Author Topic: Question about isolated ground below switching power supply in nRF7002 layout.  (Read 178 times)

0 Members and 1 Guest are viewing this topic.

Offline i509VCBTopic starter

  • Newbie
  • Posts: 3
  • Country: us
Hi, I am working on a PCB which contains an nRF7002. Some here would say "use a module instead": these are just to play around with.

Nordic's datasheet (product specification) advises to put a cut in the ground plane under the switching element and related:

Quote from: The datasheet
Connect to an isolated metal plane on top layer with vias to isolated metal plane on layer 2 connecting only near thermal vias on the IC Paddle. See the layout in layer 2.

An image of layer 2:
2340363-0

Is this advisable to copy? My intuition would say the cut in the ground plane makes an antenna. Therefore I should use a solid a ground pour. I could see the specific routing to the exposed pad on layer 2 "channeling" the return path but I am skeptical of this.
 

Offline RoGeorge

  • Super Contributor
  • ***
  • Posts: 6538
  • Country: ro
Is this advisable to copy?

That is the reference design from the nRF datasheet, so yes, copy all layers as verbatim as possible.  There are more factors at play.  That design is coming from a company that sell billions of RF chips since many generation of nRF devices.  I expect their PCB design was already optimized for best performance and least EMI.

I'm not sure I understood your argument for not following the reference design, but you can try 2 small batches, one with the cut and the other with solid layer, and compare their EMI performance.

My bet is the reference design will work much better as it is, than without that cutout.

In general, the energy is in the space between the going and the return wire, never in the wire (or the copper plane itself), so the nearby layers and the vias between layers matters the most.  (I'm talking in the spirit of this video, it's a long one but it worth a watch:  https://www.eevblog.com/forum/chat/fun-for-nerds/msg5583117/#msg5583117

If Rick Hartley's guidelines seem strange, see the book "Fast Circuit Boards" by Ralph Morrison, where it is explained the physics behind those PCB decisions.


Offline Sagar

  • Contributor
  • Posts: 25
  • Country: in
usually I have seen this type of thing with crystal oscillators and ceramic rasonators. But I think you can follow the datasheet, the cut will not make a big difference. How many layer stack-up you are using for the PCB, and how you are planning to get hands on actual PCB.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf