• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 62 subscribers
  • Views 858 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.

TPS92511 Ripple Current

Pouya Kamalinejad
Prodigy 50 points

Other Parts Discussed in Thread: TPS92511, TPS92513

Hi, 

I am using TPS92511 for a fast LED flashing application (20uS on , 2500uS off) and reading the response (reflection) with high accuracy.

Due to the sensitive nature of the measurement,   minimizing the ripple during the on-time is very desirable.

Based on my knowledge of buck converters and according to datasheet, there are two option to decrease the ripple:

A- Increase the inductance value 

B- Use an output capacitor (in parallel with LEDs)

Here are my questions:

1- Is there a limit (maximum value) for the inductor?

2- Is there a limit (maximum value) for the output capacitor?

3- How these parameters (inductance and output capacitance) affect the rise/fall time of the LED current?

Ideally I want the LED current to be as close to a square wave (20uS on - 250uS off) as possible, with minimal ripple.

Note1: I have used similar values in the datasheet (8.2: Typical application) with no output capacitor, so a large ripple (25%) is expected. For the next iteration, I am trying to minimize this ripple while not affecting the rise/fal times.

Note2: To further clarify, I am using the PWM dim input to switch the current on/off (I am applying 20uS ON time which is longer than the minimum pulse width of around 6uS). I use the Iadj pin for analog dimming (injecting current).


Note3: In reply to Niq question "TPS92513 ripple Current Requirements", Clinton responds: "you can use an output capacitor and the LED ripple current can be as small as you want to make it."

Thanks

  • 0
    TI__Mastermind 25500 points

    1. The maximum practical limit is probably around 2.2mH. Any higher and the cost and size would probably become unreasonable.

    2. There is no limit to the output capacitor value, but output capacitors slow the rise and fall times of the current waveform. If you need to dim fast it's best to use a low value (perhaps 0.1uF or less) or none at all.

    3. With no output capacitor it is easy to know the current rise/fall times since they will just be v=Ldi/dt where the rise time v is the input voltage minus output voltage and the fall time v is just the output voltage. If you add a capacitor it will charge and discharge slowing the edges. How much all depends on how far the cap discharges each cycle, and the depends entirely on cap value and the dynamic resistance of the LED(s) used.

  • 0 in reply to Clinton Jensen
    Prodigy 50 points
    Thanks for the fast and clear answers.
    I will increase the inductance and accommodate the footprint for an output capacitor (just in case I may need to add it later).
    Just one other quick question and I am good to go. Is it correct to say all these apply to "TPS92513" as well? TPS92513 is more attractive as the analog dimming is adjusted through voltage (as opposed to TPS92511 which is cotrolled by current).
    Thanks,
  • 0 in reply to Pouya Kamalinejad
    TI__Mastermind 25500 points

    Sorry for the slow response, I'm currently on vacation. But the TPS92513 has a different control scheme. The analog dimming is nice, but not for very short pulse widths. It would work fine if you use a shunt FET across the LEDs, but it would not be fast enough otherwise. It also has a maximum inductor current ripple which is explained in the datasheet. But I believe it would be higher than for the TPS92511.

  • 0 in reply to Clinton Jensen
    Prodigy 50 points

    Thanks again for the clarification. It would have taken me a lot of reading to figure them all out.

    I will more carefully review TPS92513 but will most probably end up using TPS92511 based on your response and my experience with it.

    The addition of the shunt FET seems a bit of a hassle at this point as I am struggling to meet a deadline. I am getting acceptable dimming through the Iadj pin on TPS92511, so if voltage dimming is the only advantage of TPS9213, it is not worth all the extra design effort.

    Thanks for the clear response, happy holidays and enjoy your vacation.