• State TI Thinks Resolved
  • Locked Locked
  • Replies 12 replies
  • Answers 5 answers
  • Subscribers 63 subscribers
  • Views 239 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.

TPS43061: 12V passively passes to the rail, and IC gets hot.

Simon Hilton
Prodigy 130 points
Part Number: TPS43061
Other Parts Discussed in Thread: TPS43060,

Tool/software:

Hi Hope you can help find what I've done wrong... 

We have two issues though these might be directly connected. 

Firstly when the chip is not enabled, its passively passing 12V from the main rail to the target boosted rail. When I would have expected this to simply be off. 

So it appears for some reason when not enabled the output HIGH drive mosfet is getting voltage to its gate, allowing the passive voltage through. 

Not only this, while the chip is not enabled, it quickly reaches 45C. Suggesting some kind of short going on. 

Outside of this, when the chip is enabled, it appears to work fine. Boosting voltage to a stable 20V and the chip cools down. 

Here is our schematic for this circuit...

It was based from a generated drawing from the power designer(last summer), but when checking power designer again now some of the values have changed a bit... 

Now I'm going to assume I've done something wrong here, I just hope some one can highlight what? 

Or is this expected behaviour? Do I need to add a high side switch to block power when not enabled? 

  • 0
    Prodigy 130 points

    To add I just noticed the body diode, which of course means that would passively pass through voltage regardless of mosfet state.

    Been reading https://www.ti.com/lit/an/slvafo4/slvafo4.pdf which has suggestions regarding the passive power passing through...

    Doesnt explain the IC getting hot though. That Id like looking into still. 

  • 0 in reply to Simon Hilton
    TI__Mastermind 19595 points

    Hello Simon, 

    You are right, A regular boost converter cannot disconnect the output, The input voltage will always go across the body diode of the high side FET.

    The schematic looks good.
    But I would recommend adding a ceramic 10uF capacitor also on the input side. 
    From that schematic, I cannot see any obvious reason why the controller is running hot.

    For test purposes (to see if any current is drawn from the gate drivers), you might try disconnecting the FETs by removing the gate resistors.
    But please do not enable it in that case.

    With the gate resistors connected, please use an oscilloscope and have a look at the voltage on the VCC pin and the Boot pin.

    Is definitely nothing connected to the PGOOD pin? 
    Please make sure that there is no short between any of the pins (also not to the thermal pad).


    Best regards
    Harry

  • 0 in reply to Harry Hofner
    Prodigy 130 points

    PG is floating, if theres any shorts that would have been done by assembly/PCB fab. Ill probe and check.

    So apart from lacking an input mosfet to cut off the circuit and maybe a extra cap. Nothing really is wrong in the design? (thats some what a relief)

  • 0 in reply to Simon Hilton
    Prodigy 130 points

    just checked for shorts PG not shorting nor any other pins when powered off.

    Even stranger I've reprogrammed our micro controller to not turn on the power rails so I could run these tests, and the IC is still behaving the same... It heats up when I power the board, then cools off after a couple of seconds. It turns out its nothing to do with the enable signal at all. 

    Im still going to assume this heat up and cool down on power on isnt normal behaviour?

  • 0 in reply to Simon Hilton
    Prodigy 130 points

    Just probed each pin during this hot stage on power up. And nothing really stood out...

    1:0V

    2=0V

    3=0V

    4=600mV

    5=12V

    6=12v

    7=12v

    8=0v

    9=0v

    10=0v

    11=11.5v

    12=11.5v

    13=11.5v

    14=0v

    15=0v

    16=0v

    The only thing, I guess the HIGHDrive mosfet gate being held HIGH, but that doesnt change when the chip cools down. 

  • 0 in reply to Simon Hilton
    TI__Mastermind 19595 points

    Hello Simon, 

    It is not normal behavior when the controller runs hot.
    How hot goes it get? Coes it feel warm or hot? Do you burn your finger?

    If you can exclude the PG pin and also the LDRV and HDRV pins (by removing the gate resistors), 
    the following ideas are left:

    - There is some kind of "short" on the VCC pin. Either you are connecting something to VCC that is not shown in the schematic, or the capacitor has a high leakage or is broken. Maybe replace the VCC cap with a different one (from a different vendor, if possible)

    - There is some kind of "short" on the BOOT pin. Maybe the capacitor has a high leakage or is broken. Try to remove the BOOT cap for test purposes

    - Some negative voltage is applied on any pin (e.g. the EN pin)

    - The controller IC is broken

    By the way:
    The EN pin has three levels: 
    Between 1.2V and 40V: active
    Between 0.4V and 1.2 V: low current standby mode
    Below 0.4 V: shutdown mode. 


    Best regards
    Harry

  • 0 in reply to Harry Hofner
    Prodigy 130 points

    I was just putting in another post when i got thie reply...

    I hotwired the enable pin to make sure it enables as quick as possible... I measured it reach about 50C (thermal camera) on power on but then instantly cool down...

    Is this being caused by some kind of internal IC inrush current as part of its normal boot/power up? (as you now say not normal) 

    Again please note it cools down shortly after board power on regardless of enable being applied or not. It does not stay hot after this short couple of seconds. 

  • 0 in reply to Simon Hilton
    Prodigy 130 points

    Because I dont have any spare in my bin I just removed the caps in question to test, and even with them removed the chip heated up in the same way. I dont think they were shorting. 

  • 0 in reply to Simon Hilton
    TI__Mastermind 19595 points

    Hello Simon, 
    I just testes a TPS43060 (a derivative, based on the same silicon) and it does not get warm at all.
    The voltages you measured look good, but they do not tell us if too much current is drawn somewhere.

    With most of the connections to externa components removed, I would more and more believe that the controller IC itself is to blame.

    Best regards
    Harry

  • 0 in reply to Harry Hofner
    Prodigy 130 points

    You mean the TPS43061 I have on this board might be a faulty part? I dont have another prototype board with me here... But I can try and get some one in the office to have a look at another one we have. 

  • 0 in reply to Simon Hilton
    TI__Mastermind 19595 points

    Yes, this might be an option. 
    I am off for the weekend now until Monday afternoon.

    Best regards

    Harry 

  • 0 in reply to Harry Hofner
    Prodigy 130 points

    Okays thanks for the help thus far, have a good weekend. Ill likely get hold of one of the other boards to test monday or Tuesday as well. 

    Kind regards