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.

TPS4005x POWER PAD layout

Other Parts Discussed in Thread: TPS40057, TPS40055, TPS40303

Dear Engineers,

I'm designing TPS40057 power supply, and to avoid any problems I'm using reference design in http://www.ti.com/lit/pdf/sluu186.

Please review EVM datasheet in link for following questions:


1.Why is power pad isolated from gnd planes in layer 2 and bottom layer?

2. Why is gnd plane missing under SW plane in layer 2?

3.Why are input capacitors from both sides of high side FET? Is there any reason why to not have all input capacitors in one hive?

Thank you.

  • Looking forward for any reply because application notes and design guides don't answer these question at all, or only foggy describes problem behid them.

    Thank you.

  • That design is over 10 years old, and the original engineer is no longer at TI.

    Possible answers:

    The TPS40055 is a controller.  The main power dissipation will be in teh external FETs, so there is not a large requirement to get heat ot form the thermal pad.  It is electrically connected at the top side shunt and etch and vias near C9.

    It is not unusual to have the GND removed under the SW node to prevent capactive coupling.

    I suppose you could place the input caps either way.

     

  • Thank you for reply and clarify of questions. Does it have any performance cons to have power pad isolated in lower gnd layers? In other words, is possible to connect them without affecting noise performance?

    If this design/part is over 10years old, is still good idea to design low noise PSU with it? Power designed suggest this part for given performance 12V IN, 8V OUT, 20A Iout and out ripple 8mV PP.

    Thank you.
  • There should be no issue designing with it. So far as I know there are no plans to discontinue TPA40055. I'm not sure about "low noise". Typically it is difficult to get the ripple down to 8mV with 20 A current. Let me see if I can get a recommnedation for you.
  • TPS40303 seems like the consensus choice for your application.
  • Thank you very much for fast reply and tip. As it have spread spectrum control it looks very good for noise sensitive applications. I will try both designs for prototype to just to test SwitcherPro accuracy.

    Jiri Otisk