Subject: | |
From: | |
Reply To: | |
Date: | Wed, 2 Oct 2013 11:53:41 -0500 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
You're a bit above what I would be comfortable with. Your gold content at nearly 6% is well over the recommended limit of 3%.
You would need to age the boards before testing to be sure of the reliability. When the gold absorbs into the solder joint, it forms partcles/plates/needles of AuSn4 in the matrix. Over time, the gold-tin intermetallics will migrate to the nickel plating interface, and when this happens, there will be a solid layer of this brittle material that is more prone to fracture with stress. We've used aging at 125C for 1000 hours to allow for this migration. At that point, testing should show an issue, although with the variety of tests that we've tried, none are quick. I would opt for a test with shock and vibration, since it's brittle fracture, not creep fatigue, that is the most likely failure mechanism. Plus it's faster, especially if you've already had to age the boards for 30 days.
What is your PCB surface finish? We've also found (with SAC305 and it should be true with SnPb) that if copper is available, it inhibits the formation of the solid layer of AgSn4. The individual particles are more likely to stay in the matrix, so are less likely to fracture. Less likely not unlikely. At 6%, you have a lot of gold-tin particles and much of your tin is consumed into the intermetallic.
|
|
|