What’s the big blue area sticking outside? Is it an actual pad?
I downloaded the footprint and it was what I thought, and even more.
The THT pad has circular hole, and the copper ring of it is also circular. The footprint adds oblong edge.cuts hole on top of it, and also a polygonal oblong copper which is not a pad. Also the mask hole is a polygon, not part of the pad.
Delete the copper polygons and edge cut holes and edit the THT pad and hole to be oval.
I don’t understand why this is so. The SH1/2 pads are correct even though they have the same shape, just a bit different size.
Hi @rocketry
The best way to get a correct, working footprint (or symbol) is to start with something from the Kicad libraries. They work.
You start by finding something suitable to modify, “save as” it in a personal library, and modify it. One of the big joys of personal libraries is you can then copy/paste your Snap footprint into the same part and use it as a reference to get the Kicad version exactly correct, even temporarily overlay your modified Kicad footprint with the Snap to confirm changes. Just delete the Snap part when you find the Kicad part is correct. See below.
It doesn’t take long to set up personal libraries (if you haven’t already) or learn to use the available tools. It would probably take no longer than the time you have spent with your current USB problem.
1 Like
Just to wrap up my side of things, which looks like a totally different issue (apologies for hijacking) - I’m pretty confident it’s a bug, so I’ve submitted a report with examples if anyone’s interested:
1 Like
Your pads appear to have no nets assigned. This behavior is expected.
1 Like
You shoulde check if there isn’t more than one pad on top of each other.
That what I think from your screen capture. The pad seems to have different size at the TOP copper and at the bottom copper. The only way to do that is to stack 2 pads on top of each other.
But if the 2 pads don’t have the same number (or name) it will not work properly.
You could also run the DRC check on the board (by putting just the footprint on a new board), you may get some DRC error that could help to find the issue.
1 Like
Dear Bertrand and others, it’s not useful to do guesswork and try to answer based on that anymore – I already found the cause of the problem and explained it in a post above. Continuing a long thread can only cause confusion and cover the correct answer.
1 Like
I’d suggest marking it solved and closing it then?
I’d like to have the original poster to mark the solution. The thread could go on with possible KiCad specific solutions (like the suggestion given by jmk) and details about pad design.
1 Like
Yeah, but only give them a limited time. It’s more important long run to have the solution ticked for future searches. In a couple of days time, if no one bumps the thread, it’s forgotten. 
1 Like