South Africa has many brilliant electronic design engineers, whose innovative designs have succeeded both at home and internationally. But even the best among them are not necessarily familiar with the complex discipline of manufacturing, and thus often suffer the unintended consequences of ending up with a product not ideally suited to an efficient manufacturing process.
Tellumat has amassed a vast body of knowledge in the field of ‘design for manufacturability’ (DFM), offering practical guidelines to help designers take cognisance of the manufacturability of their design, thus avoiding costly mistakes. “DFM principles are a set of good, practical design practices. These guidelines aim at easy, efficient and cost-effective manufacturing, as well as impeccable first pass yields – without detracting from or altering the functionality of a design,” says Murison Kotzé, managing executive, Tellumat Electronic Manufacturing.
What follows is a selection of common issues electronics designers should take note of during the design phase. More detailed or additional information can be obtained from Tellumat during manufacturing engagements.
General approach to DFM
Involve the manufacturer from the start – functionality is typically foremost in a designer’s mind, but the success of a design is just as dependent on the manufacturing process efficiencies it can enable. One should involve a reputable manufacturer early in the design phase and continue the engagement throughout the product development lifecycle. DFM is a dynamic process – what was considered best practice may well have changed since a previous manufacturing run. DFM principles might also differ in keeping with the process configuration or equipment of a specific manufacturer.
Confirm the capabilities of the manufacturer – especially in the case of leadless devices such as ball grid arrays (BGA), it is necessary to make sure that the intended manufacturer has the capability to place these devices accurately. Using a very fine-pitch device in a design may also limit the choice of manufacturer that would be able to produce the product.
Design for automation – designs should aim to minimise the number of assembly and reflow operations. In particular, designs which require manual insertion, placement, dispensing or soldering should be avoided. The board size should be kept within the limitations of the equipment needed for the assembly process.
Design and layout of PCBs
Panelisation – one must allow for fitting multiple PCBs onto a single panel. Panelisation allows for more efficient placement of components during automated assembly by pick-and-place machines as well as quicker inspection and testing.
De-panelisation – the panel design must allow for de-panelisation (separating circuit boards from the panel). Care needs to be taken in having sufficient space between components on the edge of the PCB and the separation edge to ensure that components are not damaged or dismounted during de-panelisation.
Surface utilisation – PCB manufacturing costs depend in part on the size of the PCB surface area. One cost-saving technique includes specifying no more space for breakaway strips for board handling than necessary.
Make provision for test points – the layout of a PCB has a significant impact on the ease and speed with which a board can be tested. In the case of products intended for high-volume manufacture, a PCB with appropriate test points can significantly reduce the time and cost of producing the product. Manufacturers like Tellumat can also assist with test engineering by helping the designer to develop the production test equipment in conjunction with the product.
Pad design – the shape and size of pads significantly influence surface mount technology (SMT) production yield and solder joint reliability. Poor design may result in ‘tombstoning’ or other solder failures when the populated board passes through the reflow oven. In addition to the shape and size of pads, track/pad interaction, solder mask design and component orientation must also be considered to reduce process defects.
Component orientation – when a product requires wave soldering, orientation of components is of utmost importance – especially of SMT devices. This is due to the fact that SMT devices are not necessarily well suited to a wave solder process, but sometimes the manufacturing process demands this. As an example, quad flat packs (QFPs) are susceptible to bridging between pins when they go through a wave solder process. In this case it is advisable to place these components on the PCB at 45° to the direction of the wave with solder ‘thieves’ at the corners. Any slight rotation will dramatically increase the incidence of bridging between pins.
Fiducial and registration marks – in order for SMT machines to accurately place components like fine-pitch BGAs, a reference point or ‘fiducial’ is required on the PCB. Registration marks placed on opposite corners of a device like a QFN (quad flat no leads) or BGA, where the pins are not visible once placed, make it easier to confirm proper alignment of the device during inspection.
Thermal balancing – most metals have thermal inertia, or a ‘delayed reaction speed’ to being heated. The bigger the pad or ground plane used, the greater the inertia. Coupled with the fact that the pads or ground planes on a PCB act as heat sinks, their size and distribution on the board may cause thermal irregularity and cause warping or other undue influence on the PCB when passing through the reflow oven.
Dual sided population – if population of both sides of a PCB is necessary, larger, heavier components should be placed on only one side of the PCB if possible. PCBs populated on both sides need to go through the reflow process twice (once per side). The surface tension of molten solder is sufficient to secure lighter components, but heavier components on the underside of the board may fall off during this process. Having these components all on one side allows the manufacturer to populate and solder this side last.
Component and PCB materials selection
Components dictate processes – one’s choice of components dictates board size and the method of assembly. All parts must be compatible with their intended assembly processes, that is, components must be able to withstand the reflow temperature (typically 220°C–235°C) or the wave-soldering temperature (240°C), if applicable. Should an assembly require aqueous cleaning, suitable parts should be chosen to withstand this process.
SMT – where possible, SMT components should be specified rather than through-hole components in large production runs. This allows the use of automated pick-and-place machines, which offer accurate, cost-efficient placing.
Array packages – when using identical value capacitors and resistors, array packages should be considered. While these packages have slightly higher material costs, this cost is offset by a significant reduction in placement costs and a possible reduction in PCB size (and thus cost).
Component vendor – one should consider carefully before specifying a specific component vendor. Tellumat has built up an excellent component library, offering the benefit of cheaper, more readily accessible or better components.
Component sizes – standardised component sizes have the advantage of more efficient reel loading, resulting in reduced setup costs. It may also contribute to a reduction in machine run-times.
Leaded or ROHS-compliant – soldering with leaded paste aids self-correction of misplaced components, whereas unleaded soldering does not, due to a difference in surface tension properties. The size and distance between lead-free pads on PCBs must also be considered, as lead-free materials require higher processing temperatures, which can cause board warpage.
Stencil design
Aperture reduction – choosing the appropriate stencil parameters, such as aperture reduction, significantly reduces process defects like solder beading, bridging and component float. Aperture reduction is not a ‘one-size-fits-all’ solution. Each design has unique requirements, and a manufacturer should be able to guide the designer through this process.
Cross-hatching – large stencil apertures, especially in the case of ground pads below QFNs, need to be cross hatched. Neglecting to do this may result in excess solder paste being applied to the PCB, causing the component to float and not make proper contact.
In summary
“Tellumat has a wealth of experience to offer designers in the domain of designing for manufacturability,” Kotzé says. “Consulting us early on in the design process and employing the correct DFM techniques will result in a product that is optimised for efficient manufacture, while staying true to the functional design.”
For more information contact Murison Kotzé, Tellumat, +27 (0)21 710 2241, [email protected], www.tellumat.com
© Technews Publishing (Pty) Ltd | All Rights Reserved
printer friendly version