On Tue, 23 Aug 2011 07:15:54 -0400 (EDT)
Cory Papenfuss <papenfuss@xxxxxxxxxxxxxxxx> wrote:
Hey, all. I've used PCB on an off for 10 years and recently
have been getting familiar again with the rest of gEDA which has
become a great set of tools!
Anyway, I've been using a board mill to make 1 and 2-layer
prototypes of typically through-hole components. The trouble is when
the autorouter chooses to use a component pin as a via to transition
to the other layer. If it's a part that doesn't have room to solder
the topside (most parts... DIPs, ribbon-cable connectors,
can-electrolytic caps, etc), it's very difficult to connect both
sides.
I typically check manually, but it's tedious and I often
miss some. Is there any way to set up the autorouter to ignore
component-layer pads, but still route on the component layer
elsewhere? I've scoured the web for ideas but come up only with
kludging a "no-autoroute-zone" by placing copper on the component
layer before autorouting.
Since you have no plated holes, the component footprints are really
incorrect. Through-hole footprints generally specify plated holes for
all pins. However, I think this does not need to be the case. I've
had similar problems with manual routing, when I accidentally route
traces to through-hole pins on non-plated-through-hole boards which
necessarily be soldered on that side.
Here is an idea of how you could use a footprint that specifies that a
copper annulus exists on the far side of the board only. Instead of a
plated pin, it includes a pure hole and an SMT pad on the far side.
You can test it and see if it does what you want. You could also put a
pad on the component side too, if you were going to solder the leads
there as well, and thus create a via with the component lead.
See attached footprint file.
Regards,
Colin