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.

TPS92692-Q1: PWM dimming frequency

Part Number: TPS92692-Q1

Hi

I read some similar question on E2E. However I couldn't get the answer what I wanted. So that , please let me ask.

https://e2e.ti.com/support/power_management/led_driver/f/192/t/667924?tisearch=e2e-sitesearch&keymatch=TPS92692

https://e2e.ti.com/support/power_management/led_driver/f/192/p/622086/2292667?tisearch=e2e-sitesearch&keymatch=TPS92692%20dimming#2292667

Our customer are going to use this device at the following conditions;

  Buck-Boost ; Iled=2.5A ; SW Freq = 450kHz ; PWM dim Freq = 20~100kHz

How much is the max duty cycle possible at Fsw=450kHz and Fpwm=100kHz?

As it mentioned at past post on E2E I read, it is recommended between 10Hz ~1000Hz.  Is Fpwm=100kHz possible?

Regards,

Koji Hamamoto

  • Hello Koji,

    100kHz is possible, but the range will be limited. With series disconnect FET dimming there will be a minimum pulse width to ensure COMP settles each cycle. The min pulse width will depend on a lot of factors, but it is around 3 or 4 switching cycles (the 450kHz switching cycles). So 100% PWM duty cycle isn't an issue at any frequency, it is your minimum duty cycle that will be limited.

    100kHz is pretty fast for this type of dimming. So if you need a wide dimming range at such a high frequency you may need a two stage with a boost followed by a buck converter. The buck converters can dim much faster.

    Regards,

    Clint

  • Hello Clint,

    Thank you for your comment. I have some question about your comment. 

    >100kHz is possible, but the range will be limited.

      How much is the min duty possible? (How much the minimum duty of dimming possible at 100kHz?)

    >With series disconnect FET dimming there will be a minimum pulse width to ensure COMP settles each cycle.The min pulse width will depend on a lot of factors, but it is around 3 or 4 switching cycles (the 450kHz switching cycles). So 100% PWM duty cycle isn't an issue at any frequency, it is your minimum duty cycle that will be limited.

      At the datasheet, it is described that the dimming range is 1000:1. What is the frequency condition? 

      As you mentioned , COMP voltage does not follow the response if the dimming so fast. We need some switching time ( 3 or 4 switching cycle?) for the response.

      So, that means the 3 or 4 switching cycle correspond to the min pulse width of the dimming. Is my understanding correct?

     

    Regards,

    Koji Hamamoto

  • Hello Koji,

    Yes, 100% duty cycle is always possible because you are not actually dimming, just running continuous.

    So the limitation is minimum duty cycle. 3 to 4 cycles is really just a typical number. It will depend on control loop bandwidth, switching frequency, inductor value, output capacitance, the speed of the dimming FET, and other factors. So unfortunately it really would need to be tested out. But it is likely that at 100kHz you will be limited to 40% or 50% minimum duty cycle or so. As I mentioned, these are not really designed for very high speeds and contrast ratios at the same time.

    The 1000:1 number given in the datasheet is at a much lower frequency, probably 250Hz or less.

    Regards,

    Clint

  • Hi Clinton-san,

    Thank you for the information.
    I had talked with the customer about this device. They are concerned how much the min duty is possible at Fdim=100kHz.
    And so I explained as you mentioned. (min duty may be 40 ~ 50%)

    They do not want the frequency into audible frequency (30~ 50kHz). If we keep 3us or less of min ON time, how much should we set the switching frequency? (They want min Duty 10% or 20% at Fdim > 50kHz.)

    Best Regards,
    Koji Hamamoto
  • Hello Koji-san,

    They might be able to get down to 20% at 50kHz, but it would need to be tested. They would have much more luck at lower frequencies. Isn't >20kHz above audible range?

    In any case it would be worth testing on an EVM to see if the performance is what they need. But as I mentioned, at such high PWM frequencies a buck converter (two stage boost + buck) would be able to do much better.

    Regards,

    Clint

  • Hi Clinton-san,

    Thank you for the information. I talked with the customer and suggested your recommendation.

    Regards,
    Koji Hamamoto