1
\$\begingroup\$

I recently designed and ordered my first batch of PCB's (first ive ever created) They were great except for one thing; I thought that setting my SMD pads layer to 'All' in the software would create a pad which was accessible to all PCB layers. The pad was created on both layers but was not electrically connected. So my question is: what is the standard way of allowing an SMD pad to be accessible from multiple layers? Would it just be to use a via? Many thanks.

\$\endgroup\$
6
  • 1
    \$\begingroup\$ besides that there is the question why you would want to have smd pads on inner layers, of course they are just the pads, you have to specify if you want to have them connected at all. good pcb design software should be able to show you a 3d model or thelike of the finished product which you can use to verify your assumptions. \$\endgroup\$ – PlasmaHH Nov 29 '17 at 13:04
  • \$\begingroup\$ Connect it (on its own layer!) either to a via or a leg of a through-hole component, if there is one nearby on the same net. \$\endgroup\$ – user_1818839 Nov 29 '17 at 13:07
  • \$\begingroup\$ I would also wonder which is a better practice: 1) put the via on the pad, or on the side? \$\endgroup\$ – Nazar Nov 29 '17 at 13:18
  • 1
    \$\begingroup\$ Outside the pad ... unless you have cleared it with your assembly house first. A via in the pad can suck molten solder away from where you need it, leaving random open circuits. \$\endgroup\$ – user_1818839 Nov 29 '17 at 13:24
  • \$\begingroup\$ of course pads would not be accessible to components on inner layers, i just thought (in my naivety) that setting the pad layer to 'all' would create a pad which went all the way through the board. I now see that this was incorrect. The consensus is to use vias though i see. I am considering putting them in the middle of my 1206 components pads where possible to save space. is there any reason why this would be advised against? \$\endgroup\$ – user2105725 Nov 29 '17 at 13:51
1
\$\begingroup\$

Surface mount device (SMD) pads, are, well, on the surface. Since the point of these pads is to connect to the device, it makes no sense to have these pads on inner layers the device can't possibly touch.

What you are apparently asking about is how to make connections between different copper layers of a board. That is done with special plated holes called vias. A via is a hole in the board, but the inside is plated with copper, and traces from each layer can connect to this plating. Vias connect all the traces that touch them from all the different layers.

You seem to be confusing SMD pads and thru hole pads. Since thru hole pads are for connecting to a pin of a device that extends thru the board, they are made much like vias. You can think of a thru hole pad as a pad and a via all in one.

One of several reasons we don't like thru hole devices is because the pads interfere with all the layers. This puts more constraint on routing and requires more space as a result.

With SMD, you don't have this problem. All the connections to the devices are on top (or sometimes also on the bottom) layer. During placement, you try to minimize the distances between connections and the cross-overs. You then use vias only when you need to switch layers. You can run unrelated signals right under device pads on other layers.

\$\endgroup\$
2
  • \$\begingroup\$ OTOH adding blind vias to a board dramatically increases its price \$\endgroup\$ – Dirk Bruere Nov 29 '17 at 14:08
  • 1
    \$\begingroup\$ @Dirk: Yes, for blind or buried vias. Normal vias that go completely thru the board add little or no cost in most cases. \$\endgroup\$ – Olin Lathrop Nov 29 '17 at 16:06

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.