I've been occasionally asked for my view on what is the most common problem that engineers face in circuit design. My answer is simple: misuse of the word and concept of "ground." I am not referring to the many puns, word plays, and jokes you can make using the word in various technical sentences. I have a more serious concern.
Unless your circuit (and system) has a connection to earth via an AC-line ground, or is connected by some independent wiring scheme to earth, you haven’t got a ground to stand on. Yet, as today's products increasingly become portable and battery powered, there is no true ground that you can use as a big bucket of electrons to absorb excess charge or transients, or establish a widespread potential plane. (The non-existence of a true earth ground also applies to many non-handheld products, such as autos, aircraft, and satellites.)
I know that when engineers and other specialists talk, they often use words with not quite their exact meanings, but everyone knows what is really meant (although I did attend a meeting where the term "IP" was tossed around a lot, and it turned out that one group thought they were talking about intellectual property, another thought it was an intermediate processor, and the rest thought it was Internet protocol). It's also true that when most engineers use the term "ground" -- even when they don’t really have one -- that they sort-of know in their hearts and heads what kind of ground it is or isn't. But over time, this mythical ground becomes the hopeful solution to a whole slew of problems with noise, transients, equipotential planes, zero-potential voltage reference points, and more.
It's really worth making the effort to be careful with your words here. If you don’t have a true earth ground, what you actually have is a common point or plane, a current return, a reference, or many other things, but that's it. Being sloppy in terminology leads to careless thinking, especially as the pressures to debug problems mount and missed schedule milestones increase. It then leads to false solutions based on projecting the availability and implications of that true earth ground.
So, if you are a new engineer to the world of real circuits (not just simulations and models), don't fall into the bad and lazy habit of casually referring to "ground" when there isn’t one. If you're a more experienced engineer, you need to be careful, too, both for yourself and for junior members of the team. Sometimes casual thinking becomes ingrained, and before you realize it, what was once just a slang term has become a technical "given" -- except that it's a phantom.
Have you ever been misled or even trapped by sloppiness in technical terminology or descriptions? And how did you find out?