Ingemar,
Dont know if its too late to offer anything on this (been away a
couple of days), but if the parts a "daisy chain" cant you identify
roughly where its failed, and then cross section real careful about
that area only?
I dont know if it will work, but have you considered dye
penetration? I guess this needs a crack up on the surface to allow
penetration but you may be lucky?
Other possible ideas, but Im really starting to clutch at straws
are
Eddy Current methods, unsure of detection limit or noise
effects due to feature size?
Acoustic routes, which work on the idea that flaws in
joints etc generate microsound waves under thermal loading, again,
dont know if its suitable, but find an expert and ask them
If all else fails you could try asking somebody like the
Institute of None Destructive Testing, Never used them myself, but I
guess they do destrucive methods to, find them at
http://www.bindt.org/
Good luck, and let us know if you get anywhere, never know when
we need to find the needle in the haystack.
Roger
Motorola AIEG
______________________________ Reply Separator _________________________________
Subject: Re: [TN] HISTACK, IBM-type, japs
Author: "TechNet E-Mail Forum." <[log in to unmask]> at #email
Date: 16/09/99 22:15
Another excellent shop is Hewlett-Packard (name changed recently???) in
Santa Rosa. They know plating better than any one I know (including IBM).
-----Original Message-----
From: joyce <[log in to unmask]>
To: [log in to unmask] <[log in to unmask]>
Date: Thursday, September 16, 1999 4:03 PM
Subject: Re: [TN] HISTACK, IBM-type, japs
>wow! blame the test vehicle for the failure on ESS (based on the
>daisychain...I assume you are doing qualification and failed in
>T/H/B)...Blame the Doctors too on the design ...gee, really should just let
>you "swim or sink" on your own...but, I do like Big blue and IBM-guys...do
>not want them got beatten for no reason at all!
>based on the limited information you provided, here is what I would do if I
>were you:
>(1) pay IBM ($)for Failure analysis...If it is SIR type of test vehicle
>(with daisychain), I think you may need LSI tester to pin point where is
the
>leak or short. I believe they have a component lab and FA lab for doing
>outside work (at least used to be).
>(2) Pay IBM well ($$) for root cause analysis (you may have to open your
>material/process in order to get some real benefit of root cause. "Window
>dressing" will lead you more costly FA).
>(3) Pay IBM more (big time $$$$) for training some of your personnel to do
>proper FA. (be willing to spend sleepless nights to make sure you
>understand the steps and why!)...send someone with background in FA and
>design/process...Believe me, it is a good investment...You may have to sign
>some non-disclosure agreement.. watch what you sign...(you may sign off
your
>day dream and nightmare too)..
>(4) look like so far the only 2 people replied your e-mail are Dr.s! stop
>bashing us... you may need us someday (like as a drinking buddy).
> jk
>At 09:12 AM 9/16/99 +0200, you wrote:
>>Guten Wienerschnitzel, netters,
>>challenge for an IBM-guy if he reads (no, of course he doesn't, big Blue
>too big) , maybe my mate and competitor Bev can rise his eyebrow and move
>his finger to the keyboard. Topic: multilayer 3D "boards" with IBMs liftoff
>and PI dielectric. Reactive Ion etch (O2/He RIE) through mask for removing
>PI, followed by WSiN sputter for biting and low current electroplating for
>first step making the vias. Vias are finally O2 plasma, SF6 RIE'd. Now,
most
>seems maybe familiar to you, but, the problem is that the conductors are
>only 4um wide and the vias 1um in diameter. Test "board" consists of single
>to two layers with very short conductors, daysichained, let's say a hundred
>such strips. See it in front of you: gold, under is the via, via goes
>horizontally some um, goes up to next gold, .. up, down, up down, like
>chaining an ordinarey PWB, but one decade smaller. And, PI is the
>dielectric, approx 2um thick. Now, palomas, the Q:
>>when these "boards" are temcycled we get interrupts somewhere in the
>daisychain, seen as RF signal anomalies. How on earth do we analyse?
>Ordinary SEM of surface tells nothing of what's under the conductor.
>Finefocus Xray has unsufficient resolution for finding microcracks.
>Sonoscan, no. Cross-sectioning, no, weeks of job, little chanse of finding
>rupture. Reactive etching will heat and possibly take both Gold and PI. Wet
>etching, e.g. KaC? Takes Au, but what do you do with the W under Au? Agree,
>more like semiconductor techs.
>>
>>Oftenly when we have to do with the japs, we describe with hands, paper,
>mail, photos and words, and they reply 'yes', 'yes', which is oftenly of
>little help. So, Bev (I'm sure you have a recipe in your pocket) and like,
>what do I do with the drunken saylor?
>>Aussilek, give up, you can't be specialist on everything, you must have
>some limitation, please, tell that you are a human and not a Cray. (on that
>isle you may say crayfish, I don' t know).
>>
>>Cu?
>>
>>Ingemar Hernefjord
>>Ericsson Microwave Systems
>>
>>PS. With HISTACK, I mean that our test samples remind of this EU-thing,
but
>does not belong to EU at all.
>>
>>##############################################################
>>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.
>>If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
>>847-509-9700 ext.5365
>>##############################################################
>>
>
>##############################################################
>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.
>If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
>847-509-9700 ext.5365
>##############################################################
>
##############################################################
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.
If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
847-509-9700 ext.5365
##############################################################
##############################################################
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.
If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
847-509-9700 ext.5365
##############################################################
|
