The message below highlighted a general problem I have run into on numerous designs. We constantly are trying to make the tradeoff regarding whether to punch, route, or V-Groove a board. Each of these techniques disrupt the board edge differently during PCB fab and also when the part is broken out of the panel. One question that always comes up is how close can we put parts & vias to the board edge. In high density designs we are always trying to push this to the limit, but I haven't found any general guidelines other that perhaps keep features at least one board thickness from the board edge. The concern is that either the FAB process will disrupt features and things such as plated through holes, and the assembly depanelization process will crack solderjoints as the board is flexed. Do any Industry design rules exist for how close the following features can be put to different types of board edges: I'd love to have the following design rule table: SMT parts TH Parts PTH Vias Routed Edge Punched edge V-Grooved Edge I have never seen IPC specs that delineate between type of edge, but then I'm not well read. Thanks, [log in to unmask] ______________________________ Reply Separator _________________________________ Subject: Re[2]: ASSY: Depanelization of SMT Board Assys Author: [log in to unmask] at Dell_UNIX Date: 4/25/96 8:42 PM There is nothing wrong with scoring and a proper panelization and scoring/routing can result in major PWB cost savings. You just have to determine in advance how the boards will be depanalized in production and design in the required criteria at the PWB level up front. If production plans on using equipment to depanel (punch, V-blade, etc.), then you need to know the component placement keep outs required around the edges of the board to clear the blades as well as component height restriction,etc. If boards are broken by hand, then DO NOT place any component close to the edge since solder joints will get damaged. As Gary has indicated below, the board dimension is not appropriate for breaking by hand and I don't even know if there is enough room on the board (1/2" X 10") to move components away from the edges for a punch equipment. [log in to unmask] Leo, There are equipment out there (Pizza cutter!!) designed to depanel V-grooved board. The V-blade used in those equipment will minimize bending stress inflicted on the PCB to minimum. (BTW,try to reduce the groove web to 0.015" if possible) Yuen ---------- From: TechNet-request To: Leo Reynolds Cc: Technet Subject: Re: ASSY: Depanelization of SMT Board Assys Date: Thursday, April 25, 1996 8:37AM Leo; We experienced similar problems with a board the same dimensions. When we investigated the excising methodology we were astonished to find that the customer was attempting to break the boards away, by hand! Boards with unsual aspect ratios, must be broken ustilizing a fixture that applys uniform pressure at the break point. This should reduce the forces that get transmitted into the board and the componet solder joints. Regards, Gary Ferrari Tech Circuits At 04:33 PM 4/24/96 -0500, Leo Reynolds wrote: >We are a contract manufacturer and are currently having trouble with a 1 >1/2" X 10" SMT board assy. It was designed in a panel of 6, scored for >eventual break apart and the original scoring left .018" of material in the >groove. When we started to break them apart we found we were cracking some >of the SMT caps. We've since instructed the PCB mfgr to go to a more >standard .012" per one engineers recmmendation, but don't have samples yet >to try. One of our process engineers attending Nepcon had several engineers >tell him that SMT boards shouldn't be scored at all. > >We are trying to determine the best course of action to recommend to our >customer. The customers design engineer felt we shouldn't have had a >problem with the original .018" groove so we may have to put together a very >god case for changing to routing or anyuthing else. > >Anyone have any recommendations? > >Thanks in advance for your consideration. > >Leo Reynolds > >