2
\$\begingroup\$

I create one 4-layer board using Altium's PCB Wizard . By default, it set the plane pullback to 0.5mm, the keep-out area is 1.3mm from the board edge. That is, the top/bottom layer keep-out region is less then the power/ground plane, as below, the red layer are the power plane, and the blue one is the top layer pour.

enter image description here

Is it a good design considering EMC?

\$\endgroup\$
  • \$\begingroup\$ Suitable... for what exactly? \$\endgroup\$ – rdtsc Dec 18 '15 at 15:03
  • \$\begingroup\$ Um, mainly for EMC/EMI. \$\endgroup\$ – diverger Dec 18 '15 at 15:33
2
\$\begingroup\$

You don't want the signal lines to come too close to the edge of the ground/power planes and you definitely don't want them to cross over or they will emit EMI.

The default values sound reasonable to me unless you have some kind of special requirements.

The 0.5mm pullback on the planes keeps the copper from showing up on the edges where the boards are routed (as in cut with a router bit) or perhaps pushed up against a metal part (though in the latter case I would add a bit more clearance for luck in case the PCB manufacturer doesn't get the cutting registration perfect).

\$\endgroup\$
2
\$\begingroup\$

There is no problem with power and ground planes extending a little further to the edge of the board than other traces. The important thing is that all copper is far enough from edges for whatever process you are using.

The ground plane should extent at least as far as any signal, but the small distances you are talking about wouldn't make a significant difference either way.

\$\endgroup\$
0
\$\begingroup\$

The frequency of a quarter-wavelength of 1.5mm (5.0mm) is

\$\lambda=\frac{v}{f}\$

\$5mm=\frac{3\times 10^8}{f}\$

\$f = \frac{3\times 10^8}{5mm}\$

\$f = 60GHz\$

so for 1.5mm of exposed power plane, it will couple electrically (behave like an antenna) at 60GHz.

\$\endgroup\$
  • \$\begingroup\$ could you please elaborate a bit on your answer? What antenna is formed? What "coupling" do you mean? My understanding is that for EM radiation a standing wave must be formed (resonant condition), so how is the standing wave formed in the case? 10x. Your answer seems interesting but lacks details, imho. \$\endgroup\$ – Sergei Gorbikov Dec 21 '16 at 12:08

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.