Like Hoffenmeuller, I worked with a bunch of technicians, but at the semiconductor fab I was in, many of them were not as...let's say "diligent". The majority of computers that ran the fab equipment in the RIE process sector had at least one damaged USB port, and one sorry example had both pairs of front and rear ports rendered unusable. This was entirely due to a careless attitude- I even dressed down one tech after I witnessed him "unplugging" a piece of test equipment from about 10 feet away by whipping the cord, and his reaction was as if I was just picking on him- he really did not have a clue!
With proper use and clear access, I have rarely seen a USB port fail, and EVERY one that I HAVE seen fail was through the result of physical damage of some sort. A recent example that comes to mind is a frame mounted computer with a heavy vertically opening hinged door in front of it with no latching mechanism slamming into a protruding thumb drive- proper Engineering controls would not have allowed this to happen.
The damaged USB ports on computers at this fab were clearly the result of physical abuse and carelessness. A contributing factor was the haphazard installation of allot of our "newer" equipment with jury-rigged brackets, no forethought into accessing the equipment for service, and a severe lack of knowledge and discipline in our Equipment Engineering and Maintenance teams. Needless to say, this fab has since went out of business.
Alternatively an even more robust form factor could be the 1X4 position 0.156" centerline, 0.040" square pin system used in many appliances, but with upgraded gold system contacts and deep-drawn EMI shields (Cheap in high volume, but more robust than having a formed seam) for electronics applications !!!
This could be marketed with a catchy name indicating it it robust such as "Super USB", or "USBrute" interface. The pin assignments and protocol could be identical to the current USB specification for less ambiguity. This could lend itself to an adapters from plain old "USB Type A to USBrute", and USB Type B to USBrute" for back compatability and use of current devices such as jump-drives.
From a connector design robustness perspective, I like high density D-sub, the .156" centerline pin and socket connectors, and 0.100" centerline .025" square pin connectors (e.g. - the old Berg Latch-N-Lok) with EMI shield, a rectangular key & keyway polarization, and latching.
The EMI metal-shielded versions of these tend to be more expensive and no standard has been built around these form factors, axcept the VGA Monitors around the 15 pin high density D-sub.
What would stop IEEE from specifyng a 1X4 pin .100" centerline .025" square pin shielded and latching version of USB? This connector is not indestructable, but it could be more robust than the current form, could have staged power engagement with staggered pin lengths and could be with latches or jackscrews.
What would stop electronics designers from then designing in these more robust forms so that connector companies would fall all over themselves to produce them? Yes it may cost slightly more per line at first, but for printers, mice, or laptops that the connectors keep getting pulled-out it could be marketed as a better, more robust product. There are some very cost effective single-beam select gold plated female contact designs out there (e.g. - 3M IDC, Tyco, Hirose, Methode, etc.), all the way to 50 microinch or thicker gold plated military spec box (e.g. - Tyco Box, Berg PV, et.al.) contacts in this form factor. It would be best for the specification to define the plating system compatability such as Nickel/15 microinch minimum Gold plating and mate compatable such as Nickel/Paladium-Nickel/Gold. This is to avoid incompatable mixed mating between gold systems and solder plated systems. Incompatable plating systems can lead to fretting corrosion failure problems.
How about it, can you techy consumers out there support building products with a more robust interface with your purchasing power?
Very nice POV from David12345. From my personal experience working in industrial electronic maintenance, supervising and working with a group of technicians, I've never experienced or heard of a USB port failure. I expect that a group of technicians are probably more aware and careful than a group of students or some individuals, but both in our line of work and personal PC use we use USB many times daily. While D-subs have their unique benefits, we have experience failures there the bulk of the connector and cable has caused PCB connection failures. While acknowledging some of the idiosyncrasies noted, such as insertion orientation (even these have not caused damage for us), I have no recommendations for a better connection. For repeated use, USB is better than RJ-45 or any other design in my experience. The approach of "If it doen't fit the first time, flip it" has not caused damage that I expect could result from a brute force, jam it in there approach. It sounds like a little practicle patience is more applicable than redesigning the most convenient and widely adopted port in memory.
Thanks for the great discussion form the engineering point of view, David. I never thought about all that went into a "simple" connector. As for me, I have never experienced a great extent of USB problems from the personal / office type usages. My previous employer's machinery is another case, but that is to be expected for the extreme environments they are operating under.
As a former electrical connector design engineer (including one design USB type B Header), I recognize the design trade-offs. The USB was very cost sensitive and I believe the standard form factor was dictated by a number of companies, but initially was largely driven by DEC in Taiwan around 1995. This was then joined by Compaq, NEC, IBM and others.
Yes, "D-subminiature" connectors are very robust; however, they are somewhat more complex to fabricate, bulkier, and have some issues with pin damage or female contact damage. They also have mechanical limitations with staged engagement as desired for hot-mating.
The "Centronics" style connector was cheaper and robust, but still bulky. Also, with new standards there is often a desire to have a clearly unique form to avoid incorrect plugging of a non-USB device into a USB port. Again, they also have mechanical limitations with staged engagement as desired for hot-mating for the USB application.
The "Circular DIN" connector has some inherent manufacturing challenges that keep the cost higher, and less attractive for a price-sensitive application.
The USB also provided staged engagement between logic and power to permit "hot-mating" with minimal risk to the electronics, and teminals in an eclosed shell to reduce static shock damage to electronics.
I can see how the USB connector is prone to damage from abuse. It is very inexpensive and therefore attractive to OEM designers pressed on the bottom line results. It was designed to have a longer shield engagement to minimize angular loads onto the plastic, but with enough force and leverage sticking out, the stamped shield will bend and allow the plastic to break. It has very diverse operating environments. If broken, it can be replaced, but admittedly only by removing the PCB and replacing the header using a soldering iron. Of course, now there are the USB A, USB B, and at least 2 versions of miniature USB connectors for small applications such as phones and cameras. I am amazed how much abuse these delicate little form factors take and still work.
Design is often a trade-off compromise. Many military and German products are very robust, but meet little market demand due to the price and lower cost competition. Many electronic devices are obsoleted before they die even with these inexpensive constructions.
About the 500 cycle life contraint. This system is to be designed for that life and meeting the required performance. Naturally, the device can be destroyed in one cycle with abuse. If there is no environmental exposure to salt spray, humidity chambers, or other corrosion accelerators, I would expect the gently handled connector to function for 2000 to 5000 cycles (or even more). However, once the gold plating integrity is violated (usually after 250 to 1000 cycles with 15 to 30 microinches of gold), it becomes very sensitive to "flakey" intermittant high resistance connections from oxides. The environment can determine how much oxides are a problem and a mate/unmate cycle can crack through the oxides again for a good connection for awhile.
Sorry I got to this discussion so late, and I hope there is someone still following the thread. I teach part-time in a private boys school. Our computer lab regularly has broken USB connectors on the socket portion. It is not unusual to just hsve a hole in the case through which you can see the hardware dangling just out of reach. I will be the first to admit the boys are less than gentle, but I cannot remember ever having a problem with any of the other connectors. I think there is a leverage factor at work here, particularly with flash drives.
All of the posters who have never had a problem, have probably never shared a computer with a group of kids who could break a cannon ball. As far as the markings, I have (2) flash drives, which were purchased very cheaply at a Big Lots, that will fit the socket in either direction. The only way you can be sure they are properly inserted is when the computer recognizes the hardware on the screen.
You're right, KGround, my memory sticks do not have a USB symbol at all. However, following the logic of Apple's connectors, the name of the maker is on the top of my stick for inserting it top up into the horizontal slot on my Mac laptop.
Yes, Kim, since your posting, I have become more aware of the orientation of the USB connectors. I never realized, but they are all marked with a symbol of some sort that indicates orientation. How about that.
In many engineering workplaces, there’s a generational conflict between recent engineering graduates and older, more experienced engineers. However, a recent study published in the psychology journal Cognition suggests that both may have something to learn from another group: 4 year olds.
Conventional wisdom holds that MIT, Cal Tech, and Stanford are three of the country’s best undergraduate engineering schools. Unfortunately, when conventional wisdom visits the topic of best engineering schools, it too often leaves out some of the most distinguished programs that don’t happen to offer PhD-level degrees.
Focus on Fundamentals consists of 45-minute on-line classes that cover a host of technologies. You learn without leaving the comfort of your desk. All classes are taught by subject-matter experts and all are archived. So if you can't attend live, attend at your convenience.