Humans understand language in a contextual way, and as such I think the discussion is almost (note almost) mute. Any experienced engineer will instantly know from the context what is meant. In training on the other hand there is a need to be more specific. I tend to use the term ground reference in measurement and power ground in supply situations an earth when I'm talking about true connections to "earth"
Much ado about nothing. We have different symbols for both. We often talk of returns but quite frankly, if I were to tell somebody that this expanse of copper on the PWB/PCB was a return plane, they would look at me strangely. Our dialects are not always syntactically correct but understood. If there is a chance for misinterpretation, then precise language is called for, e.g. SOW, requirements document, contract, etc. Otherwise, it's no big deal.
Great point you made in the article. I work with power supplies so the difference is crucial. I have taken to using the terms return ("analog return" or "digital return" or "3.3 V return") and "protective earth".
The challenge is that old habits die hard. There is a "True Ground", most of us stand on it. Common is certainly a much better term, particularly when it has the modifier preceeding, as in "analog common". The actual concept is a bit more complex, because the physical implementations, (planes), are at best fine networks of resistances, and at worst a complex web of various resistance and impedance values. So the result is that there is one more thing that needs to be explained much better back in the basic circuits class.
I work in the certification industry, where the conflict in reference has created a lot of cunfusion for our new engineers and clients. There are specific requirements for creepage and clearance distances to ground, voltage to ground, etc. Also we see voltage maps that indicate 0 volts for the common, which of course the same potential as the circuit referenced to it while true ground is required to be at 0 volt potential under normal operation. Adding to the problem is submittals from Asia and Europe where what we refere to in the US as "ground" is called "Earth." The documentation we receive refers to common as digital ground, analog ground, etc. which as stated have no reference to true ground or earth. I have been trying for years to train our engineers and clients to use the correct terminology, with limited success.
I also have seen confusion arise over the use of "return" and common and ground. Currently (pun intended) I use Batt Neg or Neg as the label for the "common" in a battery operated system. In some cases, say with H-Bridge connections, the polarity is switched between the two output connectors and in this situation I label one + and the other - to indicate the normal and reverse polarities.
I agree with your comments- when it comes to EMI/EMC issues, it's imperative that we talk about "common planes" and "return planes" rather than "ground planes".
On my schematics these days I no longer use the arrow shaped ground symbol. Instead, I have created a symbol called "0V" to denote the common 0 Volt connection.
I think many of us, perhaps most of us, misuse the terminology all the time. For some reason it is uncommon for engineers to refer to a "common" (sorry, I couldn't resist that one!). I hear engineers use phrases like "circuit ground" or "supply ground" when they really mean "common." Then of course a mixed-signal system always has an "analog ground" and a "digital ground," with the tacit understanding that these should be at the same potential, but often without the understanding that these aren't really "grounds" either. It seems like the only time I hear someone refer to "Earth ground" is in the lab -- usually when probing a circuit node with a scope probe and unwittingly connecting the Earth ground of the test equipment to the common of the circuit. Thanks for reminding everyone about this issue. I like your final sentence -- this is a great rule for all engineers to remember!
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...