COMP should be more of a flat signal across the line frequency at all load levels.  

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Prodigy 200 points

    Hi Ulrich,

    Thank you for your suggestion on modifying the compensator components. I 10x the capacitors and 1/10 the resistor and the performance improved in the sense that the PFC now works steadily and not constantly in burst mode. I then increased the capacitors event more (20x) and reduced the resistor (1/20) and i think it is even better now.

    The current values are:

    C_p = 470n

    C_z = 4.7u

    R_z = 10k

    I also found out why the PFC did not regulate the output voltage to 500Vdc. It appears that the PCB fabricator either used a different model of the UCC28056A, which is not inteded for output voltages higher than 400Vdc, or they used a fake chinese alternative. Either way, once I changed the IC with one I bought directly from TI the output increases to 500Vdc.

    I still have one more issue, and this is burst mode. My PFC goes into burst mode as soon as the output power drops below 10~15W which is not good since this causes high output voltage oscillations. Ideally I would like burst mode to happen when the ouput power is less than 3W. The datasheet also states that burst mode should start when the load drops below 15% Load.

    I dont understand how the IC "knows" when it is at 15% of rated load and what can I do to change the moment when it transitions. Our PFC is rated for 30W so starting burst mode at 15W is really unwanted. Can you advise me on this?

  • 0 in reply to Samo Smodiš
    Prodigy 200 points

    Here is capture of drain and COMP voltages when the PFC operates in burst mode.

    It is apparent that v_COMP drop to 0.750V which is the threshold to start burst mode. How do I keep COMP voltage higher to prevent this?

  • +1 in reply to Samo Smodiš
    TI__Mastermind 31130 points

    Hello Samo, 

    To answer your question in your earlier post:  The UCC28056 "knows" when it is at 15% of rated load based on the COMP voltage.  COMP voltage is defined to be at the 5-V clamp at 100% load.  COMP voltage is (mostly) independent of input line, so 15% of load = Vcomp =0.15 X 5V = 0.75V.
    In the UCC28056A, when Vcomp falls to 0.75V, it stops switching and won't resume switching until Vcomp rises to 0.875V.  That constitutes Burst Mode and your observations confirm this.   

    Since the threshold is not programmable, the only way to shift Burst Mode to 3W or less is to define 3W as 15% of 3/0.15 = 20W.  So your 100%-rated load must be only 20W.  But that does not match your stated design target of 30W.  
    Meanwhile, if bursting starts at 15W, this is 15% of 100W, so your PFC circuit is behaving as if it is designed for 100W.  
    Your Excel tool indicates at maximum design load target of 50W (67% higher than your stated target of 30W).  

    The UCC28056 controller is designed for applications with AC input range of 85Vac to 265Vac.  With only 3 pins available to manipulate VOSNS, ZCD/CS, COMP), there is very little capability for deviation from the "usual" application.  Several assumptions have been built-in to the controller design. 
    One of them is that Vout = ~400V (not 500V like yours); another is that Kzc = 401 (not 501 like yours); another is Vin_min goes to 85Vrms, (not 110Vrms like yours); and another is that actual Lboost is very close to the recommended value (not ~23% lower like yours (680uH instead of 880uH)). 
     
    I think these deviations from the assumed "usual" PFC design combine to make the apparent max power level about 100W, which leads to Burst Mode behavior at ~15W. 

    I am not sure what to do about all of this, but I do recommend to modify your Excel inputs for 30W max load (if that is indeed your maximum target) and use 85Vac as the minimum input (even if you don't intend to run it that low).  
    Using 85Vac input at 30W out will adjust the current sense resistor value higher which should move the Burst Mode threshold down to ~4.5W. 
    I'm not sure what to do about Kzc.  Obviously, you must configure VOSNS feedback for 500V output. 
    I'll have to think about how these factors affect the UCC28056 operation.

    Regards,
    Ulrich  

  • 0 in reply to Ulrich Goerke
    Prodigy 200 points

    Hi Ulrich,

    I found out that the issue was in the boost inductance Lboost. I had it too small and when I increased it to the value suggested by the Excel calculator, the PFC behaviour improved significantly. Now I transition into burst mode at power levels around 4W, which is acceptable.

    Thank you for your help, I really appreciate it.

    Regards,

    Samo