6/12/98 8:24 AM RE>Re: [TN] Land Pattern Formulas 4 IPC-SM-782? I do agree that IPC-SM-782A is a very workable system. I don't agree that the formulas are missing. They are there. What you need to do is familiarize yourself with the assumptions on "3.3 Dimensioning Systems" thru "4.3 Design for Producibility". You should understand these assumptions so you can resolve the handling of tolerances on values from manufacturing (auto-place), fabrication and component supplier. Determine your Company's requirements for values for Jtmin (minimum toe fillet size), Jhmin (minimum heel fillet), Jsmin (minimum side fillet), board fabrication tolerance (peak-to-peak), assembly positioning tolerance (peak-to-peak). Some of these values may be '0'. But you need them to maintain a workable formula that is relational for all new records of new or adjusted values. These values should be determined by talking to manufacturing, etc. Find a solid worst case number for the fillet values (RSS will take care of tolerance unilaterally toward MMC). We also used a 'rounding' factor such that our CAD pad library didn't have to deal with sub-mil values or values that were 1-2 mils in size from each other. This allows a standard set of CAD data across a large design group. Manufacturing will appreciate consistancy of CAD library data as an aide for creating paste screens, determining solder volumes, controlling temp profiles,etc. Then we captured this all on a spread sheet ( modified from IPC's ) and use it as the mechanism for creating footprint geometry automatically! It uses slightly different assumptions than the ones in IPC. The point is WE can change values in the spread sheet "on the fly", if we find a yield problem. Or, if another design group has looser design contraints, they can take our spread sheet and 'tweek' the Jtmin and jhmin values...or the manufacturing tolerances. Here's the skinny of our formula based on IPC-SM-782A. We don't use all the same nomenclature, so you'll have to translate. L=length of component, W=width of component, S=component gap, T=term, X=land length, Y=land width, G=land pattern gap, Z=overall land pattern length. See attached "seds.doc" for graphic. See "seds.xls" for adjusted IPC Excel formulas. We elected to compromise on the optimum resultant land pattern, for a more uniform CAD design base ACROSS our Design Dept's. The feeling was that Manufacturing can realistically adjust their processes IF they receive uniform land geometry. This was accomplished by a 'rounding' factor, so that no matter who designs a land pattern using the formula, the pattern will satisfy the fillet and tolerance values. Below is a review of assumptions: (You can use the spreadsheet to re-enter your own) Surface-Mount Footprint, Calculation 08/28/9 Let: CL = Lmax - Lmin CW = Wmax - Wmin JTmin = 0.020 in. Toe fillet required JHmin = 0 Heel fillet required JSmin = 0 Side fillet required F = 0.004 in. Board fabrication tolerance, peak-to-peak P = 0.004 in. Assembly positioning tolerance, peak-to-peak L, T, W defined per figure 1. Calculate initial Z': Z' = Lmin + 2JTmin + (CL2 + F2 + P2)1/2 Calculate G: G = Lmin - 2Tmax - 2JHmin Round G DOWN to next lower multiple of 0.005 inch. If G is less than 0.020 inch, set G = 0.020 inch. Calculate X: X = (Z' - G)/2 Round X UP to next higher multiple of 0.005 inch. Calculate final Z: Z = G + 2X Calculate Y: Y = Wmin + 2JSmin +(CW2 + F2 + P2)1/2 Round Y UP to next higher multiple of 0.005 inch. ------------------------------ Date: 6/11/98 5:08 PM To: Patten, James From: R. Gregory, Stephen In a message dated 6/11/98 3:58:52 PM Pacific Daylight Time, [log in to unmask] writes: > IPC support, > > I would like the exact formula used for generating the chip resistor > land pattern values for IPC-SM-782 manual. If the formula does not reside in > the SM-782 spec, please E-mail it, Fax it, or explain how I could get ahold > of it. > > FYI - From looking through IPC-SM-782 manual (published in 1996) I don't > find a straight forward formula for a chip resistor. In your IPC-D-275 > publication (dated September 1991) I find a formula for chip resistors > within section 5.3.4.3 that's very clear. Upon using this formula, when I > plug in values from SM-782 subsection 8.1, heading 4.0, for the 1206 chip > resistor, results doesn't match the land pattern dimensions supplied on the > following page. > > If you have any questions.. please E-mail, or call me. > > - Thanks for your help.... Al Arellano > > E-mail = [log in to unmask] > Phone = (303) 581 2407 > FAX = (303) 581 9973 > Hello there Al, There really isn't any "formula" for determining your pad dimensions...I think you're reading too much into it and confusing yourself. What 4.0 of subsection 8.1 actually is, are just the component dimensions...that's all. It just illustrates what the part basically looks like and gives you a min and a max dimension for each axis of the part. For instance with axis "L" for a 1206, that dimension needs to be within 3.05mm to 3.35mm in order to use RLP103A (RLP means recommended land pattern) for a 1206. Just look at 4.0 as a part of the document to determine for sure what the component should be identified as. Trust me, the 782 is a very good reference for pad geometry and footprints...if only EVERYBODY followed the 782 I might not have the gray hair that I do now...(GRIN) -Steve Gregory- ################################################################ TechNet E-Mail Forum 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 To unsubscribe: SIGNOFF TechNet ################################################################ Please visit IPC web site (http://jefry.ipc.org/forum.htm) for additional information. For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700 ext.312 ################################################################