Mail*Link(r) SMTP               FWD>DES:ASSY: component mounting

--------------------------------------
Date: 5/29/96 4:11 PM
From: [log in to unmask]

[log in to unmask] inquired about the following
clinched-lead requirement:
     
>In reference to IPC-D-275 Section 4.2.1.2 Clinched Leads (Type 1) the third 
>paragraph states "Type 1 is not applicable to leads of dual-inline packages 
>(DIPs)...." Q
     
>Question: What are the technical reasons why Type 1 clinching is not 
>considered applicable for DIP packages ?  

To my knowledge, the only known reason is the difficulty of replacing a
component and mimimizing the probable damage to the printed board with a
"high" number of component leads.

With straight-through, Type III leads, it's relatively simple to remove the
solder with well trained personnel and good solder removal equipment.  The
"partially- clinched", Type II, component leads it is still relatively easy to
remove the solder.  With both Type II and Type III component lead mounting,
there should be very minimal damage to the printed board during solder removal
operations -- it's called good process control.

With Type I "full-clinched" component leads solder removal is generally a
two-step procedure:  First to straighten the component lead's clinch; then
secondly to remove the solder.  In all cases, after the solder is removed, the
component leads are checked to make sure they are free in the holes, then you
can remove and replace the component.  This generally requires two heating
operations, though sometimes the solder extractor tip can be pushed up against
the the end of the component lead, and as the solder melts hollow end of the
solder extractor tip is forced under the component lead so as to be able to
remove the clinch as the solder extractor is angularly rotated to "straighten"
the component lead, then after the component lead is straightened the
extractor tip is "wiggled" (generally in a circular patter) to make sure (by
feel) all the solder in the joint is molten, then the solder is extracted
(sucked out).  Type I fully-clinched component leads does add to component
replacement complexity.

There are other component removal methods, such as straighten the leads and
then use a solder fountain or a "gang component lead heater block" and then
use the ol'  "heat-pull-and-pray" method.

>We have a customer who has designed a Class 3 board where all DIP packages 
>(as well as all pins for each DIP) are PTH with full clinch pads, the 
>clinch is towards the body. 
     
>Are there mechanical advantages to this method ?

A somewhat serious (;-) statement:  The component will definately not fall out
of the printed board in mis-handling prior to soldering, and it wont "lift-up"
out of place during flow soldering. 

I haven't seen any good reliability data that supports the need for
fully-clinched component leads for most environments.  IMO, the
fully-clinched, Type I, component lead requirement is still there because of
one of the USA's military organization's wishes --- IMO they just wont give up
-- IMO (;-) if they could, they would wrap the lead around something, tie the
lead in a knot, then solder it.
      
>What are the disadvantages ?
     
As mentioned above, it makes component removal/replacement more difficult.

>This board is subject to both high temperature and high vibration.
     
This one is best for the metallurgists and reliability techneters.  IMO, the
concerns are how high is a high temperature (how "strong" is the solder at
temperature), what level of vibration, what is the mass of the component(s),
what's the thickness of the printed board, and what are the combinations and
durations.

IMO, I'd take an exception to D-275 and recommend a Type I.5,
partially-clinched componet lead as a reasonable compromise, clinch it to
about 60 degrees off the axis (normal of the printed board) of the (assumed)
plated-through hole.  This approximates the full-clinch, still lets you to get
at the end of the component for component replacement, and minimizes damage to
the printed board. 

In contrast -- If you've got super-good high-quality componets, you have very
few ppm failure rates with no component replacements (this includes ALL REWORK
- or whatever else you want to call it) -- then take the risk -- go for it,
Full-Clinch, Type I, and hope you don't kill an assembly during component
replacement.

>Any advise would be helpful.
     
>Thanks
      
>D.Drake
     
Ralph Hersey
e-mail:  [log in to unmask]     
     


------------------ RFC822 Header Follows ------------------
snip-snipped all