• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 63 subscribers
  • Views 366 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.

TPS1663: Why is 10nF the recommended minimum value for the dVdT capacitor?

Dave Mesquitta
Prodigy 20 points
Part Number: TPS1663

Hello,

I'm currently evaluating the performance of a prototype design, originally made by someone else (and no longer contactable), using the TPS16630 (QFN package). This is primarily used in a 24V industrial device, to enforce a current limit of approx 5.5A to downstream circuits (externally connected circuits).

We've experienced some issues with startup, under full-load - that have been identified as relating to the device's thermal regulation during startup, effectively due to the inrush current and short-term power-dissipation across the device. We have plans to vastly improve the power handling, through better PCB layout (heatsinking of the device) and switching to the TSSOP version. I am aware also of the design calculator tool for this device, to help estimate whether a given design *should* be able to start-up or not.

However, my question specifically relates to the dVdT capacitor. I understand the the function of this component, but it has been noted that the datasheet:

  1. mentions that "The fastest output slew rate of 24V/500 μs can be achieved by leaving dVdT pin floating." (section 9.3.1); but
  2. also lists 10nF as the recommended minimum value, according section 7.3

Is there a reason for this recommended minimum value?
Other than increasing the output slew-rate, is there any other known side-effect, or behavior that I should be aware of, by having a dVdT capacitor less than 10nF (or open/floating)?

Thank you kindly in advance.

Dave

  • 0
    TI__Expert 5575 points

    Hello Dave

    Welcome to TI E2E forum and thanks for considering TPS16630 for your application.

    As you are aware C(dVdT) capacitor is used to control inrush current during start up. This is to make sure the output ramp is glitch free.

    With C(dVdT) capacitor <10nF, there can be erratic current spikes during start-up and there is possibility that output ramp is not smooth. Minimum recommended value of 10nF is mentioned so that there specific inrush current limiting at start up and you can see smooth ramp even with higher inrush current values.

    There is no side effect as such apart from possibility of seeing a glitch at the start up.

    May I know your exact use conditions such as Input Voltage, Current Limit value, Output Capacitor and Output load during start up.

    As I see from your post, you are facing issues with start up with full load.

    I would like to highlight that, during start up with full load, device has to provide inrush current as well as load current. This will set your current limit value as (ILOAD + I-inrush) during start up. May be increasing ILIM threshold or C(dVdT) can help here.

    Thanks & Regards

    Abhijeet

  • 0 in reply to Abhijeet Godbole
    Prodigy 20 points

    Hi Abhijeet,

    Thank you for clarifying the 10nF (minimum) recommendation.

    In this particular application, the output load is technically undefined (the TPS16630 is actually being used to enable a supply to an external / off-the-board load). However, an overall 5.5A continuous (resistive) load is considered sufficient for our needs.

    • Input voltage is 24V +/- 5%
    • Current limit is nominally 6A (R-ilim = 3K)

    We are using an electronic DC load for testing, set to constant-current mode (and with negligible effective capacitance). From some scope waveforms obtained, we can see that the output current from the TPS16630 is ramping as expected (as is the output voltage), but the device is running into thermal regulation during startup & before the current-limit value has been reached.

    By changing C(dVdT), we can see the output slew-rate change as expected, and consequently change the power being dissipated by the TPS16630 during startup. With a fast enough slew, or reduced load, or use of freezer-spray!, we can get the device to startup.

    Whilst we anticipate that with an improved PCB layout, we certainly hope to be able to set C(dVdt) to 10nF - or higher; we wanted to confirm that by having this as a lower value (or leaving open), we would not be running the device out of specification somehow.

    Thank you kindly for your prompt feedback & suggestions!

    Kind Regards,
    Dave