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.

UCC28780: PMP30631: Modify design using PMP30631 board

Part Number: UCC28780
Other Parts Discussed in Thread: PMP30631, , PMP, TL431

Dir Sir/Madam,

I have successfully simulated the design on the kit .

Now I want to change the design to output 5V ~ 6A. 

Simple iteration of the ratios reported in the PMP30631 kit  result in actual turns of Np=80 : Ns=20 : Na = 14.

Please let me know what is the best wire ratio I should use for the design. 

I tried to simulate the number of turns Np=80 : Ns=5 : Na = 14.

Hower, ripple output is very high. 

I send the calculation on the PMP30631 kit & for the part I calculate for 5V output. 

Please help me with the issue.

I attached the file as below !

  UCC28780 Excel Design Calculator (Rev. C) SLUC664C_PMP30631_5V.xlsxUCC28780 Excel Design Calculator (Rev. C) SLUC664C_PMP30631.xlsx

   UCC28780_TRAN_Si_PMP30631_5V.rar

  • Hello Jack, 

    This post is very similar to your other post that we are currently working on.  
    However, I will address this one separately, despite the similarities.  If they are both converging to the same project, then I'd like to close this thread out and continue with your earlier one. 

    To address your revised-PMP30631 simulation results where the 5V output appears to oscillate around 5.05V at ~3.7kHz, I think this comes from making only partial changes to the power stage and control design from a previous design.  

    Your PMP simulation for 5V, 6A started with the same structure as the standard UCC28780 20V, 2.25A Simplis system model available from TI's website, and you made only a few changes to it based on your new 5V, 6A design target.  
    Specifically, you changed the transformer turns ratios, some values around the TL431 regulator, the clamp capacitors, LM, Co1, Co2, RDM, RTZ, and a few other parts to match values from the PMP30631 schematic. 

    However, there are several other important details that you did not address, which I believe is causing the oscillations in your 5V simulation. 
    Rint and Cint are still the same as in the 20V design, but should changed per the Excel tool recommendations for a 5V design.
    Rcs is also still the same as for the 70W PMP design, not scaled to a 30W target (per the Excel tool).
    Lm is changed, but Llk is still the same 2.6uH as in the standard 45W design.  Llk is an important parameter to know or estimate, since it is key to sizing the clamp cap.
    Cclamp1 and Cclamp2 are set at the nominal values from the PMP schematic: 470nF & 680nF.  However, the Excel tool and real design uses their effective values when DC-bias is applied, which drastically reduces their actual values in operation.  The simulation does not do that, so Cclamp should be what the Excel tool recommends with the bias effect included. 
    The PMP design is a Primary-Resonance design with Lo shorted and Co2 open.  Your simulation still has the Secondary-Resonance structure in place.  

    The PMP design uses IPAW60R360P7S Fets for switching, whose Coss values influence the Cclamp, RTZ, and other component values.  
    The Simplis model uses IPL60R360P6S Fets which have different Coss properties, so PMP component values aren't appropriate. 
    The Excel tool has IPL60R185C7 Fets which have even different properties, so the recommendations of the tool are only appropriate with those MOSFETs. 

    I recommend that you make an initial decision on which MOSFET (can be different than or the same as any of the above) you are actually going to use in your real design, then enter the parameters into the Excel tool and use that same model in the simulation, so that Calc and simulation are consistent with each other. 
    The same goes for all the other parameters that are calculated.  These details are important to keep consistent.   

    All details are important at some point or other.  But details of the gate-drive structure are not so important when trying to get the power stage running and stable. 
    They are important only later when trying to optimize the gate-drive.  So some details can be put off until later, but for the power conversion and control components, the simulation should match the calculator tool results first. 
    And the inputs to the Calc Tool should be based on the actual components that you intend to use, not values left over from some previous design. 

    Since the Excel tool is based on Primary-Resonance and the PMP design is also Primary-Resonance, I recommend that you start a new Calc Tool file for you 5V-6A design with all User inputs based on actual parts and values that you will use.  You can ignore the Secondary-Resonance sheet and Burst Mode for USB-PD sheet since they are not relevant for your design. 
    Once it is fully filled out, set up your simulation to reflect all of the parts and values that you will use.  Delete the parts you won't use, to avoid confusing the sim. 
    Hopefully, you'll have better simulation results. 

    Regards,
    Ulrich

     

  • Hello Ulrich,

    That's right, it same a project with the previous project thread.

    However my client has changed from 10A ==> 5A and and requires a wide range of input, design on only one side of PCB.

    I've found Reference designs:

    PMP30720 — High efficiency, ultra-wide input (20 VDC to 375 VDC) isolated power supply reference design

    and

    PMP30631 — High efficiency, 70-W AC/DC active clamp flyback reference design

    I will base on the last principle is PMP30720 and  I need to change the least components, especially the transformer.

    Turns of Np=80 : Ns=20 : Na = 14. (Partnumber 750318431_Wurth ~ SMD mount package)

    I tried to simulate the number of turns Np=80 : Ns=5 : Na = 14.

    Can you tell me what is the right turn ratio?

    Currently I am simulating the flyback stage, i will reselect the mosfet and send you the simulation.

    ok i will close this thread and just use previous thread.

    Best Regards,,

    Jack

     

  • Hello Jack. 

    Since the 20V design has an 80:20 ratio for Nps, it provides a 4*20V = 80V for reflected voltage on the primary winding. 
    Also, 20V *14/20 = 14V on the AUX winding.  This is good enough for VDD.

    For the 5V design, to keep the same reflected voltage, 80V/5V = 16 for Nps ratio.  For 80T primary, that is 5T secondary. 
    Also 5V *14/5 = 14V on the AUX winding.  This is still good enough for VDD.

    So 80 : 5 : 14 turns can be a good choice for turns ratio.   Other turns ratios may also work okay, but will have some differences in operation. 
    The reflected voltages can be higher or lower, depending on whether it gives you an advantage somewhere (choice of MOSFET ratings, for example). 

    There is no single right answer and all other ratios are wrong.  There is mainly the best tradeoff between semiconductor stresses and their ratings, and how many actual turns can fit on a transformer bobbin and have acceptable losses and minimum size to deliver the power needed. 

    Choosing the final ratio and actual turns is all about making tradeoffs to find a good balance between size, loss, performance, reliability and cost. 
    I will close this thread now.

    Regards,
    Ulrich

  • Hello Ulrich,

    Many thanks for your assistance in our project. 

    Best Regards,

    Jack