Trimming heels under the component body

Hi Tom and the rest of the forum members,

I recently posted to the Altium Designer forum regarding a land pattern in the Altium vault that has land pads that extend under the component body.  In response, someone posted a link to a slide presentation that Tom Hausherr wrote that says that heel trimming is now deemed unnecessary.  Here is a link to the document I'm referring to (page 19):

http://communities.mentor.com/mgcx/servlet/JiveServlet/download/28883-8838/PCB%20Design%20Optimization%20Starts%20in%20the%20CAD%20Library.pdf

Thanks for preparing this document.  It has a lot of great information in it.

Would you please explain the reason(s) why it isn't necessary to trim the heels?  In the component I've been looking at, the A1 distance has a minimum value of 0.05mm, which would put the body of the component into the solder paste (assuming the component leads are pushed 0.05mm or more into the paste, which I expect it would).

Thanks and regards,

Greg

Hi Tom,

Thank you very much for your detailed explanation.  I am really looking forward to the new calculator!

I need some help understanding what you said about Note 4.  I just did a little research on leadframe materials and it looks like Alloy 42 is still used (true?), though copper is is being used more for its superior heat transfer ability and a CTE that matches the solder much better.  Is the concern in J-STD-001 that of a CTE mismatch?  (In my opinion, the statement "or similar metals" is a pretty vague statement.  Similar in what regard?  Shiny? .  A friend of mine is a machinist and someone wanted him to cut some metal for him.  He asked him what kind of metal it was, and the guy said "well, it's shiny").

It is good to know that the solder never wanders under the component package to create a solder bridge.  It makes sense that the solder would have an affinity for the metal of the pad and lead and so it wouldn't push outward.

I'm happy to know there haven't been any assembly issues with regards to solder under component bodies.  But what about later on in the life of the product?  If the component body pushes down on the solder, what happens when it cools?  I expect it would shrink a little.  My concern is the stress on the package after it cools, since it would be constrained by solder on all sides that have pins.  I am thinking thermal stresses are probably not a concern as long as the solder and package shrink a little when they cool off.  The package and solder won't be seeing those temperatures again so they probably wouldn't interfere with each other in the future, assuming a positive CTE for each.  (The only issue I could see is if the component sticks to the solder on one side and gets pulled that way as it cools).  But what about mechanical torsion of the PCB (possibly from vibration)?  Might that cause stress on the parts and possibly crack them?  I expect the leads will bend better than a component body would against the solder, but I could be wrong. Maybe the plastic body is more pliable than I'm thinking.

I hope you don't mind all of these questions!

Thanks and regards,

Greg Cary

"Don't fix something that is not broken" can be looked at from both angles.  Has anbody complained about solder joints since the calculator started to trim lands?  The IPC calculator has been around for some time and it would be interesting to hear about issues related to using trimmed lands.  My thought is that it makes sense not to place solder under component bodies that sit so close to the board...no matter the lead material.

I am aware that the feature is still there, but off by default.  I was more interested to know if there if anybody has ever commented about solder joint reliability due to trimmed lands.  Thanks for the reply.