True plug-n-play is the holy grail manufacturers strive to attain when systems must work with each other. The ideal scenario lets users take a device out of the box, plug it in, and start using it. But in the real world it’s more like “plug-n-pray” that users don’t experience too many difficulties getting a device connected and operational.
Just because two units have the same communication hardware, like a USB port, doesn’t mean they’ll be able to talk over that connection. Hardware is only half the equation. The other half is making sure each device understands what the other is saying. A telephone isn’t much good for talking if the person at one end only knows English while the other speaks only French.
Fortunately for the sake of international diplomacy and system integration, the wire, cable, and cable-assembly community has embarked on a revolutionary plan to help OEM-equipment designers change the way they deal with interconnectivity via the cables that connect devices. Through the use of thinking cables, system designers can literally think outside the box — data translation services can reside within connecting cables.
Listen to Joe Iamartino and Bob Repas discuss thinking cables in a podcast at:
Like most communication technologies, there are many forerunners that claim rights and precedence for thinking cables. One inventive thinker rushed to aid a struggling progenitor in the early 1990s. A designer of handheld bar-code scanners had to communicate with electronic cash registers from IBM, NCR, ICL, and others, each with different communications needs. One answer: Embed active components on a PCB directly into the cable assembly. Users merely bought the right cable for their register. The cable translated the scanner data into the required format. Power for the circuit came from the cash-register connector port.
While this solution appears simple and straightforward, the final design had to overcome significant engineering challenges. For example, the circuit pod needed strain relief on both sides to withstand the severe flex and movement of a consumer-oriented, high-volume, point-of-sale (POS) application. The active circuit board inside the pod had to be potted to secure components against that same abuse while the connections to the board demanded enhanced soldering techniques to ensure long-term reliability.
By 1994, shipments of handheld bar-code scanners that used thinking cables reached millions of units per year. One manufacturer of active cable assemblies, C&M Corp., had hundreds of thinking- cable designs in production throughout the 1990s. Some designs eliminated the midcable pod by moving the active circuit board into the connector end.
In the 1980s designers from Digital Equipment Corp., IBM, and others were incorporating resistors, capacitors, inductors, and other components into their cable assemblies to modify the waveform of the signal. Some designers used the electrical properties of the cable alone to affect the modification. While these cables could be called active cables because they did modify signals, most did not need external power to perform their magic. The signal only had to travel from one end of the cable to the other. Today, it is commonplace to find compensating circuits embedded in Fibre Channel and Infiniband copper-cable assemblies. They are just one form of thinking cable.
Thinking cables help exceed the limitations of standard cabling. In fact, there doesn’t appear to be any limits as to what a thinking cable can do. For example, a copper connector with an optical cable hybrid assembly could be 100 longer than a standard multiconductor copper cable while remaining lighter and more flexible. Of course there are still technical limits to consider such as the heat dissipation of the embedded electronics and meeting UL and other regulatory safety requirements. Thinking cables also work better in point-to-point situations, such as machine-to-machine communications, rather than as part of an infrastructure wiring plan.
When datacom fiber optics became commonplace in the early 1990s, OEMs produced either copper or fiber-based equipment. The media converter and pluggable port was created to give customers the option of selecting different media (copper versus fiber) or transceivers for different optical distances. Thus, the same base equipment could serve different connection needs by merely swapping media converters. They were given names such as GBIC (gigabit interface converter), SFP (small form-factor pluggable), and others.
Thinking cables can replace the multitude of media converters and add-on accessories. In place of an external pluggable device, the equipment carries just one interface port that handles the thinking cable. The cable carries all of the electronics to adapt the interface to the medium of choice, whether it is low-loss copper or fiber optic. Changing media is as easy as swapping cables.
Standards committees, notably the Power over Ethernet (PoE) group and Infiniband Trade Association group, recognized that providing power at the connector port could support a variety of uses. For example, it could spur development of multifunction devices, such as rotating security cameras; trickle-charge laptops; run sensors; and power transceivers and equalization circuits in cables. Seeing the benefits of such an option, the Infiniband standards committee changed their specifications to allow power at the connector for active high-speed interconnect applications. Powered emphasis circuits in cables can boost operating distances over noncompensated assemblies. Whether in industrial automation, high-speed communications, or general communications, the idea that power at the connector port can spur innovation has taken hold. All of these changes individually are interesting; but, taken collectively, they point to a monumental change for the cable-assembly industry.
While most people prefer wireless devices, many devices cannot be wireless due to reasons of security, distance, EMI/RFI issues, and power consumption. Our current technical ability to supply power to equipment without plugging a cable into an electrical outlet is limited, though wireless transmission of power has been demonstrated and is certainly possible. But for the near future it appears that cables will carry most of the power. Some OEM designers, recognizing this fact, have opted to change the cable rules completely.
Until recently, most OEMs looked at cables as a component to be sourced to the lowest bidder, built in places that only the National Geographic Society had ever visited. But designers now see that cables can do more for them than just passively carry signals. In fact, thinking cables can spur cost reductions that far exceed the cost of the cable itself. Used properly, they can reduce costs across an entire product platform.
A prime example would be the creation of a single interface for both copper and fiber. This would free OEMs from having to carry both copper and fiber solutions for their LAN, SAN, or WAN equipment. Currently OEMs use copper- cable assemblies for the copper port and fiber-optic assemblies for the fiber ports. Each type of port has either a permanently mounted transceiver or a plug-in style transceiver that needs a cage or carrier system to hold it. By changing to a single powered interface, the OEM benefits from having just one electronics unit to inventory. The customer then chooses from a range of conventional and thinking cables that support the speed and distance needed for the application.
Instead of choosing fiber or copper, the two mediums merge into a hybrid copper/fiber system. With power available at both ends, copper connectors with embedded electrical-to-optical converters can extend the reach beyond conventional copper cables with fiber optics, yet still keep the two best features of copper connectors: their robustness and self-cleaning gold contacts. The approach eliminates ceramic or polymer-fiber ferrules with the fiber bonded directly to the optical components within the connector, wringing out significant cost and keeping dust, oils, and scratches from degrading the light path. Conventional copper cables can plug directly into the same ports when shorter distances don’t warrant a fiber connection.
OEMs can have quick-turn assembly vendors build to order either conventional copper or copperfiber hybrid cables. There is no need for the OEM to inventory standalone transceivers or fiber-optic patch cables. Pushing the assembly of embedded optical transceivers to the cable suppliers lets them make use of vertical integration with the low-end cable assembly business to cut costs. Meanwhile, the OEM realizes savings on transceiver packaging and fiber connectors.
Chips capable of self-identification and test embedded into cable assemblies let manufacturers run real-time wiring analysis between active devices. Other awareness activities and services can be built into a thinking cable, off-loading the main electronics that the cables are plugged into. Infrastructure tracking costs can be greatly reduced by self-aware cables with broadcast abilities. Even a lack of messages from a defective thinking cable can be a call for help to attentive electronics.
With RoHS and other green initiatives driving designers to take a second look at their methods, thinking-cable assemblies would have enough value to justify return, repair, or recycling — the three Rs of green design. Manufacturers can design the cables to be upgradable or recyclable and offer these services for OEMs.
Thinking cables can also be a boon to data centers. As blade servers and ever-shrinking switches and routers implement data speeds of 10 Gbps or faster, the heat density grows significantly in those concentrated racks of equipment. Cooling specialists Emerson/Liebert and APC have told customers to take a close look at those hot spots. Moving relatively low-power transceivers and the electronics into the cables frees up valuable printed-circuitboard real estate. It also pushes the minimal heat load from the hybrid assemblies to the cable assemblies. These cable assemblies typically sit in the ambient coolair environment of the data center, not in the sweltering electronics cabinet next to some overheated processor chip. If speed constraints make cable elimination impractical, designers might as well use the cables for other purposes, like doubling as heat sinks outside the active box, to cut costs or to improve mating reliability.
Thinking cables are not a panacea to all interface problems. First, they work well in machine-to-machine connections, but their use in infrastructure areas is limited. With power applied to electronics within the cable assembly, the cable now falls under regulatory areas for safety and other considerations. This may mean submitting the cable to UL, CSA, and other agencies for approvals as needed. The plastics making up the cable have to be safe and recyclable today yet still meet the needs of the application. And all this has to be cost effective for the best dollar-togigabit ratio.
It won’t be long before thinking cables become mainstream in our dayto- day lives. Of more interest will be the impact on OEMs that adopt the technology early to take advantage of the long term cost benefits. Those waiting too long to develop suppliers capable of working with hybrid copper-fiber designs and embedded active circuits might find themselves engaged in a cost war they might not survive.Make Contact
C&M Corp., (877) 830-6057,