I'm making a prototype that uses, among other things, a bluetooth mcu connected to a 2.4GHz chip antenna. Its using a 4 layer PCB, and I'm thinking about what to do with the feedline, whether it should be buried on one of the middle layers, or left on the top layer.

I've been using this website (https://www.eeweb.com/tools/microstrip-impedance) to look at the various calculations. To get a 50 ohm line, it looks like the top layer would need to be 13 mil wide, and the buried one would need to be 7 mil. Both of these would be fine.

Is there an argument to be made for burying this feedline? Is that a standard thing done in industry (to reduce unwanted radiation, for example)?

  • 1
    \$\begingroup\$ How many more vias are you introducing on the feedline by burying it? \$\endgroup\$ – Lior Bilia Jul 4 '18 at 17:15
  • \$\begingroup\$ If the feed to the antenna is less than 1/10 the wavelength, the impedance doesn't matter. This is probably the case. \$\endgroup\$ – Leon Heller Jul 4 '18 at 17:28

As mentioned in comments, reducing vias in the signal path will likely improve performance.

If you have the space, there are advantages to a wider microstrip line:

  • Can handle higher power (not likely an issue for Bluetooth)

  • Better impedance control due to etching errors being smaller relative to the trace width

The trade-off is of course the board area used. You need to consider not just the trace width, but also the desire for 3-5 trace widths of clearance around the trace to maintain controlled impedance.

If there are no other issues, a top layer microstrip is probably better than a buried one. If the space required is too much, then consider using a thinner dielectric layer to allow you to reduce the trace width, rather than burying the microstrip.

One scenario where the buried microstrip (or stripline) would be better is if the board will be used where there are conductive materials (like the lid of the enclosure) close enough to the board to disturb the impedance of a top-layer microstrip.

  • \$\begingroup\$ Isn't it true as well that high-frequency losses are higher in buried microstrip, due to more lossy dielectric-resin around? \$\endgroup\$ – Ale..chenski Jul 4 '18 at 19:10
  • \$\begingroup\$ @AliChen, true but probably not significant at 2.4 GHz \$\endgroup\$ – The Photon Jul 4 '18 at 19:11
  • \$\begingroup\$ And of course highly dependent on the laminate material used \$\endgroup\$ – The Photon Jul 4 '18 at 19:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.