• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 322 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.

TPS56637: Use CV synchronous buck converter for CC high power LED driver

Marius dddd
Prodigy 10 points
Part Number: TPS56637
Other Parts Discussed in Thread: INA180, TPS92640, TPS92641

Tool/software:

Hi everyone,

I want to use two TPS56637 fed from a 3-cell Li-ion pack for drive high power LED. One buck is set for ≈ 3 V, the other for ≈ 6 V (both up to 4 A). Each runs in FCCM to avoid skip-mode flicker. A 7 mΩ sense resistor plus an INA180 (gain 20) closes the loop on FB, so the parts behave as constant-current drivers. I chose it because WEBENCH® Power Designer shows me very good efficiencies at 96% over a wide current range. I want to be able to vary my current in several steps between min 100 mA and max 4A.

efficiency curve from WEBENCH® Power Designer

Dimming Solution with Op-Amp from TI Application Report (https://www.ti.com/lit/an/slva419d/slva419d.pdf?ts=1752345436091)

I still have three doubts:

  • High-speed PWM on EN. With the fixed 2 ms soft-start, a 25 kHz EN signal never lets the inductor current reach its target. Is the right approach to keep EN high and do any fast PWM with a downstream series FET instead or do analog dimming ?
  • Analog dimming through the INA is a good idea ?. I inject 0–1.2 V from a DAC onto FB (simple 10 kΩ divider) to cover roughly 50 mA → 4 A with no PWM. Any practical feedback on accuracy and noise at the low end
  • Parts choice. If there’s a TI LED-oriented buck with synchronous FETs ≤ 20 mΩ, good stock, and similar efficiency but an even cleaner CV-to-CC path, I’d love to hear about it or other solution. It is important to stay with a very high efficiency > 95% if possible.

Thanks a lot!

Best,

Marius

  • 0
    TI__Intellectual 2930 points

    Hello Marius,

    I will check this internally and get back to you.

    Thanks,

    Hitesh Chandwani

  • 0 in reply to Hitesh Chandwani
    TI__Intellectual 2930 points

    Hello Marius,

    1. High Speed PWM on EN is not recommended as it Switches BUCK converter ON and OFF.
    2. Analog diming with INA as shown in the application note is good Idea, accuracy of the LED current will depend upon the Rsense accuracy and the amplifier circuit you use, also consider the tolerance of the internal Vref.

    Noise is controlled by the Cin and Cout and the filter capacitor in amplifier loop. (You may need to apply the filter capacitor across Rsense refer INA datasheet you are using for filters to eliminate noise in the sensing path) .

    Risks and Prevention of ESD, EOS, and Latch Up Events for Current Sense Amplifiers refer Figure 3-5. Input Filtering for EMI Induced Noise

    Thanks,

    Hitesh Chandwani

  • 0 in reply to Hitesh Chandwani
    Prodigy 10 points

    Hello Hitesh,

    Thanks you for your response,

    After going through the datasheets and TI application notes, I’m leaning toward replacing the TPS56637 + INA180 idea with a synchronous LED buck based on TI’s TPS92640, driven by a low-RDS(on) dual MOSFET such as the SiZF5302 and a 1 µH inductor. I hope stay stay around 95–96 % from about 1 A up to 4 A on a 6 V LED string powered from a 3-cell pack lithium (around 12 V), and it offers straightforward analogue or PWM dimming without the soft-start penalties that show up when switching EN. Because the TPS92640 regulates current directly through a sense resistor, the external amplifier disappears, simplifying the bill of materials.

    I’m still not certain the real-world efficiency will actually match the optimistic WEBENCH numbers I was getting with the TPS56637, and I’m curious how well the TPS92640 holds regulation below roughly 200 mA If anyone has hands-on experience with the TPS92640 (or the TPS92641) in this current range, I would really value your thoughts before I finalise the choice.

    Thanks in advance for any insights.

    Marius

  • 0 in reply to Marius dddd
    TI__Intellectual 2930 points

    Hello Marius,

    To get more information about LED drivers please start a new thread for TPS92640.

    LED driver team can give you more insights on this.

    Thanks,

    Hitesh