Hi, Technet
Here is a problem I was asked to solve today. It relates to this string.
I was shown some small daughterboards that are .031" thick, with components only on the topside. On each end are 4 posts that have flat gullwing leads roll-staked and soldered to them. The gullwings in the corners go out at a 45 degree angle, the inside two leads on each end point straight out.
The daughterboard is soldered to a massive, .250" thick 24 layer pwb motherboard.
No comments about my mother please, Dewey!
The assemblies were shown to me after a few months in the field, and the SJ's are fractured, some where they are roll-staked to the post on the bottom of the daughterboard, some on the opposite end where they were soldered to the motherboard pads. There does not appear to be any pattern in the cracks. I know of no physical stresses that the daughterboard would see. I do not know what the service environment is, but I don't believe it sees a lot of temperature cycling. Probably an office environment.
The solder on the leads appeared very irregular, with small peaks or mounds of solder here and there and large areas in between where the solder did not wet on the tops of the leads. Some of the leads displayed good wetting. Again, no pattern.
So I pulled up the print on the daughterboard, and found the following information. The leads are soldered to the post in addition to being roll-staked using Sn10%Pb90% wire solder. However, the leads are pre-finished with Sn90%-94%Pb6-10%.
And the daughterboard is soldered to the motherboard pads with standard 63/37 solder. The motherboard pads have an immersion silver finish.
I suspect the SJ's fractured because the Pb distribution throughout the SJ will vary widely when the entire SJ on the lead gets some 63/37 solder from the paste, and some 90tin/10lead from the lead finish, and some 10tin/90lead from the solder on the post. Not to mention whatever effect the small amount of immersion silver from the motherboard pad has in this distribution.
I am going to recommend changing the post solder to the same as the lead finish, Sn90Pb10. I have to continue to use 63/37 to solder the subassy. I suspect the Sn10Pb90 was an error on the print. Then I would qual test all over again.
Does that make sense? I do not have a lot of experience with reliability issues where there are several different alloys like in this situation. Anybody else have any advice?
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Ingemar Hernefjord
Sent: Thursday, November 17, 2005 3:47 PM
To: [log in to unmask]
Subject: [TN] Reliability prediction of LF soldering LAST CALL!
Still awaiting for an answer from the experts: can we neglect the varying distribution of lead when soldering BGAs and other parts? If not, how do we handle the situation on the floor. Production people want know what to do when they do sample metallurgical investigations and see Lead distributed many a way. See example sent to Steve, one and the same BGA. Bright spots is Lead.
Inge
-----Ursprungligt meddelande-----
Från: TechNet [mailto:[log in to unmask]] För Ingemar Hernefjord
Skickat: den 17 november 2005 03:07
Till: [log in to unmask]
Ämne: [TN] SV: [TN] SV: [TN] [LF] Reliability prediction of LF soldering
Bingo jk,
That's what I mean with reality. Scientist being, as you are with your background, you have to admit, however, that Arrhenius law is adaptable on hot solder joint as well as cold, but it's a tough job to 'estimate'
all the thousand individuals because of the different life they live.
I would like to pay attention to my own central question which was indeed a super'estimation': can we omit the various lead distribution in BGA solder joints (and others as well)? I did not contribute with a solution myself, because I wanted to wake up the dragons to see what treasures they were lying on.
Our company has been working for years with the announced topic, and in cooperation with other companies. All do the same. If you do your 'estimations' without too many of my reality factors, your MTBF calculations give an interesting answer. As the result belongs to our company, I can't give you the exact figure, but I can go so far as to say this:
MTBF for LF BGA solder joints is : MTBF(LF)=MTBF(Sn/Pb) times a multiplicand that is less than 1.0. All is confirmed by numerous tests.
Upon this number, you may build your up-and-down pyramid by putting in all the real life factors that I mentioned. We are working on that, but it seems to be a long working day before we have nice and reliable (!) model.
Inge
-----Ursprungligt meddelande-----
Från: Joyce Koo [mailto:[log in to unmask]]
Skickat: den 16 november 2005 23:34
Till: TechNet E-Mail Forum; Ingemar Hernefjord
Ämne: RE: [TN] SV: [TN] [LF] Reliability prediction of LF soldering
"Some GaAs chips are so hot
that you burn your fingers when you touch them,"...
Laptop by the name of it should be able to put on your lap... how many of lap top is cool enough to sit on your lap? By the way, I couldn't understand why the laptop guys always ad that "ultra-light" "compact", meanwhile you need one of "cooling pad" to operate these ultra stuff...
http://www.maplin.co.uk/Module.aspx?ModuleNo=45656&&source=14&doy=search
As for reliability, heat dissipation sure get rid of all the MTBF prediction (how many of prediction based on the hot location and cool location in the same unit?)..don't get me started..
jk
-----Original Message-----
From: TechNet [mailto:[log in to unmask]]On Behalf Of Ingemar Hernefjord
Sent: Wednesday, November 16, 2005 4:21 PM
To: [log in to unmask]
Subject: [TN] SV: [TN] [LF] Reliability prediction of LF soldering
Exactly. A processor chip can go really hot, while other chips are biased in a way that they have nearly no heat dissipation at all. The boards general temperature is another factor, starting a board at 30,000 feet or at Saudi ground is two extremes, that affect the calculations.
And the placement on the boards, we have boards mounted on a cooler waffle, big differences in BGA temperatures. Some GaAs chips are so hot that you burn your fingers when you touch them, others are working once in a month etc. Some balls carry a lot of current, others just pA.
Center of superBGA differs from perimetrically placements. Warpage from the boards have an impact. Vibration and other mechanical injections have an impact. And so on. I'm not doing MTBF predictions myself, but I realise that we must have to do with very complex models to be able to predict the real life for a given number of components on a specific board. I still believe in making tempcycling, switched life test and life tests to failure and other practics which I call hand-on. I'm probably rigid, conservative and too old to learn new things...
Inge
-----Ursprungligt meddelande-----
Från: TechNet [mailto:[log in to unmask]] För Ahne Oosterhof
Skickat: den 16 november 2005 21:48
Till: [log in to unmask]
Ämne: Re: [TN] [LF] Reliability prediction of LF soldering
>From my "engineering" experiences:
I used the "scientific" method to calculate an MTBF before finishing a paper design. Then changed the design if the number ended up too far out of the desired ballpark.
The changes could be made (I thought) because for most components the demonstrated MTBF depended a lot on how they are being used. A 1W transistor dissipating 0.1W lives longer than one running at 1.0W. (Kind of like comparing one solder joint on a 0402 resistor in a desktop modem (super benign environment) to one on the corner of a BGA located close to the shear edge of a board that is used in a portable computer by a forester in the Pacific NW in the middle of winter (throw it, shake it, cool it, etc.).
After building engineering samples of the design I required life cylce tests to find a 'demonstrated MTBF', a number that would go in the spec document for customers to see.
This discussion is really "fun".
Ahne.
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Ingemar Hernefjord
(KC/EMW)
Sent: Wednesday, November 16, 2005 10:25
To: [log in to unmask]
Subject: Re: [TN] [LF] Reliability prediction of LF soldering
Agree, wasn't an answer on my question. We have complicated situation with mixed technology, and we have few persons with a deep understanding in the mystery of metallurgy and kinetics and so on. Perhaps someone else can give a tip. The scientific reports are sometimes too academic, and my job is to 'strip' the report and make it understandable for those who work on the 'floor' (old fashioned expression, guess we are all standing on floor, except for the high bosses, who are in the blue). So, again, can you omit the variations of lead distribution or not? It's very important that I give the soldering responible people a correct information.
Inge
-----Original Message-----
From: TechNet [mailto:[log in to unmask]]On Behalf Of John Burke
Sent: den 16 november 2005 19:09
To: [log in to unmask]
Subject: Re: [TN] [LF] Reliability prediction of LF soldering
Sharp words for Technet..........................8-)
John
------------------------------------
Avanex
John Burke
Senior Manager RoHS Compliance
[log in to unmask]
40919 Encyclopedia Circle
Fremont
CA 94538
tel: 510 897 4250
fax: 510 979 0189
mobile: 510 676 6312
------------------------------------
-----Original Message-----
From: TechNet [mailto:[log in to unmask]]On Behalf Of Whittaker, Dewey
(AZ75)
Sent: Wednesday, November 16, 2005 10:04 AM
To: [log in to unmask]
Subject: Re: [TN] [LF] Reliability prediction of LF soldering
At least if he did that he could take it for granite that he was working with cutting edge technology.
Dewey
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Werner Engelmaier
Sent: Wednesday, November 16, 2005 10:59 AM
To: [log in to unmask]
Subject: Re: [TN] [LF] Reliability prediction of LF soldering
Hi Ingemar,
Well, tell us--what choice do you have? Are you a 'scientist' or an 'engineer'? Shall we stop making product? Go back to the stone age--assuming of course that the quartz is uniformly distributed in the blade of your stone axe?
Since when do we live in a perfect world?
Werner
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or
(re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the
posts: send e-mail to [log in to unmask]: SET Technet Digest Search the archives of previous posts at: http://listserv.ipc.org/archives Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100
ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or
(re-start) delivery of Technet send e-mail to
[log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the posts: send e-mail to
[log in to unmask]: SET Technet Digest
Search the archives of previous posts at:
http://listserv.ipc.org/archives
Please visit IPC web site
http://www.ipc.org/contentpage.asp?Pageid=4.3.16
for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or
(re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest Search the archives of previous posts at: http://listserv.ipc.org/archives Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100
ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or
(re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest Search the archives of previous posts at: http://listserv.ipc.org/archives Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100
ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or (re-start) delivery of Technet send e-mail to
[log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the posts: send e-mail to
[log in to unmask]: SET Technet Digest
Search the archives of previous posts at:
http://listserv.ipc.org/archives
Please visit IPC web site
http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100
ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or (re-start) delivery of Technet send e-mail to
[log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the posts: send e-mail to
[log in to unmask]: SET Technet Digest
Search the archives of previous posts at:
http://listserv.ipc.org/archives
Please visit IPC web site
http://www.ipc.org/contentpage.asp?Pageid=4.3.16
for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or (re-start) delivery of Technet send e-mail to
[log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the posts: send e-mail to
[log in to unmask]: SET Technet Digest
Search the archives of previous posts at:
http://listserv.ipc.org/archives
Please visit IPC web site
http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100
ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest Search the archives of previous posts at: http://listserv.ipc.org/archives Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-----------------------------------------------------
|
