• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 636 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.

TPS92691: UVLO / power down issue

Phuzzy
Prodigy 40 points
Part Number: TPS92691

I am currently working on a design using the TPS92691PWP part.  The design is a fairly high power (cica 100W). (12/24v input).   I am currently having an issue on power down with a very short current spike of around 3A on the LED chain (normally driven at ~2A)

What appears to be happening is that on power off, the input voltage drops rapidly, and hance input current (supplied bu the input caps) rises as the boost converter keeps trying to regulate the output.  Due  to this the inductor current limit is hit (~18Amps when using a 0.03 Ohm current sense resistor). This triggers a fault that turns off the output PWM FET.  The Vout voltage seems to rise to 10V or so above the LED Vf  (even though the switcher is off), and after ~35uS the fault is cleared so the output FET is re-enabled, dumping all the charge on the output caps through the LED chain - causing the spike.

Any ideas of how to stop this happening - other than implementing a higher UVLO on the PWM pin?

  • 0
    TI__Mastermind 25500 points

    Hello Toby,

    Ideally you would adjust the UVLO as mentioned so that the device stops switching before you hit current limit. If you hit current limit you will get exactly what you see because the PWM FET is supposed to turn off if a fault is detected to protect the LEDs.

    You could also pull the PWM pin low before the input voltage drops to the current limit level and hold it there. That way when the fault clears the PWM FET remains off and you will not get the current spike in the LEDs.

    Regards,

    Clint

  • 0 in reply to Clinton Jensen
    Prodigy 40 points
    Thanks for that - as I expected really, need to shut it down before the input voltage drops off too much.

    Also, regarding the PWM - how can I reduce oveshoot on PWM turn on? Will increasing or decreasing C on the comp pin improve the response, or slowing down the turn on of the PWM FET help?
  • 0 in reply to Phuzzy
    TI__Mastermind 25500 points

    Hi Toby,

    Reducing the COMP capacitor can give you higher bandwidth and it can help with the overshoot. But a simpler way is to actually slow the FET turn off rather than turn on. You can use a schottky to get a fast turn on and a parallel resistor to get a slower turn off. If tuned correctly (which is pretty easy by experimentation) the output capacitors will discharge just the right amount due to the slow turn off that the current will not overshoot at turn on.

    Regards,

    Clint