Keeping Connected
Connector technology still grapples with smaller platforms, higher speeds and placement issues.
There are thousands of application-specific electronic connector products, with pin counts numbering from one to thousands. Connector varieties mirror the huge diversity of electronics products and applications. Connectors are also used in electrical applications, such as automotive and appliance, and where electronic power connections are required.
A few types of electronic connectors include:
Terminals and splices.
Rectangular and circular connectors (including RF/coax).
IC sockets.
Connectors for electronic systems (production).
Connectors for IC test, burn-in and other test.
Printed circuit connectors.
I/O connectors.
Wire and cable connectors.
Fiber-optic connectors and associated hardware.
Many, many application-specific connector designs.
Table 1 breaks out electronic connectors by equipment type. Connectors can be made from a range of technologies, adopting a multitude of new materials and processes, as well as value-added properties such as embedded silicon (memory cards), signal conditioning components (filter connectors), and printed circuits (card edge, FEC – in short, anywhere circuit elements need to be connected).
General use of connectors will not change until PCBs become more integrated. As with other electronics components, connector technology is driven by higher frequency signals, increased packaging densities, assembly processes, industry standards and environmental regulations. Some of the key areas where challenges are expected to occur are described below:
Sub-miniaturization of electronics packaging platform (e.g., HDI, 3-D, printed electronics). Requirements below 200 mm pitch and 0.5 mm mated height may require prototype/experimental interconnect technologies, such as etched foil or MEMS-type micro machining, which would also require some form of micro-robotic insertion. For vertically integrated OEMs, this would have been possible, but is less likely with today’s pervasive outsourcing. Mainstream merchant PCB technology – an area where suppliers tend to have low R&D budgets – is moving in this direction with high-density interconnect (HDI) and microvia technology. Flexible etched circuitry comes closest to the boundary where subminiature connectors are approaching limits. This is primarily in small handheld portable applications such as cellphones, digital cameras, etc. Connector technology continually adapts to new circuit and packaging requirements, and future technologies are expected to enable smaller, subminiature connectors with very thin sections, tighter tolerances and higher temperatures.
Very high-speed interfaces. Serial circuitry and advanced signal conditioning have upped the ante to 10 Gb and beyond. Differential signaling and air dielectric backplane systems have increased copper connectors’ reach to 20 Gbps and beyond. Fiber optics, essentially a point-to-point technology, is a viable alternative (with architectural modifications), as are hybrid Cu-FO designs, which are already in limited use.
Materials cost inflation. Materials cost was a serious issue for connectors and other products in 2006-2007, when copper, nickel, tin gold and many plastic materials experienced major price escalation. Except for precious metals and some other materials, global pricing deflation has occurred with the current recession. Proprietary efforts have had some success in developing minimalist/substitute materials and processes. Connector costs have bottomed, with many units already being assembled in Asia.
Wireless interconnect. A multitude of wireless technologies are on the market. Some, such as Bluetooth and the recently released Certified Wireless USB, attempt to reduce cable clutter. This will shift some applications into the wireless realm, replacing some connectors. The net effect, however, will be more system-level product volume and, overall, more connector volume. Even in so-called wireless applications, connectors remain.
Potential supply disruptions. Supply-chain risks are associated with a global marketplace, including reduced or eliminated local manufacturing infrastructures as production migrates to Southeast Asia. Many mainstream products are now built in offshore manufacturing venues: notebook and desktop PCs, motherboards, most handheld devices and board-level products for many applications. This dependence on emerging world economies – where there are concerns about infrastructure stability, IP theft, counterfeit products, climate and other issues – could result in supply disruptions. A supply disruption with China, for example, would be catastrophic, and unrest elsewhere could constrict market and EMS manufacturing access.
Use of Connectors with Press-Fit Assembly
The Board Assembly chapter of the 2009 iNEMI Roadmap discusses some of the challenges of inserting connectors when using press-fit assembly. Some include:
Automation. Assembly automation continues to be a challenge for some connectors, although the industry has widespread tape-and-reel, tray and other pick-and-place packaging. In addition, odd-shaped component placement has come a long way. However, miniaturization increases the need for higher contact insertion precision. With press-fit assembly, for example, the industry (in some cases) uses slower, manual placement methods to load connectors on the board before the press operation. Manual insertion has become more difficult as pins get smaller, shorter and have higher density, and cycle time for manual placement has increased by an estimated 20%. But even here, some suppliers have designed new press-fit pins.
Currently, there are a limited number of equipment vendors or high-cost robots to automate odd-shaped connector placement onto boards. One reason for the lack of equipment is the lack of standardization of connector trays. Connector manufacturers tend to use trays that match their packaging, and no collaborative industry effort has been undertaken to standardize the trays. Such standardization could go a long way toward being able to automatically place parts for the press operation. Better communication from OEMs and EMS providers to connector vendors would also help.
As cost pressures increase, throughput will become more of a focus, and there will be more demand for automatic placement machines, provided they can do some (if not all) of the following steps: pre-optical/laser inspection of pintails, placement, and inspection before the press fit operation.
The size of the connector pintail, alignment, true position and average offset of the wafers play a role in the ability to place a connector. AOI use for pintail alignment and true positioning would provide some assurance that connector pins are placed into the holes without stubbing against the PTH wall.
The need for a standard system for automatic pick-and-placement of connectors has been understood for more than 10 years. Different platforms to address this need were released to the market, but due to market conditions, sales were weak and the systems were withdrawn because of demand.
Current systems on the market range from semiautomatic presses through fully automatic press systems. However, the automation only pertains to the actual press cycle, while loading the connectors onto the PCB is still manual.
A re-evaluation of the need for a pick-and-place solution is expected with the recovery of the telecommunications and networking industry. The move to smaller pins and denser pin fields is also expected to further drive the need for a pick-and-place option. As the market evolves, one potential solution is to modify an existing automatic press system to act as a pick-and-place or a pick-place-and-press solution. Another potential solution is to take an existing odd-form pick-and-place-system and modify it for use with connectors.
Inspection. Inspection is a critical area of the press operation. Regardless of the type of deformation (accordion, smashed, etc.), the primary concern is a connection that may pass an open/short test, but fail in the field due to the connection opening up. This can occur if the pin shorts to the top of the barrel and passes the electrical test.
If the pins are long enough to protrude through the board, AOI or visual inspection can catch a pin that does not protrude fairly easily. However, if the board is so thick that pins cannot protrude, inspection becomes tougher. Different methods to detect pin presence have been attempted, but no solution works in all situations and is scalable. The feedback system of the press is used at many sites to determine if the press was successful, but even this method is not 100% accurate.
The most difficult and elusive inspection in the press-fit connector process is when two connectors are pressed from both sides of a board, with pins from each connector entering the same hole. There currently is no high-volume algorithm, machine or technique to accurately determine whether both pins are properly inserted into the barrels. As shown in Figure 1, the bottom pin can be bent and still maintain contact with the barrel during assembly and testing. Without a means of accurately identifying that one of the pins is bent, this defect can reach the field and then open up.
Repair. EMS companies are challenged by many different connector types and the rework methods required to remove pins, wafers and connectors. Many assembly sites use pliers to pull the pins, although most connector suppliers sell repair tooling. The need for connector manufacturers to design unique features to permit easy removal, while preventing damage to the PCB, is encouraged, and common tooling to rework connectors is suggested. Connector bodies need to be designed for removing the connector, as well as placing it. Strategically located tooling holes for removal tools need to be designed into the housing. Also, the mechanical strength of the body should be sufficient to permit removal of all pins at once, without mechanical failure of the body.
Key Areas in Development and Commercialization
Connector developments follow OEM requirements, which are now shared with – or referred to – EMS providers. Key areas of interest will include micro and nano materials and process developments, high-speed electrical performance and miniaturization. Mobile/system-in-package (SiP) interconnect requirements may drive future micro-scale robotic connector design plus other dimensional requirements outside the realm of conventional stamp and form/mold connector processes:
BGA attachment with mechanical integrity for advanced SMT applications is now a reality.
10 Gbps now, 20-40 Gbps in the future, with higher-density connectors (> 100 signal contacts per inch), both differential and single-ended signal applications, or low-pin-count serialized interfaces (i.e., backplane systems).
High-speed copper and FO cables (Infiniband, Fiber Channel, 10 Gb Ethernet, etc.).
LGA socket I/Os have reached 1366 with PC processors; I/Os may drop to 1 or 0.5 mm pitch.
New high-performance memory sockets – DDR3, DDRx.
Sockets with multi-GHz level signals are dependent on PCB layout; future DIMM may be a two-piece or new design.
Serial or parallel optical interconnects; board-level waveguide packaging with optical ICs.
Value-added connector assemblies to reduce system cost and complexity, yet sustain the connector function.
Connectors are unique among electronic components because the applications, designs, materials and processes are almost limitless. Many designs have unique customer requirements, while others are multiple sourced industry standard products.
Thus, the industry really has two parts: a huge multi-billion-unit industry or OEM-specific standard segment, and a large semi-to-full customization segment with many market niches and unique customer designs. The roadmap will combine all these diverse characteristics as it seeks to fulfill future industry requirements.
John MacWilliams is a senior consultant and analyst with Bishop & Associates (bishopinc.com.com) and chaired the Electronic Connector chapter of the 2009 iNEMI Roadmap; uscompetitors@yahoo.com.
