Colin, I haven't come across any such figures, and I doubt that they would have been very meaningful if I had, as there are too many variables in terms of material choice and choice of fab house to pin numbers to. I suggest you determine what flatness you need and the materials you want to use and then go talk to fab houses for advice on how possible your requirements are. My view is that, if you're using BGA's, I would prefer a thinner board to a thicker one from a manufacturing point of view. It takes a much longer time to get BGA contacts up to soldering temperature compared with most other components where the leads stick out the sides. A thicker board means an even longer dwell time at high temperature, which to me is a risky time for a board. Peak soldering temperatures, especially where BGA's are involved, are pretty close to the max case temperatures of a lot of components, and if you increase the length of time components cases are exposed to high temps just to heat up a thick board, you may induce internal damage. You can put the boards in support jigs during processing until the stiffeners are fitted. Pete Duncan Colin Weber <colin.weber@vari To: "TechNet E-Mail Forum." <[log in to unmask]>, DUNCAN aninc.com> Peter/Asst Prin Engr/ST Aero/ST Group@ST Domain cc: 07/20/01 05:43 AM Subject: Re: [TN] Bow & Twist Peter, We are designing the board to IPC specifications for scientific instrumentation. There is no special case for these designs. I don't have a problem with bow and twist as I haven't designed the boards yet. My question is more directed to how to avoid the issue if I want to design a large PCB. Apart from stiffeners and the like I am specifically wondering how thick I need to make a PCB if I am going to be placing fine pitch devices and/or PBGAs on them. If I were to go with 2.4mm I have EMI stackup and aspect ratio concerns, but feel comfortable the board will be more rigid. I could and would prefer to step down in thickness but do not have a feel for how the board would fair during handling and assembly. I am talking about 14 to 15 x 9 to 10 inch board sizes for processor designs. An obvious solution is to split the PCB up into two boards, but I do not have that luxury. I was just curios if there are any guidelines or figures fro bow & twist with respect to size versus thickness? At 02:44 PM 19/07/2001 +0800, you wrote: >Much of the bow and twist these days is down to the choice of material used >for the substrate coupled with 'balance' in the board design and ultimately >on the capability of the fabrication process. > >If you have heavy planes in the board, make sure they're distributed evenly >throughout the layers, otherwise any good PCB Fab worth his salt will keep >any bow and twist within the bounds of the usual specs. It used ot be that >the thicker the board was, the more rigid it was and the less inclined it >was to warping. These days, more than 20 mils bow or twist on a 63 mil >thick board measuring 9 x 6 inches is unusual. > >Are you designing boards to be made to MIL-PRF-55110, to an IPC spec or to >something more commercial? > >Pete Duncan > > > > > <colin.weber@VARI To: [log in to unmask] > > 07/19/01 12:18 PM > >Folks, > >Is there a method of determining the maximum permissible size of two board >types? >a) A SMT PCB containing Fine pitch QFP devices >b) A SMT PCB containing PBGA 388pin 1.27mm pitch & Fine pitch QFPs > >with respect to Bow and Twist. > >I am interested in how the thickness of the board relates to maximum bow >and twist. >I'd like to determine this before I design the layout, rather than find out >it is wrong >after performing the tests of IPC-6012A, IPC-TM-650 2.4.22. > >Regards, > >Colin Weber Regards, Colin Weber --------------------------------------------------------------------------------- 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 ---------------------------------------------------------------------------------