• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 471 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.

Original question:

TPS92518EVM-878: Shunt FET dimming circuit

TPS92518EVM-878: Shunt-FET dimming

Craig Karageorge
Prodigy 220 points
Part Number: TPS92518EVM-878
Other Parts Discussed in Thread: LM3409, TPS92518

Can TI provide scope plots of current responses of an LED string using PWM diming and shunt-FET dimming? The turn-on and turn-off times using PWM diming as well as the current responses less than 25uSec pulse-widths do not meet my requirements.

  • 0
    TI__Intellectual 2590 points
    Craig,

    The turn-on and turn off times using PWM dimming of the main MOSFET are 100% inductor slew rate limited. Therefore, you can always calculate your rise time using the inductor slope (Vin-Vout)/L. With respect to fall time the equation would include the falling slope (-Vout)/L.

    When using a shunt FET, your rise and fall times can be in ns range if your shunt MOSFET is fast enough. The current in this case is constantly in regulation and simply being steered between the MOSFET and the LED string.

    Any waveforms we have would only represent the exact configuration we have. The datasheet contains several plots already.

    Whatever your requirement is can be obtained by changing your inductance in the case of main FET dimming, or changing your shunt MOSFET drive strength in the shunting case.

    -Jim
  • 0 in reply to James Patterson
    Prodigy 220 points
    Hi Jim, The data sheets do not have comparison plots of PWM dimming vs shunt FET dimming with the same loads, scales, etc.
    Can TI provide two scope plots, one of each dimming configuration that contains PWM input vs LED current output on a 5uS time scale ?
  • 0 in reply to Craig Karageorge
    TI__Intellectual 2590 points
    Look at the LM3409 datasheet. It operates almost identically to the TPS92518. There should be two plots that show shunt FET and main FET operation. Refer to figures 17 and 18 in the typical characteristics section on page 9.

    As you will see, it is many orders of magnitude different. When you use a shunt FET you can switch extremely fast (in the orders of ns). When you use the main FET you are inductor slew rate limited. Therefore it will be on the order of us or tens of us depending on Vin, Vout, and L. If you have a fast requirement, then you need to use a shuntFET. Otherwise, it is possible to minimize inductance which will minimize rise time and use the main FET.

    I don't have any better plots than these available at the moment and really the exact timing will be specific to your implementation.

    -Jim
  • 0 in reply to James Patterson
    Prodigy 220 points

    Ok, fair enough.

    What inductance value is associated with those plots?

    Unfortunately the forum reply option does not allow me to attach my scope plot, but at 20uSec pulse-width the converter startup delay is 5us, another 5uS for the current to ramp up in inductor and the current response is significantly distorted  using PWM dimming.  So for me to use this part in my design, I will have to use shunt-fet configuration.

    1) Does the controller annunciate a fault when the output is in the shunted state?

    2) I assume the inductor current rating for shunt-fet configuration needs only to be max peak current with 10% margin,  correct?

    Too bad TI did not include a gate drive and shunt FET on the eval board.

  • 0 in reply to Craig Karageorge
    TI__Intellectual 2590 points

    Answers:

    1. No, there is no fault associated with Vout low (shunted condition).   Since the TPS92518 is a SPI based device, even trying to retrieve the Vout information from the internal ADC will be asynchronous to the actual shunt state in real time.  In general, you should not need to to deal with that.  The converter will simply operate continuously in the presence of any load condition and the shunt FET will directly control the dimming obviously, steering the current between the switch and the LED.

    2.  The inductor current rating does not change in either condition.  It is always the same calculation as per the datasheet.  The margin you mention is completely up to you.  That depends on the inductor core material you choose.  Soft saturation core material can easily be designed with tight tolerance, but hard saturation requires proper margin to ensure you don't saturate in your corner cases.