Reading time ( words)
By Todd King, E.I. Microcircuits Inc.
Design for manufacturability (DfM) is extremely important to today's electronics production. DfM can help EMS providers deliver the best product possible at the lowest production cost.
Of course, this service doesn't fit every contract customer. However, the majority of customers have a newly designed product that has not yet been tested for automated production, requiring DfM analysis. Anything can be manufactured in a design layout using modern programs. Anything can also be made to work as a prototype if engineering spends enough time in hand-designing it. That does not mean it can actually be manufactured. If the design is manufacturable, aspects of the layout, components, or overall design may add cost to production that DfM can eliminate.
Even companies designing leading-edge electronics based on innovative designs want to keep production costs down. Some begin planning for manufacturability with only a circuit design or schematic; not at the actual layout stage. Maybe they've got a custom case they want the final product to fit into. In this situation, the manufacturing provider will lay out a circuit board, take their circuit design and their bill of materials and make it as efficient as possible from a manufacturing standpoint. This is the easiest DfM to work with. However, the other scenario representing probably the majority of DfM work is customers coming to manufacturers with the full board layout. They've either done it themselves or they've outsourced the CAD layout. They then took the design and hand-built some prototypes. They have a design, and may already have a case they want it to fit into. This takes some work to make sure the product is manufacturable. As noted above, virtually anything can be designed and built by hand.
The first step in turning a handmade product into something producible using automated techniques is to verify that electrical or agency design rules are not violated. Next, examine how the parts and components are situated, top and bottom side, as most assemblies are double-sided or multilayer. Is it possible to convert most of the design into an automated surface mount format? That will always bring down cost and speed up production.
The main goal is to simplify a design's layout and try to build it in the least expensive way possible. Sometimes design suggestions can be incorporated. Others, because of the specific needs of the product or client, require more work to solve special production issues.
Figure 2. Another example of a complex board with X/Y and Z constraints on solderability.
The next step is to build prototypes. These are run on automated lines to see where any hang-ups may occur and how to correct or work around them cost-effectively. Offer the manufacturer's perspective on how to improve the product for manufacturability. Usually, prototyping will go easily, and production can start. That's the ideal scenario, but it is not the only one possible.
Adding Through Hole Within DfMWhen critical components on an assembly can't be SMDs, designers must verify that they can go through the wave solder process without adverse effects. This is particularly critical with high-density double-side PCB layouts. Special pallets and/or masking can be used to achieve success. Many times they can't, and require manual soldering, which slows production throughput, affects quality, and adds substantial cost to the end product.
Automated selective soldering is one of the processes now used to create these difficult prototypes products. Used in-line at E.I.'s facility for normal through hole processing, selective soldering reduced manual labor by 60% and equally increased product quality.
If the customer is inexperienced in laying out designs, the likelihood of hand soldering in the design goes up. Other prototyping issues, such as pad sizes that don't match the components, come up. The goal with DfM is to help re-design their board so it will lend itself to automation. The more a product lends itself to automation, wave soldering or surface mount, the price of manufacturing is always lower. Manual labor is probably the biggest cost contributor.
Even with surface mount devices, alternative components may be recommended to make the design more robust. Some can't withstand reflow temperatures needed on the board. Some alternatives are cost-driven, but sometimes a design needs what only that particular device can provide, and there is no other choice. Often it is a tradeoff.
Automated selective soldering can address many of these tradeoffs and speed up some of the things that can't be surface mount and also can't be wave soldered. In the E.I. facility (see Acknowledgements), selective soldering systems are fully automated with SMEMA line interface. The process conveyor is set parallel to the solder nozzle. An X/Y/Z axis motion articulates the mini-solder wave nozzle beneath the circuit boards. Using this technology, the system effectively performs precise point to point, dip, drag, and mini-wave soldering. The flux system is also mounted to the X/Y/Z table and can be configured with spray, drop jet, or ultrasonic nozzles. This fluxer enables easy change-over for different PCBs, components, solders, nozzles, and odd-form assemblies. Nitrogen assists the thermal capability, improving the surface tension of the solder and controlling wave stability.
The solder pots are small (35-75 lbs typical) and wave solder nozzles range from 2- to 8-mm diameter in round configurations and from 100 to150 mm in rectangular nozzle configurations. Solder temperature and dwell time is typically the most critical area for selective solder processing. Most operators try to run the solder pot at 490°F because of the FR-4 Tg of 140, while taking into consideration panel and board size.
Programming selective soldering systems can be recipe-based or accomplished through the basic program system, using scanned images, imported Gerber data or even a camera-based jog-to-teach method, making prototyping simple. Solder points are selected and nozzle configuration set to generate a selective soldering program in minutes.
This flexibility addresses issues such as components that need to be soldered at an angle due to coplanarity issues with the board. Also, some tight and deep leads are difficult to solder by hand without damaging surrounding components. With RF components, for example, even the slightest tail on a lead can become an unwanted antenna. Selective soldering offers a lot of alternatives that help difficult prototype designs fit into automated manufacturing.
ConclusionWhen working with difficult or unusual designs, opt for selective soldering if the PCB cannot go 100% surface mount, and if specific components cannot be wave soldered. Try any adjustments to the mechanisms of the selective solder system to accommodate that specific assembly before resorting to expensive hand soldering. Even the most difficult layouts can be handled by automated selective soldering systems. This takes the labor aspect out of the product design, lowers cost, and greatly increases throughput.
ACKNOWLEDGEMENTS:Selective soldering equipment information was provided by Reed Gaither, CEO, RPS Automation LLC, 3808 N. Sullivan Road, Building 14-J Spokane Valley, WA 99216; (509) 891-1680; firstname.lastname@example.org; www.rpsautomation.com.