Now, I have a very odd footprint to make, for some parts that may have different footprints that force different solder masks… a total mess. We’re not even sure that all the combinations will reflow well, so maybe we need to try designing a prototype and see what happens.
See this draft, partially taken from a DXF
Suppose the two pads are all copper, while the paste has to go only on the orange areas…
Given the isles in copper with a thermal-relief-alike shape, maybe I won’t have to shape paste, and just fill all the two pads.
What I thought to do is a collage of rectangles, but it’s awful and… well, I’m not so sure of the result.
Besides I still don’t like it too much, I have a doubt.
When I add multiple copies of a same pad number/reference, they aren’t automatically connected together, even if make them overlap. This ends up having a few ratsnest still requiring a connection.
Now, if I designed the pads well(?) they will overlap or be close enough(?!) on the same gridline, and all is well. If I ignore those additional ratsnest (not very nice…) it doesn’t matter. But otherwise?
Is there some setting to avoid all of the same referenced pads to need a connection?
I dont know how KiCad internaly works but with my experiences if I place one PAD with the same pin name assigned (eg: GND) there are no complains about unconected things or ratsnest.
The two “forks” are with lateral fingers surely overlapping the base, while the center one is just touching it: I did it on purpose, to test what result I could achieve.
Note how strangely on pad 2 there’s no ratsnest from (connected) base to central larger finger, while on pad 1 exists also that ratsnest
I have a DXF from the mechanical designer who drafted it scaled with some help from the lab technician, and mmh… I’m not that sure they did a great job for reflowing and operating thermal transfer.
Whatever, this is a 10:1 detail I quoted
I did what @Joan_Sparky said about the center of the pads:
These pads are widely overlapping, and electrically they appear as a single connection point, no ratsnest.
The connection point shows to be on the center of the middle finger, no idea why is there though.
Wow thanks for your priceless time
The solution is akin to what I thought to do, setting some pad strips with just copper, plus others also with paste.
Has the only drawback of returning ratsnest, as you wrote before, when pad centers don’t overlap…
Will try this one, IOU
Yeah… sometimes you don’t get the centers covered without more pads and somehow I don’t like that.
I even got a footprint where the centers sit essentially over each other and still get ratsnest lines, depending which of the two pads I actually managed to connected with a track (BZR6071 last occasion, no idea how current version does it), but maybe I have to check that fp again. might need some rework ;-).
So, yeah. Nothing is perfect.
PS: at least for that footprint you could lay some tracks in there and connect them. The pad-outline mode is nowadays standard for me while putting down tracks because of stuff like that (and me seeing what I’m doing).
PPS: be aware that your dimension drawing didn’t specify the position for the pads for the SMD0603 really, I eyeballed it.
Yep, I realized it myself when I tried making the pads, but the post was already sent, didn’t want to spam
Also, I have already adapted it for the SMD 3014 layout, but next week I’ll have to get confirmation on pad sizes from people that prepared that DXF (their measures are a bit conflicting with the mfr datasheet I see).
I always get a ratsnest, if my pad is as an example just component side SMD
