• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 63 subscribers
  • Views 570 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.

LM5575 pcb layout

Gioacchino Fertitta
Prodigy 250 points
Other Parts Discussed in Thread: LM5575

Hello,

I have read the LM5575 datasheet and the application note "AN-1569 LM5575 Evaluation Board". The datasheet recommends that PGND and AGND are connected through the topside copper area underside the device and the app note shows this connection.
Given that LM5575 has two ground pins, PGND and AGND, I would connect the AGND through a narrow trace to the negative pole of the output filter capacitor rather than to the topside copper area. Doing so, the regulator should be able to get a more accurate measure of the output voltage. In my opinion, the solution stated in the datasheet would lead to a high noise level on the AGND pin because of the ground connection parasitics (resistance and inductance).

I'd like to know the reasoning which underlies the datasheet  (and app note) solution.

Can someone help me?

Regards

Gioacchino

  • 0
    TI__Guru 74035 points

    I will forward this to the correct person.

    F.D.

     

  • 0 in reply to Frank De Stasi
    Prodigy 250 points

    Ok, I'll wait...

    Thank you

    Gioacchino

  • 0 in reply to Gioacchino Fertitta
    TI__Mastermind 19901 points

    Hello Gioacchino, 

    I am not the author of the appnote you mentioned, but I will comment on this anyway :)

    The portion of the circuit with high di/dt will be from the PGND to the negative terminal of the input capacitor. This section will see pulsing current during each switching cycle. The location of this on the board is approximately the red path below:

    The AGND connects to the lower end of the center pad and is not on the way of the high di/dt current, but I agree that it is not too far away either. Also the di/dt current has to go to the bottom layer to return to the input capacitor, which will "pollute" the ground plane on the bottom layer. 

    I think an improved version of this layout would be if the input capacitor is placed on the "north" side of the IC. Something like this:

    The components that were on the north side can go somewhere else. That way the high di/dt current returning from the PGND to the input cap is only on the top layer, the parasitic inductance is lower (smaller loop area!) and the ground plane on the bottom can stay quiet. 

    Here is an example of how to place the power components along with some downloadable files: http://e2e.ti.com/support/power_management/simple_switcher/w/simple_switcher_wiki/2214.lm5575lm25575-pcb-layout.aspx

    In terms of the AGND connection,  the example in the link above works well - AGND is connected to the quiet bottom layer through the pad vias and there is no high di/dt current flowing in the bottom ground plane in this case. Another modification/improvement may be to connect the AGND pin directly to the bottom ground layer and not to the pad under the IC.  Your idea may also work well, but I have not tried it myself.

    I hope this helps.

    Regards, 

    Denislav 

  • 0 in reply to Denislav Petkov
    Prodigy 250 points

    Hello Denislav,

    thank you for giving such a comprehensive answer. You gave me some valuable insights.

    Gioacchino