• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 3 answers
  • Subscribers 62 subscribers
  • Views 681 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

UCC28782: UCC28782 SCH review and IC UCC28782 BSW pin damaged issue

Hank Cheng
Prodigy 40 points
Part Number: UCC28782


Hello:

We have a 65W USB-C PD ACF with UCC28782ADRTWR ac-dc circuit finally fully operational up to 65W output.

Our intention is to productize it for USBC PD AND a versatile charger.

The base design with UCC28782 shows a WT6636F USB interface chip . 

We face a issue and the board broke @ 265VAC by lowering the load gradually to 1.15A. The oscilloscope signals (Drain LS) and the thermal were normal at the time of the break, around 40°C for the hottest point. There was no false manipulation on our part, the test environment was stable and known.
Visual inspection of the board shows no obvious failure. And GaN looks intact. 

Finally, it is found that the BSW pin of the IC UCC28782  is damaged, and it can work normally again after replacing a new IC UCC28782 .

Two things need your help,

1.Please help review the SCH

2.What are the possible reasons that can cause the IC UCC28782 BSW  pin to damaged ?

(The attachments are circuit  and design calculator and used components)

UCC28782 Excel Design Calculator RevC 的202316.xlsx

  • 0
    TI__Mastermind 40580 points

    Hello Hank, 

    Thank you for your project using the UCC28782 ACF controller. 

    In our experience, damage to the BSW pin is usually due to excess voltage stress.  The Absolute Maximum rating of BSW is 30V (with respect to BGND). 
    An overvoltage at BSW may result from a transient voltage rise from the Auxiliary winding, which may have increased BIN voltage > 30V.
    I'm not sure how that can occur while slowly reducing output current. But it is worth checking with an oscilloscope. 

    Maybe more likely is that there can be some stray inductance in the BSW track between the pin and the L2 inductor.  This stray inductance may be creating spikes of voltage normally peaking below 30V, but a small rise of BIN or VDD above normal may raise the peak of the BSW spikes above 30V, enough to damage that pin.  
    Please probe the BSW pin voltage with respect to BGND pin, using "tip& barrel" probe technique to minimize noise pickup.  Try to probe as close to the two pins as possible, using full bandwidth.  If you see voltage spikes there close to 30V, please investigate the pcb layout to see if the track length can be reduced. 

    Your schematic shows a bidirectional clamp D6 across C15 on the BIN net.  The schematic shows no value or part number.  Please make sure that this part can sufficiently clamp the input to the boost stage to < 30V.  It does not need to be bidirectional, by the way. 

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Prodigy 40 points

    Hi 

    Thank you!

    1.I will confirm the voltage on BSW to see if it is caused by Auxiliary winding or stray inductance!

    2.Could you review the SCH , I don't know if there is a small error in my design ,THX

  • +1 in reply to Hank Cheng
    TI__Mastermind 40580 points

    Hello Hank, 

    I already did review the portions of the schematic that are posted above.  I did not see anything that would cause me concern, except for the unknown value of D6.  For the most part, your schematic resembles that of the UCC28782EVM-002 except for the changes involving the GaN Fets on the primary. 

    I did not detect any errors in your design.  

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Prodigy 40 points

    Hi 

    Thank you!

    1. Yes, you are right! For D6,currently we use the CDSOD323-T24SC! And it's Bidirectional Diode , could you recommend me replacement components for D6? THX

    2.In addition ,I want to add a TVS Diode close to BSW pin to make sure BSW pin doesn't damage again! Could you also recommend me a solution, THX

  • +1 in reply to Hank Cheng
    TI__Mastermind 40580 points

    Hi Hank, 

    CDSOD323-T24SC is the same part that we use on the UCC28782EVM-002 board.  Since your application is USB-PD, I guess that the highest output voltage will be 20V, so a 24-V TVS will be necessary to provide margin to the 20-V variations.  At the BIN pin, it is probably okay. 

    For a TVS on the BSW pin, I think the uni-directional version CDSOD323-T24S (or the bi-dir part CDSOD323-T24SC to keep the parts the same) could be an appropriate clamp to protect BSW against over-voltage.  It should be connected close to BSW and BGND to minimize its own inductances to the pins it is intended to protect. 

    Regards,
    Ulrich 

  • 0 in reply to Ulrich Goerke
    Prodigy 40 points

    Hi 

    Thank you for the feedback!

    More discussion,
    Maybe the damage of the BSW pin is caused by the transformer, is it the transformer using the P22x13 core? Because I have not seen this type of breakage with the Rencousa EQ25 and RM8 transformers. I think it's always the bad soldering that causes the breakage.

  • 0 in reply to Hank Cheng
    TI__Mastermind 40580 points

    Hello Hank, 

    A transformer by itself may not be the cause of the BSW failure, but if a bad solder joint opens or a wire breaks with current in it, the inductive "kick" may generate an overvoltage on the AUX winding which may overstress the BSW pin.  

    I have not seen a this kind of core shape (commonly called a "pot core") used for a power transformer application.  It is an old technology, and pot cores have earlier been used for tuning inductors.  This core has a hollow center post for the tuning slug.  Most power cores have solid centers posts. 

    It is more difficult to bring many windings out of this kind of shape, because of the narrow opening.  Maybe a wire was stressed and broke free of its termination as you suggest. In my opinion, bad connections account for >90% of all electrical problems.  But this is just my opinion, and I don't know if there was any wire breakage that could have caused the BSW failure. 

    Unless something is repeatable, all possibilities are speculation. 

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Prodigy 40 points

    Hi 

    Thank you !

    It's interesting,

    I have found that if the soldering is good, I never have breakage problem with UCC28782 BSW pin.

    Does this happen to other case as well? THX

  • 0 in reply to Hank Cheng
    TI__Mastermind 40580 points

    Hello Hank, 

    I have not experienced a soldering problem involving the BSW pin specifically and I am not aware of any other customers who have, but that does not mean that it has not happened before elsewhere.  We just did not hear about it.

    Over the years, I have experienced connections breaking open on my own prototype boards during debug of various systems.
    Sometimes the result was benign and simple resoldering solved the problem. 
    Sometimes the result was catastrophic and I had to replace many components or even replace the entire board. 

    In any case, good solder joints are always important, whether in production or just in prototype development. 

    Regards,
    Ulrich