TECHNET Archives

December 1997

TechNet@IPC.ORG

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Ed Holton <[log in to unmask]>
Reply To:
TechNet Mail Forum.
Date:
Tue, 23 Dec 1997 14:14:35 -0500
Content-Type:
text/plain
Parts/Attachments:
text/plain (109 lines)
Richard

I understand your frustration about having your question not answered, as
it has also happened to me in the past.  But don't give up the fight, you
should be applauded for asking the question.  That is a much better
attitude then never asking. . .

I do not have any hard numbers for you about thru-hole vs SMT components,
but your design should not be based just on component cost alone.  There
are numerous intangibles that must be considered:

Placement:  SMT has the capability of 6 sigma quality for placement while
TH is approximately 3 sigma.  For example, most TH machines will have 3
mis-insertions out of a thousand.  SMT is significantly better.  You need
to consider the quality improvements of SMT over TH and the increased
efficiency.

Optimization of process.  Design your board for all SMT and you only need a
SMT line.  Add in TH and now you need TH machines and a wave solder.

Efficiency/speed:  I have yet to see a TH machine reach 40K plus placements
an hour.

Soldering:  SMT reflow is capable of higher quality levels than wave
soldering (however I must brag that I am below 100 ppm with my wave solder
process! Thank you to Bob Willis, Ray Rua, John Maxwell and SEHO who have
helped me achieve this)

In my opinion, it is better to spend a few more cents for all SMT and
increase first time yields and throughput rather than some of the known
fallout due to TH and wave soldering.  However, keep reading because there
are many more caveats to consider. (I work in automotive electronics where
every cent counts, however I firmly believe that spending more on the
components will save on rework/repair cost.  These are hidden costs, that
the customer never sees except when accounting calculates the overhead.
Have gone many rounds on this with accounting)

Component cost.  Many SMD components are cheaper than TH, especially chip
capacitors and resistors.  But there are many high voltage devices and
connectors that are still less expensive as TH.  However, I have seen the
cost of many components drop such that when the program was first
developed, TH was cheaper, but within 3 years, the SM components were a
better deal.  Consider the future of components when designing.

DIP components.  While a DIP package may be more cost effective than an
SOIC, the placement issues between the two of them, plus the need for a DIP
inserter as a special piece of equipment vs. the SMD placement machines
that place chips to quads, we eliminated all DIP components from future
designs.

Look at commonization  of components to get better volume discounts, what
components are used in design today at your facility?

Optimization of process:  Some components are not available in SMT packages
and you are forced to use TH.  If this is the case, consider using more of
the same kind of TH components to efficiently use the TH machine (this
usually happens with radial components) and balance the process time
between the two machines.

The list goes on and on.  Basically, don't decide the design just on the
cost of the components, you will get burned and your M.E.'s will hate you
(haha).  Many other factors should drive the decision.  At a previous
employer, i started to develop design guide lines based on the
manufacturing capability of the production floor.  As a designer, you
should know what your capabilities are, and what the ME group has planned
in the future.

Here is a sample for you.  I proposed to our designers/engineers when
designing a new board:
1.  make it all top side SMT.  (one process and you are done)
2.  make it top side SMT with thru-hole soldered using paste in hole
technique.  (new process development)
3.  make it top and bottom SMT reflow (same amount of equipment required as
in number 1, but double the assembly time)

Here is where it gets tricky:
if significant TH is required you have three options:
4A.  reflow SMD, insert TH and wave solder or selective solder the TH
joints
4B.  Adhesive attach SMD to bottomside (including SOIC's and Quads), insert
TH and wave solder the whole thing (but your wave solder better be in
control, pad design, etc.)(This might allow use of CEM1 as opposed to FR4,
what a cost savings!!)

5.  topside reflow, TH, bottomside adhesive and wave solder the whole thing
(yes, these are all variations on type I, II, and III)

BUT, after all the above have been considered, the whole group needs to
discuss the best design, based on efficiency, quality, through-put,
component cost, etc.

So, I hope this sheds some light on your issue.  No hard numbers, but some
other things to consider.

Ed Holton
Hella Electronics

##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
To subscribe:   SUBSCRIBE TECHNET <your full name>
To unsubscribe:   SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional information.
For the technical support contact Dmitriy Sklyar at [log in to unmask] or 847-509-9700 ext.311
##############################################################


ATOM RSS1 RSS2