• State TI Thinks Resolved
  • Locked Locked
  • Replies 14 replies
  • Answers 6 answers
  • Subscribers 64 subscribers
  • Views 351 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.

Original question:

TPS92520-Q1: The difference between CHxEN=0 and UDIM=0

TPS92520-Q1: UDIM UVLO and ch_en1

Dustin Yim
Intellectual 2840 points
Part Number: TPS92520-Q1

Tool/software:

Hi Team,

Could you please help these question?

1.  Please see the picture below.

- Where is the udim1 signal go to channel 1 block?

- Is the UDIM UVLO related with "ch1_en" or it just goes to udim1 as zero PWM signal?

2. If it is UDIM UVLO status, Can the tps92520 detect other FAULT ? or cannot? 

  • 0
    TI__Mastermind 20885 points

    Hello Dustin, 

    You can treat the UVLOx setup on UDIMx as an AND gate with the CHxEN to turn on the CHx output.  If voltage on UDIMx is not above 1.12V then even if the CHxEN bit is set, the output current will not happen.

  • 0
    TI__Intellectual 2840 points

    Hi Francis,

    Thanks for your answer.

    But my two questions are not solved yet.

    Customer asked these question to see their FAULT flag when UDIM L -> UDIM H.

    And as far as I know,  the “PWM” on the  valley current  control is dcdc sw frequency, not dimming frequency. Could you please check it?

  • 0 in reply to Dustin Yim
    TI__Intellectual 2840 points

    Hi Team,

    Could you please help my questions?

    1. In the block diagram udim1 = pwm1 = PWM

    Is it right? 

    Then how about for internal PWM signal?

    Is it equal to “PWM” in the  valley current control loop?

    2.  What FAULT it can detect and set the flag when it is UDIM low.?

  • 0 in reply to Dustin Yim
    TI__Intellectual 2840 points

    Hi Team,

    1. I guess the pwm signal path is like below.

    Customer asked the pwm net in internal block.

    Could you please give me advice for this? 

    I am asking why they need this information now.

    2. What FAULT it can detect and set the flag when it is UDIM low.?

    I guess, UDIM status don't affect fault diagnostic logic.
  • 0 in reply to Dustin Yim
    TI__Mastermind 20885 points

    Hello Dustin,

    1.  Yes that is correct.  The "PWM" for opening the Gm amp connection happens when using either the PWM registers or UDIM signal.  

    2. The faults will be flagged no matter if it is UDIM high or low, but note some of the faults could never occur unless the device is switching.  This is all explained in the datasheet in section 7.3.13 Faults and Diagnostics.  

    -fhoude

  • 0 in reply to fhoude
    TI__Intellectual 2840 points

    Hi Team,

    1. The inputs of the ‘LED fault Detection’ block are VCSN and COMP, and when a fault occurs.

    And the COMP pin is confirmed to be discharged when any Fault happen.

     - Customer wonder if every FAULT make this discharge. Is it right?

     - Can it detect all of fault when COMP pin is discharged?

    2. Could you please check o, x table below?

    Table 7-1        
      Category Make COMP discharge? Can be dected when COMP discharged? Can be detect when UDIM = L?
    Thermal Warning FAULT  o?    
    Thermal Protectoin FAULT  o?     
    SPI Error FAULT      
    V5D UVLO FAULT      
    V5A UVLO FAULT      
    VINx UVLO FAULT      
    BSTx UVLO FAULT      
    Short CHx output FAULT      
    HSS Current Limit FAULT      
    Low Side Switch Current Limit FAULT      
    Minimum Off-Time FAULT      
             
    Table 7-2        
      FAULT OR DIAG      
    TW DIAG  ???    
    CHxTP FAULT      
    VINx(UVLO) FAULT      
    CHxBSTUV FAULT      
    CHxCOMPOV DIAG      
    CHxSHORT DIAG      
    CHxHSILM FAULT      
    CHxLSILIM FAULT      
    CHxTOFFMIN DIAG      
    V5AUV DIAG      
    PC FAULT      
    SPE DIAG      
    LHSW FAULT      

  • 0 in reply to Dustin Yim
    TI__Mastermind 20885 points

    Hello Dustin,

    Please read the datasheet.  It is covered there the faults and diagnostics section 7.3.13.  

    Only a few faults cause the COMP to be discharged and those faults are stated in the datasheet as it states discharges the COMP.  

    No, we can not directly detect of the COMP voltage. We only know when it gets discharged due to the outlined fault conditions.

    No, we do not measure the UDIM voltage, it is simply a logic gate where the threshold is stated in the specifications. 

    If we did these types of things then we would have it in the datasheet.  If it isn't listed then you can assume it doesn't.  

    -fhoude 

  • 0 in reply to fhoude
    TI__Intellectual 2840 points
    1. But for this two contents,

    "Can be dected when COMP discharged?"

    "Can be detected when UDIM = L?"

    => I wonder if each fault can be detected although comp discharged.

    Could you please check my understanding.

    Fault

    Category

    Can this fault be detected although COMP is discharged?

    Can this fault be detected although UDIM = L?

    Thermal Warning

    FAULT

    O

    O

    Thermal Protection

    FAULT

    O

    O

    SPI Error

    FAULT

    O

    O

    V5D UVLO

    FAULT

    O

    O

    V5A UVLO

    FAULT

    O

    O

    VINx UVLO

    FAULT

    O

    O

    BSTx UVLO

    FAULT

    N/A

    N/A

    COMPx Overvoltage

    FAULT

    N/A

    N/A

    Short CHx Output

    FAULT

    N/A

    N/A

    High-Side Switch Current Limit

    FAULT

    N/A

    N/A

    Low-Side Switch Current Limit

    FAULT

    N/A

    N/A

    Minimum Off-Time

    FAULT

    N/A

    N/A

    The reason of N/A,

    UDIM status and COMP cap chaging status don't affect fault diagnostic itself.

    But they can chage the output status, so make it very difficult those fault situation.

    However, when considering every abnormal situaltion like ESD, short to any Voltage, we cannot say them with "x".

    1. There is a "UVLO" in the UDIM pin input block. Does it goes to ch1_en? or Just make the "udim1" as zero?

    I am asking this, since this internal logic flow question is related with the above table.

    The inputs of the ‘LED fault Detection’ block are VCSN and COMP.

    And the COMP pin is confirmed to be discharged when some Fault happen.

     - Can it detect all of fault although COMP pin is discharged? This is related with above O,X table.

     - I wonder these two input's purpose. They are just input to be sensed? Or They are affecting fault detection ability ifself?

    +

    1. Customer really want to know o,x for those table.

    Can we say x for N/A if there is no any abnormal condtion?  

  • 0 in reply to Dustin Yim
    TI__Mastermind 20885 points

    Hello Dustin,

    Please clarify if X = Yes or O=Yes????  I am going to answer the table with words not XO.  

    Fault

    Category

    Can this fault be detected although COMP is discharged?

    Can this fault be detected although UDIM = L?

    Thermal Warning

    FAULT

    YES

    YES

    Thermal Protection

    FAULT

    YES

    YES

    SPI Error

    FAULT

    YES

    YES

    V5D UVLO

    FAULT

    YES

    YES

    V5A UVLO

    FAULT

    YES

    YES

    VINx UVLO

    FAULT

    This fault is not detected but the condition dictate if the part will be ON or OFF, but it is independent of COMP.  We don't measure UDIM.  

    This depends on what you mean by VINx UVLO.  We use a resistor divider from Vin to GND where UDIM is connected to the resistor divider and the ratio of the resistors sets when the device can turn on.  The VIN UVLO is the UDIM signal.  If it is low then you are below UVLO setpoint.  

    BSTx UVLO

    FAULT

    YES

    YES

    COMPx Overvoltage

    FAULT

    If the COMP is held low then you can't have COMPx OV?  

    IF the output is off then how can the COMP rise up?  

    Short CHx Output

    FAULT

    YES

    YES

    High-Side Switch Current Limit

    FAULT

    NO (the switcher would never be told to turn on if COMP is at 0V)

    NO (the switcher would never be told to turn on if UDIM is at 0V)

    Low-Side Switch Current Limit

    FAULT

    YES (It is not directly tied to status of COMP voltage but during the start sequence the LS FET is turned on is greater than the threshold it could trigger.  That would likely mean some other fault condition occurred.  

    NO (the switcher would never be told to turn on if UDIMx is at 0V)

    Minimum Off-Time

    FAULT

    If COMP is held low then there will be no switching so you won't have Min Off-Time

    N/ANO (the switcher would never be told to turn on if UDIMx is at 0V)

    1. There is a "UVLO" in the UDIM pin input block. Does it goes to ch1_en? or Just make the "udim1" as zero?

    I am asking this, since this internal logic flow question is related with the above table. Yes, there is a threshold for UDIM which you can consider UVLO or just a threshold with hysteresis (rising and falling).  Like I said before, this signal is ANDed with the CHxEN bit status.  The channel will not be turned on unless both signals are "high".    

    The inputs of the ‘LED fault Detection’ block are VCSN and COMP.

    And the COMP pin is confirmed to be discharged when some Fault happen.

     - Can it detect all of fault although COMP pin is discharged? This is related with above O,X table. NO, That block only detects COMP OV and SHORT fault.  Technically it does not detect all OPEN circuit faults as outlined in the datasheet where we have 3 cases you have to process to determine open circuit condition, section 7.3.15.  That table i put in outlines what could potentially be detected.    

     - I wonder these two input's purpose. They are just input to be sensed? Or They are affecting fault detection ability ifself? It detects SHORT and COMPOV Faults, but not all open circuit conditions at outlined in the datasheet.    

    +

    1. Customer really want to know o,x for those table.

    Can we say x for N/A if there is no any abnormal condtion?  I don't know what X or O means so I can't say.  Please be specific about this sort of stuff so I don't have to guess.

    This table is guidance but I am trying to understand what the customer is trying to do with this information? 

    Almost every point I stated in this post is in the datasheet.  Please read the datasheet.  

    -fhoude

  • 0 in reply to fhoude
    TI__Intellectual 2840 points

    In Korea, o is Yes, x is no. So in the table your comments are similar with my assumption. 

    Thanks for your clear explanation.

  • 0 in reply to Dustin Yim
    TI__Intellectual 2840 points

    Additional question,

    If the channel off situation persists from UDIM = L or other reasons, I think the VCSN will decrease and a short fault will be detected.
    On this condition, If the channel is turned on for some reason or if UDIM = H is switched, will the TPS92520 work normally?

  • 0 in reply to Dustin Yim
    TI__Mastermind 20885 points

    Yes.  Even if there is a true short on the output this is a constant current regulator and it will regulate the current into the short.  

  • 0 in reply to fhoude
    TI__Mastermind 20885 points

    Dustin, you still haven't answered any of my questions.  

  • 0 in reply to fhoude
    TI__Intellectual 2840 points

    Hi Francis,

    Thanks for your clear explaination.

    From your questions,

    1. In Korea, o is Yes, x is no.

    2. "This table is guidance but I am trying to understand what the customer is trying to do with this information?"

    => They want to understand what happen on FAULT when UDIM changed. So, I provided the table.

    The reason I posted E2E queston or asking, I am not sure someting with 100% from Datasheet.

    For example, COMP discharge : I need to sure on other fault which don't describe in Datasheet.