RE>ASSY: Solder joint mech tests 7/31/96
Valquirio,
I spent several years working on this sort of stuff in a former life at a large OEM. Because of the lack of standards and procedures, we wound up generating our own.
We worked on gull-wings, J-bends, butt joints, TABed joints, and many others from some relatively "strange" development efforts!
Major issues to consider when approaching this type of testing are the rate of pull, the angle to pull at (and how to keep this angle constant through-out the test), how to remove the device body without damaging the solder joints, effective gripping of the leads, how to analyze the load/time curve, selection of appropriate/representative joints to test, and choice of how many to test.
We had some difficulties agreeing internally on the applicability of these techniques. Some people argued as to their viability in predicting reliability, while others claimed that they were good process development tools. Still others thought that they were simply a way to stay employed ;^). I tend to think that they have *some* utility in optimizing certain assembly processes -- perhaps tweaking a reflow profile or choosing between a couple of different solder pastes. There were many instances where I don't think that it gave very useful data, mostly at times when people were trying to use this for reliability information. Correlating a pull force to the properties which tend to influence solder joint life (geometry, creep/fatigue, lead properties, etc) is difficult if not impossible to do.
Regards,
Greg Bartlett
Mercury Computer Systems
Chelmsford, MA
[log in to unmask]
--------------------------------------
Date: 7/31/96 8:44 AM
To: Greg Bartlett
From: [log in to unmask]
I am a university student who is looking to ascertain the strength
of SMT solder joints. I would like to know if there are any standard
industry practices which are used to measure the force required to
pull an SMT solder joint apart. I am mostly interested in solder
joints involving J-type and gull-wing leads. Any information would be
appreciated.
Thank You,
Valquirio N. Carvalho
[log in to unmask]
***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to: *
* [log in to unmask] with <subject: unsubscribe> and no text. *
***************************************************************************
------------------ RFC822 Header Follows ------------------
Received: by mailgate.mc.com with SMTP;31 Jul 1996 08:39:19 -0400
Received: from firewall.mc.com (firewall [192.148.197.15]) by jericho (8.6.11/8.6.11) with SMTP id IAA21935 for <[log in to unmask]>; Wed, 31 Jul 1996 08:39:09 -0400
Received: by firewall.mc.com id AA04483
(5.65c/IDA-1.4.4 for <[log in to unmask]>); Wed, 31 Jul 1996 08:39:07 -0400
Received: from ipc.org(168.113.24.64) by firewall via smap (V1.3)
id sma004481; Wed Jul 31 08:39:03 1996
Received: from ipc.org by simon.ipc.org via SMTP (940816.SGI.8.6.9/940406.SGI)
id HAA05535; Wed, 31 Jul 1996 07:31:17 -0700
Resent-Date: Wed, 31 Jul 1996 07:31:17 -0700
Received: by ipc.org (Smail3.1.28.1 #2)
id m0ula1i-0000OaC; Wed, 31 Jul 96 07:05 CDT
Resent-Sender: [log in to unmask]
Old-Return-Path: <[log in to unmask]>
Date: Wed, 31 Jul 1996 08:06:12 -0500
From: [log in to unmask]
To: [log in to unmask]
Message-Id: <[log in to unmask]>
Resent-Message-Id: <"KBgC73.0.vcL.0mq_n"@ipc>
Resent-From: [log in to unmask]
Subject: Unidentified subject!
X-Mailing-List: <[log in to unmask]> archive/latest/5459
X-Loop: [log in to unmask]
Precedence: list
Resent-Sender: [log in to unmask]
***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to: *
* [log in to unmask] with <subject: unsubscribe> and no text. *
***************************************************************************
|
