I need a double E (two E in front of each other) pattern for a conductive rubber keyboard, like
__ 1A__ __
__ 2A __ 11A__ __ 2_
__ 2A __
__ 1A__ __
(I could’t attach a picture:(( )
The two real connectors are pads “1” and “2”. The other 1A and 2A’s are different pads, which --of course-- are not connected to anywhere n schema. The only problem is that DRC doesn’t like it: pads “1” and “1A” are too close to each other.
I also tried to have 4pcs of “1” and 4pcs of “2”, but I had to manually wire the same-name-pads together… Otherwise DRC considered them unconnected.
It would be fine having implicit wires in footprint declaration: all the pads with same name could be connected together when footprint gets placed.
Or having an option when there are pads with the same name
need to be connected together or
any one of them must be connected only
(Think of TO packages, where big area is connected to middle pin internally.
I could choose how to handle:
to connect any one of the area and the small pad, or
need connecting both separately…or
wire them together.)
Or making pad goups, where the members may be connected to each other, but it is enough to connect only one member of the group to pass DRChecking.
I can’t clearly see, what should be (and may be) done.
Opinions?
That is quite easily done by drawing lines in the footprint editor. As of now (6253) one is forced to draw on a non-copper layer first and then switch the lines to copper.
Yes, that’s it. Thank you. (How did you upload the picture? Did my low status prevent me from using picture? Drawing with proportional characters is not easy…)
But the real question is how to handle same-name pads iwithin a footprint. Need to connect all together or only one of them?
Pads with the same number are considered connected, if there is “sufficient” overlap between them. If the pads are spaced apart, DRC will always flag them.
Hmm…not working for me. I have a testpoint footprint: 4 overlapping rectangles in diamond shape with a small hole left in the middle to hold the pin of probe. All 4 pads are named “1” and ~1/3 of them overlapping. DRC still considers them “not connected”. Two scrennshots: “unconnected pads”, and footprint in editor with one pad relocated (overlapping to be seen)
I see. Then please add to the end of the wishlist: instead of center-center distances use some vector math to correctly decide if two rectangles (2x4corners) are overlapping or not (calculate the intersections of sidelines and check if any of intersecting points are within the sidequadrants).
Till then I’m littering the footprint with a lot of cute little pads.
The diode footprint (SOD232) consists of 2 SMT pads for each pin. One contains information for the soldering mask and the other the information for the copper.
They are on top of each other and have the same pin numbers to boot… still, the analyzer assumes something not being connected there.
I attached the footprint as well in case someone wants to have a look.SOD232.zip (53.6 KB)
The only problem, that if I got 48 keys on my keyboard and each of them generates 4 DRC errors, then hard to check 192 errors if there is a hidden 193th…a real one. I’d prefer the much safer DRC error free pcb.
I don’t pry into why having such a large soldermask cutout, but you may try to name soldermask-pads as 1A and 2A. They will be unconnected as these pin names are not used on your schematic. I don’t know if DRC blames two different-net pads on each other.
On the other hand this is the case what madworm mentioned: two pads with the same name and their center is at 0 distance. They should be considered connected, so no DRC error should be received.
Like in kindergarten: if you show me yours I show you mine :
my testpoint with overlapping pads. Madworm is right, this don’t produce DRC errors!!!
for that matter it’s a terrible solution
[quote=“novaktamas, post:9, topic:1488”]
…
I don’t pry into why having such a large soldermask cutout, but you may try to name soldermask-pads as 1A and 2A. They will be unconnected as these pin names are not used on your schematic. I don’t know if DRC blames two different-net pads on each other.
…[/quote]
Oh, no… the cutout in the soldermask is the smaller one… the big red pads are cooling pads in copper.
DRC won’t let you route if you got 2 pads wit different nets too close together and switching off the rules while routing is not pretty ‘helpful’
As for your testpin footprint, have you tried one big pad that covers everything and is only on the mask layer (not on copper) and then less copper pads underneath? Looks like you put up a ring of like 23 small + 4 big ones in there? [edit] probably not, as I just found out further down…
So yeah, at least it works.
[edit]
found my error… I had the smaller pads on mask layer only… just tried to put them on copper too and I got rid of the visual disturbance.
:
