SMT footprints

[William "Chops" Westfield]

Tell me about SMT footprints for passive component (resistors,  
capacitors, etc), and how sensitive automatic placement and soldering  
techniques is or isn't to the exact size of pads and whether they  
have vias in the middle or at the edges?

Ideally, I want to create some "universal" footprints for PCBs that  
would allow any of several different sized component to be mounted,  
for flexibility of component choice.  I can certainly make SMTs big  
enough for a 1206 part that are close enough together for a 0603  
part, and I'm pretty sure it will work OK for hand assembly, but I  
don't want to preclude automated assembly...

The most pressing concern is electrolytic caps; I would like to make  
a footprint that can accept SMT tantalum (rectangular package), SMT  
aluminum (cylinder/square package), OR radial TH caps with one or  
more lead spacings.

Is this possible?  Has anyone done it?  I'm only concerned with  
flexibility of part selection, not secondary effects like exact  
physical size or ability to route traces
between SMDs or whatever...

Thanks
Bill W

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Xiaofan Chen]

On Sun, Jul 20, 2008 at 10:00 AM, William Chops Westfield
<westfw@...> wrote:

> Tell me about SMT footprints for passive component (resistors,
> capacitors, etc), and how sensitive automatic placement and soldering
> techniques is or isn't to the exact size of pads and whether they
> have vias in the middle or at the edges?
>
> Ideally, I want to create some "universal" footprints for PCBs that
> would allow any of several different sized component to be mounted,
> for flexibility of component choice.  I can certainly make SMTs big
> enough for a 1206 part that are close enough together for a 0603
> part, and I'm pretty sure it will work OK for hand assembly, but I
> don't want to preclude automated assembly...

Not a good idea for the process and the process engineer will just
reject your board outright. If the pads are too big, the part will
flow. The pads are designed in a way that a "self-center" effect
will make the parts sit right in the desired position.

I've seen boards (low cost vendors) using pads for many
0603/0805 resistors. And we will just consider this
supply not acceptable.

Overlapped pads are bad idea. Quality is never a secondary
effect.

Xiaofan
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Tony Smith-11]

> Ideally, I want to create some "universal" footprints for
> PCBs that would allow any of several different sized
> component to be mounted, for flexibility of component choice.
>  I can certainly make SMTs big enough for a 1206 part that
> are close enough together for a 0603 part, and I'm pretty
> sure it will work OK for hand assembly, but I don't want to
> preclude automated assembly...


A small part on a big pad isn't a good idea, they tend to float off to one
side.  I had a board here like that a while ago, the parts were on all sorts
of funny angles.  It still worked though, just looked a bit odd.  I guess it
was a case of using whatever they had on hand.

Tony

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Djula Djarmati-2]

Xiaofan Chen wrote:

> On Sun, Jul 20, 2008 at 10:00 AM, William Chops Westfield
> <westfw@...> wrote:
>> Tell me about SMT footprints for passive component (resistors,
>> capacitors, etc), and how sensitive automatic placement and soldering
>> techniques is or isn't to the exact size of pads and whether they
>> have vias in the middle or at the edges?
>>
>> Ideally, I want to create some "universal" footprints for PCBs that
>> would allow any of several different sized component to be mounted,
>> for flexibility of component choice.  I can certainly make SMTs big
>> enough for a 1206 part that are close enough together for a 0603
>> part, and I'm pretty sure it will work OK for hand assembly, but I
>> don't want to preclude automated assembly...
>
> Not a good idea for the process and the process engineer will just
> reject your board outright. If the pads are too big, the part will
> flow. The pads are designed in a way that a "self-center" effect
> will make the parts sit right in the desired position.
>
> I've seen boards (low cost vendors) using pads for many
> 0603/0805 resistors. And we will just consider this
> supply not acceptable.
>
> Overlapped pads are bad idea. Quality is never a secondary
> effect.

What Xiaofan says is all true. We have many problems with customers who
send the same design they used for hand-soldering to be made by us.
Oversized pads and traces/vias not separated by soldermask from pad are
the most common problems.
Ideally, the pad should be symmetrical and just a bit bigger than the
metal surface touching the pad and surrounded by all four sides by
soldermask (and no vias inside the pad).
However, there is one trick that works for your purposes: make the pads
narrower (say 50%) than the width of the metal on the component like this:

          MMMXXXXXXXXXMMM
+--------MMM|XXXXXXX|MMM--------+
+--------MMM|XXXXXXX|MMM--------+
          MMMXXXXXXXXXMMM

The results are not perfect but this is much better than a huge "sea of
solder" pads. If you want I can take a picture of this layout.

Djula
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist