Analog Angle Article

The fable of the cable, or why “strain relief” is also “strange relief”

When your system stops working properly, one of the first areas to check is the interconnection cables. Let's face it, a cable is a potential source of trouble connecting two other potential sources of trouble, as a wise-guy project leader told me, only partially in jest. An intermittent cable can really drive you crazy, since it may behave or misbehave as you move, rearrange, connect, or breathe on it.

Perhaps it is due to my past experience working on large electromechanical systems with lots of vibration and moving parts, but I am a big proponent of providing serious strain relief for cables–large and small, internal and external–as a way to avoid both intermittent and hard cable failures when installed. This means that the wire bundle's outer-insulation jacket should be mechanically connected to the connector shell (or backshell), the connector should be firmly connected to the chassis or enclosure, and the cable assembly itself should be tied down along its travel path. I'd also like to see connectors which provide a reasonable amount of metal-to-metal contact area when mated.

But I think I am in the land of wishful thinking on this. Many of today's products are simply too small to allow decent strain relief along any of those dimensions, and cost pressures further aggravate the connector design decisions. I see, and have been burned by, too many cables with flimsy shells and no mechanical tie-down at all. In many cases, the only thing keeping the connector mated with its other half is the miniscule amount of friction at the metal contact point itself.

That's going to be trouble, for sure, and it will likely be intermittent, which is the most frustrating kind of trouble, since you never know for sure that you have really found the true source of the intermittent behavior; maybe the problem has just “gone away” temporarily, in its own, you wonder.

I know we are not going to see mainstream use of substantial connector pairs such as the well-known “D” connector (DB-9, DB-15, DB-25, etc.) with jackscrews to keep the pair mated. Today's products simply won't allow it, for many reasons. But I am increasingly frustrated with inadequate attempts at providing interconnection integrity which really serve only to provide false confidence.

Consider the ubiquitous, well-established RJ-11 and RJ-45 types of connector used for telephone and network-cable termination. They have that little plastic tang that is intended to lock them into place in their receptacle, with a nice, comforting click you can feel and hear. I have found that in many cases, though, the tang starts to crack at its base after just a few mating cycles, so it no longer performs that locking function reliably. However, you can't see the hairline fracture unless you look at it very closely, so you don't know that your connector (and thus the cable) is not firmly seated in the receptacle and can jiggle, resulting in intermittent mating integrity. This is a headache waiting to happen, as the cliché says.

What can we do? As designers, we can at least try to give a little more consideration to the mundane topic of cable and connector integrity, and try to provide some sort, any sort, of strain relief. As users, we have to improvise as best we can. When I noticed that reliable mating of the dc power-plug connection on my wireless router was very sensitive to cable placement and position (see “My router died, but its replacement kept me happy” for more background on that unit), I simply used a hot-glue gun to attach the power cable to the router case. I can still unplug the connector if I have to (and how often will I be doing that, anyway?) but otherwise, that cable is staying in place. Maybe it is crude, but it is still very satisfying to me, for sure.

And to those of you who wonder about the term “strange relief” in the headline, here's the story: a friend of mine working at a mechanical-engineering publication told me that one year they had a few interns on staff, and when the spell-checker flagged “strain-relief” as an error, the well-meaning intern changed it to “strange relief.” Maybe that makes more sense, in some ways? ♦

0 comments on “The fable of the cable, or why “strain relief” is also “strange relief”

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.