标签: production

In the fast-paced world of active devices and the very tangible area of test equipment, the lowly cable and associated connectors sometimes seem to be almost invisible, in both the prototyping and production cycles. That's not only unfair to their self-esteem (to use the trendy parlance), but unwise. Sure, we hear that "everything is going wireless," but the reality is that hard-wired signal interconnects are an indispensable and irreplaceable part of nearly all systems.   It gets more challenging: as more design engineers work on projects that extend into the gigaHertz and higher range, they'll have to learn to give more consideration to the cable assembly as a critical and even dynamic part of the design. If you look at RF/microwave-centric publications and web sites, about one-third of the ads and content are devoted to the subject. It's a world where second- and third-order parameters such as phase matching between two nominally identical assemblies (such as used for phased-array radar) become critical; even the temperature coefficient of a cable's specifications can be a concern. The assemblies are carefully engineered, modeled, tested, and fabricated energy waveguides with precise dimensions, special internal and external insulating materials, and more.   Fortunately, cable assemblies are getting some attention from diverse perspectives, as shown by these unrelated features I have seen in the past few months:   We hear a lot about counterfeit components—mostly ICs and passives—but we don't see much about the situation with cabling. Yet, it seems to be a serious problem, especially as the cable may work to some extent if not full spec. According to Cabling Installer, their 2011 article on the subject Counterfeit cable exposed was among their top 10 articles in 2014, three years after publication! (Also see Counterfeit cable is getting ugly .) The result is a cable assembly that doesn't fully meet the operating spec and may sort-of work but only support lower data rates, or a PoE (Power over Ethernet) installation unable to provide the specification's power.   It's not just electrical performance, either. On the safety side, a cable’s insulation in most installations must be fire-rated to not support combustion, yet these fakes fall far short— and it's an issue that you won't know about until a fire breaks out. Would you have even imagined presumed copper cable that was really brittle aluminum, but with a copper cladding? Think of the installation and performance surprises on that one!   The thing about cable is that it is so easy to make a fake, and put almost any rating you want on it. After all, who "tests and verifies" the cable's performance when data rates go down and BER goes up? Who checks the insulation? You may take it as a "given" that if it says Cat5/UL rated, it actually is, but that's naive. Note that this is not a new concern. Even in audio, non-RF applications, I often wondered about the super (aka "monster") audio cables that were sold as OFHC (oxygen free, high conductivity) ostensibly because of their superior performance potential. How do you really know the vendor isn't shipping plain copper slapped with an OFHC label?   Problems with RF/microwave cables in complex installations. I saw an ad (yes, a print ad!) from W.L. Gore stating that "a recent study for the aerospace industry showed that more than 29 percent of microwave cable assemblies fail during installation , and aircraft manufacturers have accepted the practice of simply replacing them." Wow, I thought, that has to be costly and time-consuming, in so many ways! Gore's solution is to provide a simulator to evaluate the stress of installation, with features which replicate bend radius, routing guides that induce torque, and abrasion bars to simulate routing across sharp edges of access holes in the airframe structure.   Finally, there's the world beyond extending to 100 GHz, well beyond modest 1 to 10 GHz, where spectrum is available and new components make reachable. What sort of cables and connectors can you use there? It's a brave new world of tiny, tinier, and tiniest, with hair-thin cables and corresponding matchhead-sized connectors.   This V Connector (male and female) from Anritsu provides coaxial coverage to 65 GHz, with a 1.85-mm geometry endorsed by IEC (International Electrotechnical Commission); it mates with standard available 2.4-mm connectors.   Cabling seems to be one of those things with which engineers have both love/hate and ignore/fear relationships, it often seems to me. For example, my cable modem and wireless router sit next to each other, but they are connected with a 20-ft. (7-m) Ethernet cable, because that's what I had handy on the day of installation. Yet when recently I found a 2-ft. (0.7-m) cable in my spare-cable collection, I wasn't sure if I should swap the shorter one in for the longer one. On one hand, the shorter one is more elegant and appropriate, and wouldn't need to be coiled up and tied back; on the other hand, everything is in place and working, so why take a chance and upset things?   While basic cable-assembly connectivity and continuity is very easy to verify with an ohmmeter, the actual performance to specifications of the assembly is not—especially for non-electrical parameters. What's your relationship with cables, connectors, and assemblies? Have you ever overlooked something that came back later to cause you design, debug, or production aggravation?

A number of years ago, after an irksome day interacting with production (when we still manufactured in Canada), I complained to my supervisor—the Gradin portion of this theory—that "If you took two people and added their IQs together..." At that point, I paused momentarily, and he immediately interjected, "You would get a negative number." The engineer in me wondered if this could be possible. This led me to a plausible explanation. The MBA in me wanted this expressed as a theory. Boredom while waiting for a plane's departure led me to write this down. The Gradin-Kagan effect (or worker empowerment: phooey) Any element of society is subject to the law of entropy. Unless forced to assume some entity by an external force, it will degenerate into a form that approaches chaos. In the absence of clear, unambiguous, nonconflicting, and attainable goals, the Gradin-Kagan Effect will manifest itself as follows: Hypothesis No. 1 (Kagan's Law or the inverse synergy effect): When two or more people work together, the effective IQ of the group is almost always less than the simple sum of their individual IQs. IQ is not a scalar quantity. It has at least one other dimension that prevents the simple addition of IQs when several people work together. Hypothesis No. 2 (Gradin's Law): When two or more people work together, the effective IQ will always give a negative number. In a normal distribution of the vector IQs of a group, it would be expected that the second dimensions of the IQs would cancel each other out, giving an effective IQ of zero. However, in what must be a previously unobserved effect of Murphy's Law, the effective IQ is always less than zero. Hypothesis No. 3: Any person working in a group will have his/her IQ reduced when making any simple decision. Through some perverse feedback effect, as yet unexplained, the negative number derived through Gradin's Law is added to each individual IQ, thereby reducing that IQ. For example, an individual with sufficient IQ to distinguish which end of a burning cigarette to put in his mouth, in a workgroup setting, would happily lick the hot end and still be surprised at the resulting burn. According to the Dilbert Principle, all management is incompetent. As a result, there may never be a situation where the Gradin-Kagan Effect is ever negated by suitable goals. By the way, the burning cigarette part is true. Have you ever felt frustrated enough with the stupidity of other people that you would be willing to share your experiences with the rest of us? Aubrey Kagan is Engineering Manager at Emphatec.

Several years ago, after a particularly exasperating day interacting with production (when we still manufactured in Canada), I complained to my supervisor—the Gradin portion of this theory—that "If you took two people and added their IQs together..." At that point, I paused momentarily, and he immediately interjected, "You would get a negative number." The engineer in me wondered if this could be possible. This led me to a plausible explanation. The MBA in me wanted this expressed as a theory. Boredom while waiting for a plane's departure led me to write this down. The Gradin-Kagan effect (or worker empowerment: phooey) Any element of society is subject to the law of entropy. Unless forced to assume some entity by an external force, it will degenerate into a form that approaches chaos. In the absence of clear, unambiguous, nonconflicting, and attainable goals, the Gradin-Kagan Effect will manifest itself as follows: Hypothesis No. 1 (Kagan's Law or the inverse synergy effect): When two or more people work together, the effective IQ of the group is almost always less than the simple sum of their individual IQs. IQ is not a scalar quantity. It has at least one other dimension that prevents the simple addition of IQs when several people work together. Hypothesis No. 2 (Gradin's Law): When two or more people work together, the effective IQ will always give a negative number. In a normal distribution of the vector IQs of a group, it would be expected that the second dimensions of the IQs would cancel each other out, giving an effective IQ of zero. However, in what must be a previously unobserved effect of Murphy's Law, the effective IQ is always less than zero. Hypothesis No. 3: Any person working in a group will have his/her IQ reduced when making any simple decision. Through some perverse feedback effect, as yet unexplained, the negative number derived through Gradin's Law is added to each individual IQ, thereby reducing that IQ. For example, an individual with sufficient IQ to distinguish which end of a burning cigarette to put in his mouth, in a workgroup setting, would happily lick the hot end and still be surprised at the resulting burn. According to the Dilbert Principle, all management is incompetent. As a result, there may never be a situation where the Gradin-Kagan Effect is ever negated by suitable goals. By the way, the burning cigarette part is true. Have you ever felt frustrated enough with the stupidity of other people that you would be willing to share your experiences with the rest of us? Aubrey Kagan is Engineering Manager at Emphatec.  

I was pondering my past experiences working on various cost-savings projects. I've worked in a product cost reduction group and a company expense cost reduction group. With that experience in mind, I'm looking at many companies that are going through the motions of cost cutting, saving a few hundred dollars here and there. At the same time, they are spending more money on projects and issues because of the way they tried to eliminate costs or save money. I know that no matter what position we hold, we'll come across a cost cutting activity that just makes no sense. Sometimes to a point where you just drop your jaw and raise your hands and cannot find a single word to express your combination of confusion and frustration—what I call a case of frusfusion . Have you ever been frusfused? . Here is my experience, followed by an example I have seen elsewhere. In the days when I was involved in laying out PC boards (PCBs), I regularly asked our managers to OK the purchase of a couple of Altium Designer licences for our engineering team. We had no layout tools. We still lived in the old days of drawing schematics by pencil. I always joked to management that we should buy some lanterns and candles to place in the lab for when we work late at night. . I would send off about 10 to 15 PCBs per year to have layouts and modifications done. I always spent about $10k or sometimes a bit more per year on having a third party do our layouts. Even though it was only about $6k for a couple of licences, I could not convince management that spending $6k was cheaper than spending perhaps $10k to $15k per year on layout and Gerber file creation. . I could lay out the boards myself just the way I needed them and get them more quickly than sending out. Instead, I sent out the schematic, paid more money, and twiddled my thumbs for a week or two waiting on my boards. And hoping they were laid out correctly. I just never understood the justification on why not to buy. . I put up the fight for about four years. On the fourth year, someone must have opened his eyes—I was allowed to purchase one licence for three people. Sharing the one licence was a struggle, but the biggest struggle was having to take six months to learn the program instead of paying for the two-week training. . Another example is watching engineers passing the opportunity to buy $500 evaluation boards at the front end of a project because they don't want to spend the money and think they have a better way to do the design. But then, when it is two weeks before production is supposed to start, and the design is not working, and issues are arising one after another... Woops! Should have spent $500 nine months earlier instead of delaying production and spending thousands now, just to put a big bandage on a design.. I always say that trying to save $50 will cost you $500. However sometimes spending $500 will save you $10,000. Has anyone else had similar issues on cutting costs that cost you more?   James Bowden Design Engineer