Clayton,
 
You can contact IBM Microelectronics Divison in Endicott New York (www.ibm.com). They can do 'Mechanical Deflection Test' for solder joints.
Also you may read Chapter 7 Failure Models, para 7.4.3, page 304 in 'Failure Modes and Mechanisms in Electronics Packages' (available at  www.wkap.nl) for some explanation and reference material on the method. One reference article is in 1996 IEPS Conference proceedings, pp. 435-442 by R. Kuracina and A. Zubelewicz on this subject.
 
The test methodology has been applied to prototypes or actual product in liu of thermal cycle test since test duration is very short for
mechanical stress test. A solder joint provides two functions, electrical continuity and mechanical retention. For large components like sockets, SMT connectors etc. mechanical retention is further aided by hold down mechanisms like nut and bolt or snap fit lugs. Pulling a lead and wiggling it  may not be a valid test. Bur you can assess the joint quality by 'Lead Pull Test' after the component has been cut off. The average pull values may then be compared to a calculated value based on wetted area and solder tensile strength.
----- Original Message -----
From: [log in to unmask]>Clayton Gardner
To: [log in to unmask]>[log in to unmask]
Sent: Thursday, March 09, 2000 4:22 PM
Subject: [TN] Mechanical Straining Solder Joints

Hi TechNet
 
Has anyone got any opinions on verifying solder joint quality by applying mechanical strain?
 
I have had a board rejected recently for "insufficient solder" on a corner pin of a wave soldered SO08 package.
 
The joint was inspected by Customer QC, who determined the joint was suspect due to dull colour and potential insufficient toe fillet.  To verify if the joint was good or bad, mechanical strain was applied to the lead using tweezers and wiggling it back and forth.  This caused the lead to pull from the solder joint.  Leaving a pocket of solder on the pad from where the lead was pulled (What was left of the pocket exhibited good heel fillet and sufficient toe fillet).
 
Customer QC opinion is that if a solder joint is good you should not be able to pull the lead from the joint and if too much force is applied the pad will pull from the board not the lead from the joint.
 
My understanding is that a solder joint is not meant for mechanical strain only good electrical contact therefore by applying force even a good joint will deteriorate.  Therefore, even if the lead doesn't pull from the joint, the quality of the joint has been affected.
 
Any comments?  Would anyone classify this as a valid verification technique? How do you verify a suspect joint?
 
Thanks
 
Clayton Gardner
Engineering Manager
 
A.E.M.S
11-13 Fiveways Boulevard
Keysborough VIC 3173
AUSTRALIA
Ph. 61 +3 9701 5499
Fx. 61 +3 9701 5422
Mb. 0416044552
Email:  [log in to unmask]
WEB:  www.aemspl.com.au