Text item:
Isn't it true that the 30u" requirement was in effect long before Bellcore was a
gleam in Pa Bell's eye?
This may be one of those situations where we do it 'because that's the way we
have always done it'.
I don't think that soft gold is an option, but if hard gold could be plated at
10u" or so, the benefit to the manufacturing operation is the savings in
materials and labor for tape/untape, rework for bad tape, etc. and the
bottleneck costs. It would seem that there would be higher equipment costs for a
plating line, but automation would amortize the costs quickly.
This assumes that the life of the surface finish is acceptable, of course.
I have heard that thinner gold is currently being used on some SIMMs and DIMMs.
Does anyone have any feedback on this?
Steve
[log in to unmask]
Steve Joy asked:
>I have had requests for the reduction of the 30m" minimum for tab plating
>recently. Where did the 30m" minimum average rule come from?
A rhetorical question? Bellcore, methinks. Check out TR-NWT-001217 Issue 1,
Generic Requirments for Seperable Electrical Connectors Used in
Telecommunications Hardware. It is also probably in TR-NWT-000078.
Paragraph 5.2.2.1 and table 5 specify 50 uin Au/50 uin Ni as "preferred"
and 25 uin Au/50 uin Ni as "acceptable" for gold fingers, and for some
reason 30 uin Au/50uin Ni for other (non-PWB) contacts.
You can find where to order these from http://www.bellcore.com or call
1-800-521-CORE (908-699-5800 outside the USA).
I normally go into autodyslexia and start gibbering nonsense when trying to
read more than a few paragraphs of the mil stds, but some of this Bellcore
stuff actually makes sense. ;-)
Seriously, I respect Bellcore's technical competence in these matters
implicitly. This very nice specification will give you everything they
require both generically and in terms of performance/testing. I bet you
could even get some test data from them on this matter if you call someone
in Red Bank, NJ. (any Bellcore types lurking?)
>We have recently had to deal with some turnkey fabs having less than
>30m"; in the range of 10m". We did porosity tests and accelerated aging,
>etc. We could find no problems with the boards. (The nickel was around
>50m".)
Well, you might also do some mating and unmating cycles, measure normal
forces before and after testing, contact resistance changes, etc. And if
memory serves, most thickness requirements have a safety margin to allow
for porosity. If you did in fact have a very _non-porous_ gold finish,
you'd probably see good test results with thinner coatings. Trouble is,
specifying and checking porosity is a lot harder than thickness
measurement. Likewise, hardness and wear resistance of the metal.
I can readily accept that finishes with thinner noble metals could pass a
performance requirement, it just seems to me that any routine verification
testing of this is difficult.
>The question is: Does anyone have any reliable data for gold thickness
>vs. life for any products? ..like consumer; simms; other computer add
>ons?
There is plentiful data on that subject. You may need to hit a local
technical library, but this is one area which has had a lot of research.
AT&T/Bellcore probably are the biggest experts. You may also ask some of
your better connector suppliers. I know AMP has supplied me some
unpublished technical articles on such matters in the past. I respect their
opinion, too.
>I know that gold defects are not major hitters in PCB shops, but it is a
>bottle neck, correct?
For me (a user of circuit boards), it seems that whether we put 10 uin or
30 uin is not so much an issue of material cost, and so it must be
throughput on the gold tab lines (jump in here, fab houses) that motivates
fab houses to ask for relief.
Excuse me for quoting this old saw, but you know what the first three
things you need to do are when your computer acts up?
1) check the connectors
2) check the connectors
3) check the connectors
I've seen sooooo many old PWB assemblies with discolored or worn looking
gold tabs which have obviously oxidized. I've sprayed them with DeOxIt and
seen them miraculosuly work fine. I know it costs more, and maybe I'm just
obstinate, but I like gold connectors, like 'em thick too.
Forgive the boldness, but all the Interphase boards you buy will have 30
microinches of gold on the seperable contacts. I believe the marginal
additional cost will be repaid by the extra years the boards last, and the
expectation they won't be returned for repair. But I DO agree we get some
inquiries from vendors which imply we are asking for something unusual...
cheers,
Jerry Cupples
Interphase Corporation
Dallas, TX
http://www.iphase.com
Text item: External Message Header
The following mail header is for administrative use
and may be ignored unless there are problems.
***IF THERE ARE PROBLEMS SAVE THESE HEADERS***.
Cc: [log in to unmask]
Subject: Re: Where did 30m" for Tab plating come from?
From: [log in to unmask] (Jerry Cupples)
To: Steve Joy <[log in to unmask]>
Date: Mon, 25 Sep 1995 19:32:58 -0500
Content-Type: text/plain; charset="us-ascii"
Mime-Version: 1.0
Message-Id: <ac8cef4b02021004badb@[157.175.110.24]>
X-Sender: [log in to unmask]
Received: from [157.175.110.24] (mac1_comm0) by crow.iphase.com (4.1/SMI-4.1)
id AA15055; Mon, 25 Sep 95 19:27:48 CDT
Received: from crow.iphase.com by iphase.com (4.1/1.34)
id AA10104; Mon, 25 Sep 95 19:27:52 CDT
Received: from iphase.com by hermes.intel.com (5.65/10.0i); Mon, 25 Sep 95 17:29
:14 -0700
Received: from hermes.intel.com by relay.jf.intel.com with smtp
(Smail3.1.28.1 #2) id m0sxNtc-000twrC; Mon, 25 Sep 95 17:29 PDT
|