Peter, FYI, to the best of my knowledge all major suppliers of underfills for Flip Chips - flux underfill (no-flow) and capillary underfill - supply both reworkable and non-reworkable. Specifically, I know that Kester, and Emerson & Cumings do. I am pretty sure that Loctite/Hysol does as well. Each of them can supply materials having particle sizes more suitable for BGA's, but I have only worked with the Flip Chips Steven Creswick - Gentex Corp -----Original Message----- From: <Peter George Duncan> [mailto:[log in to unmask]] Sent: Wednesday, August 22, 2001 12:17 AM To: [log in to unmask] Subject: Re: [TN] thermal interface material [This e-mail is confidential and may also be privileged. If you are not the intended recipient, please delete it and notify us immediately; you should not copy or use it for any purpose, nor disclose its contents to any other person. Thank you.] Geez! My third contribution to this thread. Genny, underfill is generally a special epoxy, formulated for use mostly with BGA's. It's prime use is to cushion BGA solder joints to help dissipate the stresses that build up in them, thus increasing their mechanical reliability. These underfills are not (typically) very thermally conductive. There may, however, be other types of underfill that are thermally conductive to replace compounds, etc. I don't know for sure. Underfills (BGA type at least) come in two basic forms - 'capillary action' and 'no-flow'. No-flows are applied to the component before placement and act as a fluxing agent as well as an underfill. Their disadvantage is that they're not generally reworkable (your device is permanantly stuck), and your choice of flux becomes very circumscribed. The upside is they do have slightly better thermal properties than the capillary type. Capillary types are applied after component placement, usually with a controlled pressure fluid dispensing system. In volume manufacture the $140k price tag for a fully automated one might be cost effective, but for batch or low volume stuff, cheaper dispensers are available. I got a refurbished manual one for a touch over $300 and it works quite well if you haven't been on the sauce the previous night and thus have a steady hand. Capillary action underfills don't limit your flux options, are quite easy to apply and are reworkable, softening at just below eutectic soldering temperature and lifting off with the component. It's supplied in syringes that fit straight onto standard dispenser adapters. You stick a needle on the business end of the syringe and run a bead round one or two sides of the component. The bead is drawn under the component to fill the gap there. More than one pass is required to fill large area components (over 25mm on a side). This process would be no good for your components with such big air gaps though, as capillary action doesn't occur. And Underfills need to be stored at -40 deg C, so you need a serious freezer. Cheers again, Pete Duncan Genny Gibbard <Genny.Gibbard@W To: [log in to unmask] AVECOM.CA> cc: (bcc: DUNCAN Peter/Asst Prin Engr/ST Aero/ST Group) Sent by: TechNet Subject: Re: [TN] thermal interface material <[log in to unmask] > 08/22/01 05:00 AM Please respond to "TechNet E-Mail Forum." Thanks for the info received so far. Some more info and more questions. What is underfill, and how or when do you apply it, and where? To the IC, or to the board, before or after installation? The chip runs at varying temperatures depending on the freq of the reference. A low freq crystal that is internally multiplied up causes the chip to run much cooler than providing a reference that does not need to be multiplied. However the second causes more accurate operation. We have applications that use it in both ways. With the crystal, you can touch it with your finger without pain. With the higher reference, you cannot. The manufacturer's data sheet provides no comments regarding the heat slug at all - the dimensions are shown on the mech drawing. There are no recommendations on whether heat transfer needs to be improved or how. There would be no air flow over the device as it is shielded. I wanted to transfer heat to the ground plane on the opposite side of the board by having an area of copper under the device well connected to the ground plane with vias, but I need to make a thermal connection to the copper from the device slug. Thanks for your time. -----Original Message----- From: Graham Collins [mailto:[log in to unmask]] Sent: August 21, 2001 12:14 PM To: [log in to unmask] Subject: Re: [TN] thermal interface material Genny First off, as an assembler I'd worry about exactly what you conclude - will the pad make soldering a problem? I'd worry about coplanarity, among other things. If this were my problem to worry about I'd look at it this way: 1) is it really a problem? The designer (you?) should know the power use of the chip, as well as the expected operating environment. Is it going to run hot? 2) if it is a problem, can some sort of underfill be used? I'm not in favor of messing with an unsoldered part... good luck! regards Graham Collins Process Engineer, Litton Systems Canada, Atlantic Facility (902) 873-2000 ext 6215 >>> [log in to unmask] 08/21/01 02:54PM >>> We use an IC that sits min 0.25mm, max 0.5mm off the PCB according to the datasheet. There is a heat slug embedded in the bottom. Fine pitch 120 pin, PQFP The data sheet says nothing about requirements for heat sinking, but when there's a slug you kind of feel you should help it out, right? For reliability, MTBF,... Due to the fine pitch, placement is critical. We are looking for a thermal pad of some sort to place under the device, as that thick a layer of heatsink compound under it is probably not a good idea, causing more problems than it solves. Possible running of the compound, contamination of the lands, causing the part to move while in the oven... We get the boards reflowed out of house. The co. we use suggested a sil pad, and said they could possibly pick and place the pad. My questions: How thick a pad should we look for, 0.25mm or thicker? What if the gap is the max for some IC's - will the pad be of any use when not making contact? If we spec thicker, will the pad keep pins from soldering when the IC high centers on the pad? Anyone have other ideas? Genny Gibbard (mailto:[log in to unmask]) Product Transition and Support Wavecom Electronics Inc. ---------------------------------------------------------------------------- ----- 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 ---------------------------------------------------------------------------- ----- ---------------------------------------------------------------------------- ----- 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 ---------------------------------------------------------------------------------