Rick,
I did not mention the no-flow (or flux underfills) or a reason. We use them
extensively on our LTCC (low temp co-fired ceramic) products with great
efficiency - stencil solder paste, place SMD's, dispense flux underfill,
place flip chip, place an RF shield over the entire structure, and reflow -
post cure if necessary. Large die, up to 5.8mm sq, have passed extensive
TC's (-40+125°C). Acoustic imaging has proven it a good process.
Cannot say the same for laminates (PWB's) as a substrate. Usually, the
corresponding solder pads on the PWB are solder mask defined. If the
solderable area is not limited in some way, the solder bump will flow out
across the conductor, causing the FC to drop closer to the PWB - net result,
reducing the reliability (reliability directly related to height of standoff).
It seems that the openings in the solder mask are an excellent site for
inducing voids in the underfill. Results in gross electrical non-contacts.
Sometimes the device has rotated up to 90° from its original position due to
gross void formation (and likely help from convection oven), yet essentially
the same part made from LTCC can run 'forever' without failure. I've not
found the root cause yet. Tested many things. Maybe another reader has had
similar experience and can relate a solution on their end.
Since you likely did not have acoustic imaging capability in-house, I
figured it best not to steer you down what seemed to be an easy path, that
might be prone to latent defects.
I know that Kester and Emerson & Cumings make flux underfills, but I do not
remember that (Dexter) Hysol has them - I know they have traditional
post-reflow underfills, as do a whole slug of other sources.
I, personally, have underfilled a few of the 5.8mm devices by hand using the
normal post-reflow type underfills, with very good success, but I have made
a few boo boo's too. I have the ability to cut glass slides to the same
dimensions as the device. Combining this with placing a few wire bonds on
the PWB, creates a realisitic dummy flip chip that you can gauge your manual
(or automatic) underfill process with.
Do you know what environmental requirements you need to meet??
Hope this helps too.
Steve Creswick
CTS RF Integrated Modules
At 09:21 AM 2/16/01 -0800, you wrote:
>Steven,
>
>Lots of good info and questions, some of which I hadn't considered or don't
>have answers for yet. The device we're placing is around 3.5mm square with
>16 eutectic bumps that are approximated .15mm. Pads on the PCB are
>approximately the same size and we do have to place 2 smd chip resistors on
>the same substrate. We're only doing a 10 piece test run with only 30-40
>pieces to follow that so volume initially is low. We have an EFD dispenser
>that I'm hoping to use for underfill and intend to attempt placement with
>an SRT Summit BGA rework system. From the limited amount of research I've
>done to this point it seemed to me that some of the no flow-fluxing
>underfill systems might be a viable approach although I haven't done alot of
>review of that process yet so may be overlooking something.
>
>Thanks for the feedback.
>
>Rick Thompson
>
>-----Original Message-----
>From: TechNet [mailto:[log in to unmask]]On Behalf Of Creswick
>Sent: Thursday, February 15, 2001 5:29 PM
>To: [log in to unmask]
>Subject: Re: [TN] Flip-chip underfill?
>
>
>Rick,
>
>Could you provide a little more info, such as how large is device and the
>bump dia & pitch? Number I/O's?
>
>Is the flip chip bump composed of Sn63, or a IBM C4 structure?
>
>How large are the corresponding solderable pads on the PWB?
>
>How close are the nearest adjacent components?
>
>How many do you have to put together?
>
>Are you planning to co-reflow the FC with SMD's, or secondary reflow?
>
>What dispensing equipment do you have available to use?
>
>How do you plan to align and place the flip chips?
>
>
>Indium (and others) make no clean flip chip fluxes. One can either dip the
>solder bumps approx 1/3 to 1/2 way of their height into a paste-like flux,
>or jet or dispense a wee bit of the same flux in a very low viscosity
>version. I believe that Indium's part number designation is FC-NC-LT-A, B,
>C, D (where FC means Flip Chip, NC = no clean, LT=low temp (Sn63 like) and A
>thru D denote viscosity (A=like alcohol, D= like a paste flux).
>
>Big, oversize (or otherwise 'fat') fillets are visually unappealing, and (in
>my testing) less reliable than small, but uniform fillets.
>
>Most underfills that are applied after FC attachment will require the
>component to be heated to 70-90°C to enhance the capillary flow of the
>material. Post-dispense cures range from 125°C to 150°C, depending on
>supplier, etc. etc. Times range from minutes to about an hour.
>
>Rework after underfill cure is a pain, so essentially be assured that all is
>well before sealing it up tight with underfill.
>
>Scads of people to supply underfills - Dexter Hysol (now part of Loctite),
>Ablestik Labs, Kester, Alpha, Emerson & Cuming (Ablestik), Loctite, .....
>
>Placing the chip by hand may not be the smartest thing to do, but with a
>reasonable setup, and small die (say less than 0.100" square) you should be
>able to dispense by hand with something similar to an EFD dispenser.
>Underfill from two adjacent sides only - let the capillary flow bring it
>under the chip and to the other sides. For larger chips, one might be
>better to use a dispenser that you can program a path. Whatever you do, do
>not allow the underfill wavefront to deplete the available reservior (excess
>material not under the chip). If the reservior is depleted, you can pull
>air under the chip - oops!
>
>Do you plan to check for underfill voids? X-ray generally will NOT work!
>Got an acoustic microscope??
>
>What environmental requirements do you have to meet?
>
>Parting shots - If is a low rel, cheap & dirty PWB, you may get away with
>it. If it is a higher quality project, where the customer can come back and
>bite you - tread carefully. It may appear easier than it actually is
>
>Have fun!
>
>Steven Creswick
>CTS RF Integrated Modules
>
>
>At 10:57 AM 2/15/01 -0800, you wrote:
>>We've been asked to do a small proto run of flip-chip assemblies. I'm
>>looking for some insight on how best to underfill these parts. Is manual
>>equipment sufficient to do this? Any input as to materials & methods would
>>be appreciated.
>>
>>Thanks in advance,
>>
>>
>>Rick Thompson
>>Ventura Electronics Assembly
>>2655 Park Center Dr.
>>Simi Valley, CA 93065
>>
>>+1 (805) 584-9858 x-304 voice
>>+1 (805) 584-1529 fax
>>[log in to unmask]
>>
>>---------------------------------------------------------------------------
>------
>>Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
>>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 delivery of Technet send the following message: SET
>Technet NOMAIL
>>Search previous postings at: www.ipc.org > On-Line Resources & Databases >
>E-mail Archives
>>Please visit IPC web site (http://www.ipc.org/html/forum.htm) for
>additional
>>information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700
>ext.5315
>>---------------------------------------------------------------------------
>------
>>
>>
>
>----------------------------------------------------------------------------
>-----
>Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
>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 delivery of Technet send the following message: SET
>Technet NOMAIL
>Search previous postings at: www.ipc.org > On-Line Resources & Databases >
>E-mail Archives
>Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
>information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700
>ext.5315
>----------------------------------------------------------------------------
>-----
>
>
>
---------------------------------------------------------------------------------
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
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 delivery of Technet send the following message: SET Technet NOMAIL
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------
|
