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.

TPS92662-Q1: Power and LED topology questions

Part Number: TPS92662-Q1

I'm a software guy slowly getting deeper into hardware. My next project will be in the architectural lighting space, using TPS92662 for pixel/color control of a large fixture, where each pixel is a 6-channel, 1a per channel LED. I'll need numerous TPS92662's, probably about 50 per fixture. I've got some fairly basic questions:

  1. With the recommended boost-buck power design, am I correct that I can use a single boost converter to get to the appropriate voltage (36V in my case), and then run multiple buck converters from that source to power each individual 92662? Is there a practical limit here, other than reasonable amperage for the application?
  2. I would like to drive two 6-channel LED's from each 92662 to reduce cost. But this would mean at least one, possibly both, would be on a separate MCPCB with a ~ 6" cable to the board with the TPS92662. How bad of an idea is this, and what should I do to mitigate inductance if it's not a terrible idea?
  3. Is there guidance on distance between 92662's when using multi-drop UART? At what length should I switch to CAN?

  • Hello Brooks,

    Hopefully this will answer your questions:

    1. You can run as many buck converters from a single boost converter output as you would like as long as the boost is capable of delivering the power that number of buck converters will require. Even our EVMs run multiple buck converters from a single boost stage, so that is no problem.

    2. That should be fine actually. When you are shorting/opening LEDs you want to use no (or very little) output capacitance on each buck converter. This means that whatever parasitic inductance you introduce with wire length is simply in addition to the buck power inductor as they are in series. EMI may be the main concern here, but operationally it is fine.

    3. For any clock frequency/bit rate we recommend 8" to 12" maximum between the 662 devices. But if you are using lower bit rates you may be able to extend this. It really depends on signal integrity, external noise, reflections, and also EMI. So it will really need to be tested in your particular system. If you do not need high bandwidth you will have much more flexibility. We don't recommend CAN between 662 devices due to the Tx bus parking as described in the datasheet. However, if you really need very long distances you could consider running a separate CAN to each device although that may become more costly.

    I hope that helps.

    Regards,

    Clint

  • Thanks, Clint. This helps!